ORNL-TM-3039.txt

 

 

o

S e

 

ORNL-TM-3039
Contract No.'WeTHOS-eng—26

Reactor Division

MSRE SYSTEMS AND COMPONENTS PERFORMANCE
by
| Molten-Salt Reactor Experiment Staff

Edited and Compiled by
R. H. Guymon

JUNE 1973

NOTICE: This document contains information of a preliminary
nature and was prepared primarily for internal use at the
Oak Ridge National Laboratory. It is subject to revision or
correction and therefore does not represent a final report.

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

 

‘ 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
i legal liability or tesponsibility 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, .

NOTICE

 

' 3‘% %SEER

 

 

 

DISTRIBUTION OF THIS DOCUMENT 1S UNL\MI_TE%)\

 

 
o i

. , 111
CONTENTS

ABSTRACT « « oo v v v e e e e e e e e et et eeen e 1
1. INTRODUCTION .+ v v o o o o o o o « o v o s s o o o s o v vy 3
2. DESCRIPTION OF THE PLANT '+ « « v « « o o o o o s o s o o v+ &
3.  CHRONOLOGY OF OPERATION AND MAINTENANCE . « « « o « s o s o« o 9
4. PLANT PERFORMANCE AND STATISTICS. « o o o o o o « o + o s o o 15

 

4.1 Cumulative StatisticS. o o « « o o o o o o o o o o o o 15

4.2 Availability during Various Periods. . 4« ¢« « &+ & « « 18
4.3 Interruptions of Operations. « « « o o o o o o ¢ o o o 22

4.4 Time Required for Operational Tasks. « o« o o o « o « o 48

4.5 Changes Made in the Plant. . . . . & e o s o s s o o 59

4,6 Tabulation of Recorded Variables at Full Power . . . . 63

5. FUEL SYSTEM &+ « ¢ o « o 5.0 o o o o o o« o s s o o o 5.0 o+ ¢ 89
5.1 Descriptione « o o o o o o ¢ o o o o o s 6 o o o o o » 89

5.2 Purging of Moisture and Oxygen from the System . . . . 89

5.3 Fuel-Circulating-System Volume Calibration , « « « . . 91

5.4 Drain TImeS. « o« o o o s o o o o o s s o s s o o o o o 91

- 5.5 Mixing of Fuel and Flush Salts . o 0 4+ & & « o « o« » 93
\=j 5.6 Primary System Leak. & ¢ ¢ o « s e o o % o e s 2 s o o+ 93
5.7 Operation '« « o ¢ o s o o 0 o 0 s 0 g 0 o 0 o 0 o s 94

. 5.8 Fuel Pump and Overflow Tank. « « « o o o o o ¢ « o o o 94
5.9 Primary Heat Exchanger « « « o« o o o o s o » ¢ o o o « 110
"5.10 Reactor Vessel . and Reactor Access Nozzle , o « ¢ + o o 112

- 5.11  Fuel and Flush Salt Drain' Tanks. ¢ « + ¢ o ¢ ¢ ¢ o =+ ¢ 129

6. COOLANT SYSTEM & &« o o « o o o o o o o s o s e.s ¢ o o o o '« 137
6.1 DeSCription « « o « o ¢ o o o 0 0 o o s s o o o o o o 137

6.2 Purging Moisture and Oxygen from the System. . . o . 137

6.3 Coolant Circulating System Calibration and Drain Time, 137

6.4 Operation....'........'...".oo...-q139
6.5 Coolant Salt Circulating PUmp. « « o o o .o ¢ o o ¢ o o 139

6.6 Radiator..............,.7....—...,.140
6.7 Performance of Main Blowers, MB-1 and MB«3 , , . ., o o 147

6.8 Coolant Drain Tank o« o o « « + .+ o o s 0 s + 0 .o ¢ ¢ ¢ 159

7. COVER-GAS SYSTEM ¢ o e e o oo o & o o s o s o s s o o o + o161

7.1 Initial Testing . .7 e ® e ® ® 8 o v s e e e s e e ¢ 161
7.2 Normal Operation e« + o o + @ « o s o o o o o o ¢ v o o 162
7.3 Conclusions and Recommendations. . « o o + « v o + o o 164

 

 

 
10.

11.

12,

13.

14,

OO WOW\O W
- 3

 

iv

OFF-GAS SYSTEM L e ‘s & @ ; * e e o e e e e

8.1

™ 00 o ™
nHWwN

.

8.6

FUEL AND COOLANT PUMP LUBE OIL SYSTEMS. o o o « ¢ o »

o - .
vk

WO WO WO\
- .

=0 00N
[l =]

Description. . - | ] * L J - » & - . - o & -

Experience with Coolant Salt Off-gas System.

Experience with the Fuel 0Off-Gas System,

Difficulties with the Fuel Off-gas System. .

Subsequent Operating Experience with Fuel 0Off-gas

System -* * - - * - @ * L - - * o ._ * . *

Discussion and Conclusions « « o« o o o o o o

Description. s J'.,. * % s e e e e * s » ¢ s
Installation and Early Problems. . « « « « + &

 Addition of Syphon Tanks to the 0il Catch Tanks. .
Oil Leakage. . ; e o @ ; 0 * & & = s & 8 09 0:0 . @
Change~out of O11 PumpS. o« « « o o o o o o o o o &
" Test Check of One 0il System Supplying Both the Fuel

- » *

® » . - -

and Coolant Salt PumpsS « + « « o ¢ o o o = o s s o o
0il Temperature Problems . . « « ¢ o o o ¢ o o o o s o
Packages.

Increase in Radiation Levels at the Lube
AnaIYSis Of 011. t.o » !-n * * & & & 2
Replacement of 011 . . ¢« & & ¢ o« ¢ « o &
Discussion and Recommendations . « « o+ »

COMPONENT COOLING SYSTEMS &« o o ¢ o s s o o s .0

10.1
10.2

Primary System . « « o« o« ¢ o .¢ ¢ s ¢ o &

Secondary System ¢ ¢ o e s s oo ; . s s

VENTILATION SYSTEM « « o o o o o o o o o s ‘o

11.1
11.2

11.3

Description. . + ¢« ¢ ¢« o o ¢ o ¢ ¢ o o o

Operating Experience .. « « ¢« ¢« ¢« « « o &

Conclusions and Recommendatfons. . « « o+

WATER SYSTEm » * * - . .. 2 - - L] - . - » . @ . -

12.1

- 12.2

12.3
12.4
12,5
12.6
12.7

LIQUID
13.1
13.2
13.3

Potable Water System . . .
Process Water System , ., .
Cooling Tower Water System
Treated Water System . . .
Condensate System . . « o o &
Nuclear Instrument. Penetration. .
General Water Systems Conclusions.

WASTE SYSTEM &+ « o « o o s o o = o o ¢

* * e -
* o & @
e o e & e

Description. « « o« ¢ ¢ o ¢ ¢ o o s o o o

Preliminary Testing.“. s s » {’o . s o e
Operating Experience . « ¢« « ¢ s s o« o &

HELIUM LEAK DETECTOR SYSTEM . » o« & ¢ o o s o &

14.1
14.2
14.3
14.4

Description and Method of Operation. . .
Calibration . . . ¢ « ¢ o s o s ¢ « o
Operating Experience . « « o« o » o o +
Discussion and Recommendations . . . . .

0il

v *
e ®

. 8 . -

e 8 e e

e & e @

220
222

223
223

231
232

232
. 232
. 235
237
237

238
239
240
245
245,
246

- 247

247

. 247
248

251

« 251
- 251
- 255

256
 

15.

17.

18.

"16.1 Description. « « + « o o o
16.2 Alternating Current System
16.3 -250—V dc System. . % e e »

- 16.4 Reliable Power System. . .
.. 16.5 48-V dc System . . + . . .
16.6 Diesel Generators. . . . .
16.7 100-kVa Variable Frequency
16.8 ConclusionS. . « « o « o &
HEATERS , & ¢« ¢ & ¢ o o« o s ¢ o o

- 17.1 Description. + « « « + . .
17.2 Preoperational Checkout., .
~17.3 System Heatup, . « + « + &

~17.4 Heater Performance . . . .
-17.5 Systems Cooldown Rate. . .
17.6 Discussion . . ¢« « « & o o

'SAMPLERS. et e e eie 4w e we s

18.1 Fuel Sampler—Enricher. . o
18.2 Coolant Sampler. . « « + o

19.

- 20.

INSTRUMENT AIR SYSTEM . « o o « o

15.1
15.2

- 15.3
16,

Description . . + « & + &
Operating Experience . . .
Conclusions, . « + o« & « &

ELECTRICAL SYSTEM . e e 9 * & e

CONTRDL RODS; *® » s 2 & & o =
. % :
2 19.1

19.2
19.3
19.4
19.5
19.6

19.7
- -FREEZE

20.1
20.2

20.3

"7.20‘4_

20.5

FREEZE

21.1
21.2
21.3

Description., « + « + &
Initial Testing. . « + &
Periodic Testing . . .
Operating Experience ,
Improved Rod Scram Testing

‘Maintenance Experience . .

.

. - * . - ..

Discussion and Recommendations
VALVES . ; . .-.'.'; '7‘ . ; .
IntrOduction é.j ;@c " s 4 s s & 8 s e

the Freeze Valves
of the Freeze

~ Description of the Design of

Description of the Operation

Operating Experience , . .
Recommendations.-.-. o 6

FLANGES. .f;:.-.i!TF o“o ;

Description. « « « + & « &

Operation. ¢ o . ¢ 8 8 & u
Conclusions. . + ¢ « o« ¢ &

*

- . - . - *»

or—Generator

. - . . . . -

* * * . » o -

. [ .

-

*

&

* . * . » -

e e e e o » .

» o ® e » e

® . * . L] * *

- . * . ®

- » . * ® . .

* * - . - . - *

e o * o =

- *

Valves.

.

. » . - - .

 

e o — s i et
 

vi

22. CONTAIMNT * - » - . » - ) . * - o & - . * - . * - * . * ® . -

22.1
22.2

22.3
22.4
22.5

Description and Criteria .« . « o « o o o « o o o o o &
Methods Used to Assure Adequate Containment and
ResultSe « o s o o ¢« ¢ o ¢ o o o o o 5 o 5 o s s o »
Discussion of Cell Leak Rate Determination
Vapor Condensing System . . ¢« ¢ ¢« ¢ o « &
Recommendations. . « « o« « ¢« ¢« ¢ o« ¢ o o &

23. BIOLOGICAL SHIELDING AND RADIATION LEVELS . o & « ¢ o ¢ o « &

- 23.1
23.2

24. INSTRUMTATION .. ® [ . - * e . * @ * - * . . * ® » .

24.1
24.2
24.3
24.4
24.5
24.6
24.7

24.8
24.9
24.10
24.11

24.12

- 24.13

- 24.14
- 24.15
24.16
24.17
24.18
24.19
24.20
24.21

REFERENCES

. Description- * & s 2 s & o s e . & @

Radiation Surveys — Approach to Power. « « o« « o« o o o

Radiation Levels During Operation. . « « « s o s s & &
ConclusionS.: « + « s o o ¢ o o s s o o o o ¢ o o o » o

IntrOduction . - . » - » » » » » * e L

Initial Checkout and Startup Tests . .
Periodic Testing . « « ¢ ¢ ¢ s ¢ o « o«
Performance of the Nuclear Safety Instrumentation.
Performance of the Wide-~Range Counting Channels. .
Performance of the Linear Power Chamnels . . « . .
BFs Nuclear Instrumentation. . . + ¢« o + & ¢ o o o«
Nuclear Instrument Penetration . . « « « « ¢ o +
Performance of the Process Radiation Monitors. . . .
Performance of the Personnel Radiation Monitoring and
Building Evacuation System . « « « o ¢ o o ¢ ¢ o o o
Performance of the Stack Monitoring System . . . . . .
Performance of Thermocouples and the Temperature
Readout and Control Instrumentation. . . . . .
Performance of Pressure Detectors. . « « « « &«
Performance of Level Indicators ., . . . . . .
Performance of the Drain Tank Weighing Systems -
Performance of the Coolant Salt Flowmeters
Performance of Relays. . « « « . . .
Training Simulation . . « « ¢ « &
Miscellaneous ., « o« ¢ o ¢ & o o
Conclusions and Recommendations. .

- * * * . . . * .

-
. . - . » [ ] - .
* L] » a - e » -
* - . - . . & * *

. -2 - > * . . . - -

* - ® - s .9 - - .

403
403

403

404
404

404
408

408

409
410

411

411
412
412

413
 

MSRE SYSTEMS AND COMPONENTS PERFORMANCE

ABSTRACT

When the MSRE was shut down in December 1969, it had accumulated
13,172 full-power hours of operation. Salt had been circulated in the
fuel system for 21 788 hours and in the coolant system for 26, 076 hours.

_ Essentially no difficulty was encountered with the primary system
during operation. After the reactor was shut down there was an indication
of a leak in the drain-tank piping at/or near a freeze valve. Further in-
vestigation will be maderlater as to the nature and cause of this leak.

There was a small continuous 1eakage of lubricating oil into the fuel

pump throughout the operation. This, together with salt. mist, caused peri-

odic plugging in the off—gas system which was designed for clean helium,
Filters installed in the main lines proved very effective.,

In early operation, difficulty was encountered with the coolant radi-

~ator. The doors would not seal, there were too many air leaks, and thermal

insulation was inadequate. After these were repaired, the system operated

fine except for a failure of one of the main blowers and some trouble with

the blower bearings. , . , ,
Only relatively minor difficulties were encountered with the contain-

ment and other systems.

 

 
 

1. INTRODUCTION
P. N. Haubenreich

Operation of the MSRE constituted a major step toward the objectives
of the Molten-Salt‘Reactor Program. The g6a1 of this program is the de-

- velopment of'large, fluid-fuel reactors having good neutron economy and
producing low-cost electricity.1 The MSRE was built to demonstrate the
practicality of the molten-salt reactor concept with emphasis on the
compatibility of the materials (fluoride salts, graphite, ‘and container
ralloy) the performance of key components, and the reliability and main-
tainability of the plant.

In the course of 5 years of testing and operation of the MSRE (196h -
1969) the operators sccumulated con51derable experience with the various
components - and systems in the reactor plant. Tnis experience, properly
disseminated, should be valuable in the continuing development of molten-
salt reactors. Much has already been published in tne Molten Salt Reactor
Proéram semiannual progress reports (Refs. 2 to 14) but such reporting is
piecemeal and sometimes rathervcondensed. On the other hand, there is
much very detailed information in test reports and operations and mainte-
- nance files, but these are relatively inaccessible and specific informa-
tion is tedious to extract. It was considered.worthWhile therefore, to
extract, organize ‘evaluate and report the experience with MSRE systems
and components. o

The purpose of this report is to present a convenient, comprehen51ve
description of the MSRE experience. The first chapters.briefly describe

the plent end outline the chronology of its'Operation. Next there is a
3 chapter on the overall plant performance, 1nclud1ng statistics relative
to reliability and maintainabillty. The chapters which follow are each
devoted to one system or component. Finally, there‘is a chapter of dis--

cussion and conclusions.
 

2. DESCRIPTION OF THE PLANT

R. H. Guymon

The MSRE was a single—rigion;-circulating molten-salt'fueled‘ thermal
reactor which produced heat at the rate of about 8 Mw. The fuel was UFy
in a carrier salt of L1F-BeF2-ZrFq. At the operating temperature of 1200°F,
this salt is a liquid which has very good physical propertles: viecosity
ebout 8 cent1p01se den51ty about 135 Ib/ft3, and'vapor pressure less than
0.1 mm Hg. | | |

~ The design conditions are shown in the flow diagram (Fig. 2--1). The

general arrangement of the'plant is shown in Fig. 2-2. The salt-containing
~piping and equipment was made of Hastelloy~N, a nickel—molyhdenumyiron-
‘chromium alloy with exceptional,resistance to corrosiOn'by molten fluorides
and with high strength at hlgh temperature |

In the reactor prlmary system, the fuel salt was . rec1rculated by the
sump-type centrifugal pumphthrough the shell—and-tube heat exchanger and
the reactor vessel. - The 5-ft diam. by78-ft high reactor vessel is shown
in Figure 2—3‘ Tt was filled with 2-in. by 2-in. graphite moderator stringers
which had grooves machlned in the sides to form flow channels ‘for the fuel ,
salt. ©Since the graphlte is compatlble with the.molten salt, 1t was possible
to use unclad graphlte which is deslrable to obtaln good neutron economy.
The heat generated in the fuelisalt as»it passedfthrough the reactor was
transferred in the heat exchanger to & molten LiF-BeF; coolant salt. The

"coolant salt was c1rculated by means of a second sump—type pump through

 

: the heat exchanger and through the radiator. Air was blown by two axial

flow blowers past the radlator tubes to remove the heat which was sent

-:up the coolant stack where 1t was dlSSlpated to the atmosphere. _ /
Drain tenks were prov1ded for storlng the fuel and coolant salts at

hlgh temperature when the reactor was not operatlng., LiF-BeF, flush salt

~ used for flushing the fuel system'before and after maintenance was stored (\\jd

in the .fuel flush tank. The salts were dralned by gravity. They were | (

transferred back to the circulatingrsystems by pressurizing the tanks with

helium.
 

 

STACK  FAN

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

U

| R . el

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

S g;mq 5 ","‘
| ol A coouant [
S | i ]enricher PUMP ;

o i
| { { J TO ABSOLUTE FlLTERs-.......i -
' | ‘ !- i 1018 *F ;
i =i 850 G.PM.
OFF-GAS
AT R
oow ' 1240°F _ —
' OVERFLOW TANK
ABSOLUTE 170 *F
FLTERS ' 120G GAM,
BLDG. REACTOR : 1075 F
VENTLATION . VESSEL | power- FREEZE FLANGE {TYR)
. 8 Mw
bl geress
L OLAN FREEZE VALVE (TYP)
o wmd  COOLANT ,
F‘ SYSTEM
4
J 'h
S8
-~ " RN A
= reeresTosreme
MAN
CHARCOAL
BED

 

 

 

 

ORNL-DWG 65-14108

LEGEND
ssmmmm FUEL SALT
— COOLANT SALT
stresreveneer mm CWER GAS
----- RADIGACTIVE OFF ~GAS

:

FH.TERS

- -

.

DRAIN
TANK

 

 

 

 

 

 

 
        

 

 

ORNL-DWG 63-1209R

 
 
  
 
 

REMOTE MAINTENANCE

 

 

  

 

 

  

 

CONTROL ROOM - |
i '
|
REACTOR CONTROL
ROOM _ ‘ 4
- . ‘l
S 6 —' -"
'. : ! _l'l
i j |
E; ‘ » ——
» —_—
AN
i
(,fi\
: J
9
8
1. REACTOR VESSEL 7. RADIATOR
2. HEAT EXCHANGER 8. COOLANT DRAIN TANK
3. FUEL PUMP 9, FANS
4. FREEZE FLANGE = 10. FUEL DRAIN TANKS
5. THERMAL SHIELD 1. FLUSH TANK
6. COOLANT PUMP 12, CONTAINMENT VESSEL

13. FREEZE VALVE

Fig. 2.2 Layout of the MSRE

 
 

ORNL-LR-DWG 61097R1A

FLEXIBLE CONDUIT TO
GRAPHITE SAMPLE ACCESS PORT CONTROL ROD DRIVES

COOLING AIR LINES

      
  

o
° __3_;
o ‘ 1 ACCESS PORT COOLING JACKETS
FUEL OUTLET r REACTOR ACCESS PORT
" CORE ROD THIMBLES N i SMALL GRAPHITE SAMPLES
ROD T! ==Y HOLD-DOWN ROD
LARGE GRAPHITE SAMPLES 1 OUTLET STRAINER
. &

 

CORE CENTERING GRID

 

 

 

FLOW DISTRIBUTOR

 

 

 

 

 

 

 

VOLUTE
GRAPHITE ~MODERATOR .
STRINGER ’
Ty AL
FueL iNLeT < (|
s T~ CORE WALL COOLING ANNULUS
REACTOR CORE Can —i|

REACTOR VESSEL

 

 

 

 

 

ANTI-SWIRL VANES — ' S
' e MODERATOR
VESSEL DRAIN LINE - SUPPORT GRID

Fig. 2.3 Details of the MSRE Core and Reactor Vessel
The fission product gases, krypton and xenon, were removed continfiously_
from the circulating fuel salt by spraying salt at a rate of 50 gpm into
the cover gas above the liquid in the fuel pump tank. There they trans-
ferredifrom the liquid to the gas phase and were swept out of the tank by
a small purge. of helium. After a delay of about 1-1/2 hr in the piping,
this gas passed through water-cooled beds of activated charcoal. The
krypton and xenon were delayed until all but the 8%Kr decayed and then
were diluted with air and discharged to the atmosphere. | |

The fuel and coolant systems were provided with equipment for taking
samples of the molten salt while the reactor was operating at power. The
fuel sampler was also used for adding small amounts of fuel to the reactor
while at power to compensate for burnup

The negative temperature coefficient of react1v1ty of the fuel and
graphite moderator made nuclear control of the system very simple. However,
three control rods were provided for adjusting temperature, compensating .
for buildup of fission products, and for shutdown.

The plant was provided with a simple processing facility for treating
full 75~ft? batches of fuel salt with hydrogen fluoride or fluorine gases.
The hydrogen fluoride treatment was for removing oxide contamination from |
the salt as Hy0. The fluorine treatment utilized the fluoride vblatility
process for removing the uranium as UF5.-

Auxiliary systems included: (1) a cover-gas system with treating
stations for remofiing dxygen and moisture from the helium cover gas;

(2) two closed-loop oil-systems for cooling the fuel and.coolant pumps
and providing lubrication to’tfié’bearings; (3) = closed 100p.component
coolant system for cooling the: control rods and other in-cell components;
(L) several cooling water systems including a closed-loop treated water
system for cooling certain in-cell équipment§ (5) a ventiletion system for
contamination control; andr(6) an instrument air system. ‘
A1l of the primary salf'éystém'was located in the reactor and drain
'tank cells. These sealed pressure vessels prov1ded secondary contalnment.

For a fuller descrlptlon of the plant, see Reference 15

 
 

 

 

3. CHRONOLOGY OF OPERATTON AND MAINTENANCE -
' . R. H. Guymon |

Design of the MSRE began .in the summer of 1960 and by August 196h
installation was far enough along to permit the planned non-nuclear
testing16 to begin. Milestones that were passed in the years which fol-
lowed are listed in Table 3-1.

Teble 3-1. Milestones in MSRE Operation

 

Salt first loaded into tanks |  October 2k, 196k

Salt first circulated through core. January 12, 1965
First crltlcallty (23%5y) . June 1, 1965
First operation in megawatt range January 24, 1966
Full power reached | - May 23, 1966
Nuclear Operatlon w1th 235U concluded ) - March 26; 1968$
Strip uranium from fuel salt August 23-29, 1968
First crltlcal with 233U p._ October 2, 1968
Reach full power with 233y | ~ January 28, 1969
Nuclear operation concluded - Decenber 12, 1969

 

The activities durlng the perlod of non—nuclear testlng, zero—power -
experlments, and preparation for power 0perat10n are outllned in Fig. 3-1.
During the prenuclear testlng, except for the usual startup troubles
 due to early fallures and instrument malfunctlons, 21l systems operated
- well and there vere no unant1C1pated problems in handling the molten
_salt. Some plugglng of the offgas system valves and fllters was encoun-
~ tered, but thls did not seem serlous.__
Crltlcallty was attalned‘by addlng UFu~L1F enrlchlng salt to the
'~eafr1er salt. More uran;um{was added to bring the fuel gradually to the
. desired operating concentratien'while the control rods were calibrated and
.feactivity:coeffiéients were measured. At the conclusion of'thefzero-
power nuclear experiments‘the.reactor was shut down to finish construction

of the vapor-condensing system, to carry out some inspection, and to

 
INSPECTION AND
'PRELIMINARY TESTING

MSRE ACTIVITIES

JULY 1964 — DECEMBER 1965

PRENUCLEAR TESTS
OF COM PLEATE SYSTEM

FINAL PREPARATIONS
" FOR POWERLOPERATION

 

” N

FINISH LEAKTEST, -
INSTALLATION PURGE & HEAT

OF SALT SYSTEMS SALT SYSTEMS

TEST AUX. SYSTEMS

. OPERATOR TRAINING

LOAD SALT

TEST TRANSFER, OPERATOR.

FILL & DRAIN OPS. TRAINING

 

— ’ Al

INSTALL INSTALL CORE SAMPLES
CONTROL RODS ‘ INSPECT FUEL PUMP

SAMPLER-ENRICHER HEAT-TREAT CORE VESSEL
277227 _ ;
TEST SECONDARY

' TAI
FINISH VAPOR-COND. SYSTEM CT

MODIFY CELL PENETRATIONS ADJUST & MODIFY

REPLACE RADIATOR DOORS RADIATOR ENCLOSURE
:

 

LOAD U-235 -
INTO DRAIN TANKS LOAD & IN ZERO-POWER - _ LOW-POWER
COOLANT FLUSH CIRCULATE CIRCULATE . NUCLEAR , ~ (0-50 kw)
SALT SALT C & FL- SALTS CARRIER EXPERIMENTS | EXPTS.
2222772 g
L } 1 { 4 | y | 1 { 1 1 § { . _
I I 1 RN 1 t I T b I 1 1 I 1 i I 1 1
J A S 0 N D J F M A M J J A S 0 N D
1964 1965

Fig. 3.1 MSRE Activities July 1964 — December 1965

-

0T

 
 

11

meke some repairs and modifications, including replacement of the heat-
warped radistor doors. rThe'firstfteSt,of the secondary containment was
made during this shutdown period

After tests in the kilcwatt range showed that the dynamics of the
system were as expected,,the approach to full power was started in
January 1966. Plugging in the fuel offgas system occurred immediately
after increasing the power tc one megewatt. Almost three mouths were
spent investigating and remedying the offgas prdblem-by'installation of
‘8 large, efficient filter;'”As:shown in Fig. 3-2, the power ascension was
‘resumed in April and full power, which was limited by the capsbility of
the,heat-removal system, wasjsttained in May.

High-pcwer,operation'wes'abruptly halted in July when the hub and
blades of one of the main.blcwers?in.the heat-removal system broke up.
While the reactor'was down, 'the arrsy of graphite and metal”specimens'was
removed from the core and & new array instelled. During the flushing
operations before the specimen removal the erl-pump bowl was accidentally
over-filled and some flush selt froze in the attached gas lines. ~ Tempo-
rary heaters were installed remotely to clear the lines and in November
the offgas line at the pump bowl was cleared by running a tool through it.

Most of the remaining operation with 235y was at power. Restrictions
continued to develop in the offgas system.but caused llttle delay in the
program. Shutdcwns were necessary to make repairs on the fuel sampler-

enricher, the component coolant pumps, and in-cell air line,disconnects

o whose leakage caused an indication of high cell leak rate. The core

r,:,specimens were replaced periodically, annual containment tests were made,

 

'_ and verious experiments conducted. Operation with 23°y fuel was termi-
nated on March 26, 1968 8o that the urenium could be stripped from the
fuel selt and replaced with 23311
During the shutdcwn which fbllowed the core speclmens were replaced
and the fuel piping and. vessels were surveyed using remote gamma ray spec-
r;troscOpy to determine the distribution of fission products. Necessary and -
pipreventive maintenance was also done in preparation for operatlon w1th
-:-i-zsaU fuel. | |
| After testing end extensive modification of the excess fluorine dis-

' posal system, the flush salt and fuel salt were processed to strip the

 
 

 

 

 

DYNAMIKCS TESTS -

INVESTIGATE
OFFGAS PLUGGING

REPLACE WALVES
AND FILTERS

| RaISE POWER

} REPWR SAMPLER

| ATTAN FULL POWER
} ek conTammenT

 

FULL~POWER RUN
-— MAN BLOWER FALURE

REPLACE CORE SAMPLES
TEST CONTAINMENT

 

AuUsH ]

12

FUEL

SALT N
FUEL LOOP POWER

 

 

0

RN

FLusH ]

 

 

2 4 668 10
POWER (Mw) °

 

 

ORNL-0WG 69~ T293R2

HIGH-POWER OPERATION
TO MEASURE & fag

INVESTIGATE COVER GAS,
XENON, AND FISSION
PRODUCT BEHAVIOR

ADD PLUTONIUM
RRADIATE ENCAPSULATED U

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

REMOVE CORE ARRAY
PUT REACTOR IN STANDBY

Fig. 3.2 Outline of the Four Years of MSRE Power OperatAion
 

 

 

 

13

ursnium and remove théfcorroeion products produced during the fluorina-
tion.!?
attached to a drain tank began immediately thereafter. The cells were

closed and leak-testednbefore the reactor was made crltlcal by small ad-
ditions of 233y through the sampler-enrlcher. ‘On October 8, 1968, Glenn

Seaborg manipulated the controls ‘to raise 'the power to 100 kw maklng -
233 S
U.

Loading of 233U into the carrier salt through spec1a1 equlpment

the MSRE the world's first reactor to operate on

Early in the zero-power phy31cs experlments, the amount of éas en-
trained in the;circulatingbfuel had increased from <0.1 to ~ 0.5 vol %.
Investigation of this, as well'as dgifficulties with the sampler-enricher
end plugglng in the offgas system, delayed the approach to full power |
until Jenuary 1969. - | |
| As the power was 1ncreased into the megawatt range, ‘there were ob-
served for the first time sporadic small increases (~ 5 to 10%) in nuclear
power for a few seconds, occurring with & varying frequency somewhere '
~around 10/hr. The characteristics of the transients pointed to changes
- in gas volume inh the fuel loop and it appeared that they were most 1ihely"
caused by occasional release of some gas that collected 1n the core.wln
This hypothesis was -supported when, late in February, a varlable frequency
generator was used to operate the fuel pump &t reduced speed. The gas
ent_rainment in the'ruel circul_atfng loop decreased sharply (from 0.7 to
?O.l vol %)-and the perturbations ceased entirely.’

Operations continued at various nuclear powers and. fuel-pump speeds
until Junerlnwhen the reactormfias shut down to replace the core specimens,
investigate the'distribution:offifission products-in the-prinary loop by
,gammafscanning, remove the'offgas restrictione and-test the secondary
containment. For the first tlme one of the control rods dld not scram
h‘l,and it was replaced durlng the shutdown. | e o
77; The remaining months of operet1on were spent in various studles of
’{1the behavior of tritium, xenon, and certain other fission products .in. ther

'reactor.:fln Novemberflt became ev1dent that . sufflclent funds would-.not
7 be“available_to'continne operation and on December 12, 1969 nuclear’ '
operation wes concluded. Another flrst occurred shortly after the fuel

was drained'when a small amount of flssion—product activity appeared in

 
 

 

14

the cell atmosphere, indicating a leak in the primary system apparently
near a freeze valve. , | | |

| After the final shutdowfi, fihe facility was_placed_in a standby.con—
dition to await post-operation examinatiohs plahned for early in the next
fiscal year.19 The operating crewsrwere disbanded and overrthe next six
- months most of the engineers were reassigned as they completed analyses
and reporting of the reactor expefience. As of this writing, the post-
operation examination has not been accompiished.

- In this report and elsewhere MSRE run numbers are often used to
identify the period of operation. Generally a new run number was assigned
at the end of a major shutdown or when there.were substantial changes in
. the purpose or type df operation. The starting dates for eaéh of the 20

runs are listed in Table 3-2,.

Table 3-2. Dates 6f MSRE Runs

 

 

Run No. ' Starting Date® Run No. ' Starting Date®
1 January 9, 1965 11 January 24, 1967
2 Msy 11, 1965 12 June 8, 1967
3 May 31, 1965 13 September 3, 1967
L December 5, 1965 1k September 19, 1967
5 February T, 1966 15 August 1k, 1968
6 March 26, 1966 16 December 10, 1968
T June 11, 1966 | 17 January 10, 1969
8 September 1k, 1966 18 April 11, 1969
9 November 6, 1966 19 July 31, 1969

10 December 6, 1966 - 20 ' November 20, 1969

 

Slhese are the dates on which samples, logs, etc., began to receive
new numbers, and are not generally the dates of reactor startup.
15

4. PLANT PERFORMANCE AND STATISTICS
R. H. Guymon ~P. N. Haubenreich

"They 've kept the darned thing running and when they
shut down they can get back on the line; you can't
knock that!" An unidentified "AEC spokesman', ag -
_quoted in October 12, 1967 Nucleonics Week.

"So far the Molten Salt Reactor Emperzment has operated
successfully and has earned a reputation for relia- .
bility." USAEC Chairman Glenn T. Seaborg at the
‘ceremony marking the first operatton of a reactor
fueled'wtth 233y, October 8, 1968.

The MSRE ran long and it ran well. It ran long enough with 235U .,

~ fuel to attaan the original goals of the experlment then continued

through more than a year of added experlments with 233U fuel. These
statements are supported by the statlstlcs on the overall plant performr
ange-presented in this chapter.. Later chapters describe the performance

of individual components and systems which made up-the plant.

4.1 Cumulative Statistics

Several different\ipdiggtions of how long the MSRE ran ére’presented

in Teble L-1. The time critical is a commonly used index for reactor ex-

- periments. Another basis of cOmfiarison With“othér reactors is the number

of equivalent full-power hpurs; _Integrated power (Mw-hrs) is closely re-

, léted. (The figures quoted'fiéféfaré'based on 8 full power of 8.0 Mw,
- which is the value 1nd1cated by heat balances.20 Analyses of isotopic
-changes 1nd1cate that full pcwer was 7.34 Mv, 1n which case the tabu-
 -_lated Mw—hrs should be mnltlplled.by 0.92.). The amounts of tlme that
o salt circuleted in the loops are of 1nterest from the standp01nt of the
 'demonstrat1on of materlals compatlbility.- Each pump operated: a length of '
':tlme equal to the salt circulat1on plus helium circulation in that loop.

. The number of thermal cycles of various kinds on dlfferent parts of

the salt system are llsted 1n Table 4-2, The numbers are small in com-

p&rlson with the perm1581ble numbers of cycles except in the case of the

 
 

 

 

16

Table 4-1. Accumulated Operating Data

 

Time Critical (hrs)
Integrated Power (Mw-hrs)®
Equivalent Full-Power Hours

Salt Circulating Time (hrs)
Fuel Loop
Coolant Loop

Helium Circulating Time (hrs)

~ Fuel Loop |

Coolant Loop

Time Above 900°F (hrs)
Fuel Loqp |
Coolant Loop

Fill and Dreain Cycles

Fuel Loop
Coolant Loop

 

235y operation 23%U Operation Total
11,515 6,140 17,655
72,401 33,296 105,737
9,005 4,167 13,172
15,042 6,746 21,788
16,906 9,170 26,026
4,046 3,384 7,430
3,172 1,535 4,707
20,789 10,059 30,848
17,44Y 9,99k 27,438
| 37 14 51
13 6 19

 

BBased dn heat balances which indicated full power was 8.0 Mw.
 

 

Table 4,2 MSRE Cumulative Cycle History

(. ‘Thaw Usage
_ ' ! | | : Quench ‘& | Factor
Component, - 'Heat/Cool |Fill/Drain|Power|Quench , Time . On/Off| Thaw: Trans. &

 

- *These figfires are based on the original éalculations} If they were based on freeze flangé thermal
cycle tests, the usage factors would be 23.04% and 13.37%. '

LT

 
 

. 18

| freeze flanges. Here the fuel flanges experienced 99% of the permissible
number of cycles based on early calculations or 23% ‘of the number perm1351-
ble on the basis of extended tests of & prototype flange.

In order to put the MSRE record in proper perspective, 1t is neces-
sary to measure it against those of other reactors in a similar stage of

development. A comparison on the basis of equlvalent full-power hours is

made in Table 4-3. It is not invidious to say that the MSRE compares well

with other reactors having the same purpose that is, to demonstrate the

.practlcality of a reactor concept.
4.2 Availability during Various Periods

The best index of the rellabillty of a plant should be the fractlon
of time (over some extended period) that it is available for its intended
service. Perhaps inevitably, however, in experimental,plants where the
. obJectives include.more:than simply_generating‘pcwer, the definition of
-availability'is not always s0 clearcut. Therefore we heve presented in
Table U-4 an index which, in principle, isfiless significantfibut-whose' A
definition is quite clear; namely,.the time that the réactor wes actually
critical in-each 3-month period during the 4 years of power operation.
Percentages of elapsed time are shown for selected intervals, but these
must not be regarded as e measure of reliebility since the reactor was
suberitical much of the tlme because the planned program requlred it.
',(Durlng shutdowns for core specimen removal or the substitution of
tzaau for example.)
~ In the MSRE test progra.m16 it was planned that there be a perlod of
’ 0peratlon for the prlmary purpose of demonstratlng plant rellablllty.
_iThlS phase of the program covered the last 15 months of operation with

235y, Table 4-5 gives a'breakdown of the tlme durlng thls period. The
 reactor was critical 80% of the time and the . avallablllty of the plant,
as defined in this table, was 86% This amply Justifled Dr. Sesgborg' s |
statement quoted at the head of this chapter. ‘

 Another indication of rellabllity is how long a plant can be kept
continuously on line. In the final run with zasU, the fuel salt was in

the loop continuously for just over 6 months, before it was draimed for

“
19

Teble 4.3 - Equivalent Full-Power Hours®
Produced by Early U. S. Reactors ‘of Several Types

 

- Operation

 

e w First Interval L
Type Reactor Critical Terminated = {years) -~ EFPH
Aqueous - _
Homogeneous HRE-2 12/57 5/61 3.4 3,100 -
Organlc-Cooled OMRE 9/57 . 4/63 5.6 5,934
" Piqua 6/63 " 1/66 2.6 5,642
Sodium-Graphite 'SRE W57 2/6h 6.8  8,1h0
co . Hallam 8/62 9/6k4 2.1 2 661
HTGR Peach Bottom  3/66 ' ’ib 3.8 10,836°
LMFBR EBR-1 8/51 | 'ié/62 11.3 5 sohb
EBR-2 11/63 b 6.1 11,713,
Fermi 8/63 b 6.3 102
PWR Shippingport . o
(Core #1) 12/57 2/6k 6.2 27,781
BWR EBWR 12/56 6/67 10.5 11,164
VBWR 8/57  12/63 6.3 11,814
MSRE 6/65  12/69 4.5 13,172

MSR

 

®Calculated from.thermal'MW—hrs and installed capaoities reported
in USAEC-DRDT booklet "Operating History of U. S. Nuclear Power Reactors —
"'1969" except Shlppingport Core—l data from April 1964 Nucleonics.

bOperation is not yet termlnated.

EFPH quoted is through 1969,

Total for Shipplngport (with two cores) through 1969 is h3 400 EFPH.

 

 
 

 

20

Table 4-4 Time Critical Each Quarter during the
Four Years of Power Operation (1966 - 1969)

 

Hours ‘ - Critical Time
- Year - Quarter = Critical Elapsed Time

 

62 v | )
1070

413
1221

1066

& w o

1852 | ’ T4

' 235
1186 73.9%
1292 (1967)

| « 9T.2%
21k (Qtr.)

1967

 

= w N R

20L45
0
0

735

1968

Eow o e
_,i

1800
1375
1054
1176

1969 61.5% 56%

- (1969) (233y)

= w e

 
 

 

21

- - Tdble 45 Breskdown of Time during Sustained Operation Phase: =~
Of MSRE Program (December 1k, 1966 - March 26, 1968)

 

 

Activity or Condition - - - Time - 'sePercentsge
 Criticel - 8934 nr  79.6
Available "~ Changing specimens - o 26d 5.6 086.3
- 'Annual tests - 54 1. 1
_Air-line disconnects (1/67) 13 d 2.8
| Sampler latch (8/67) | 35 d 7.5
Maintenance Component cooling pump (9/67T) 3d 0.6 1%?
Sempler wiring (12/67)-_: ... -.34a - 0.6}
e r View in. reactor cell (6/67) 6.4 1.3
Q"ther__-. 1 Mlscells.neous o Lo oo 0.9 2.2

 

'I'otal. ela.psed o S | | k68 4 ‘100.0_ |

 

 
 

 

22

| the planned processing. During these 188 days the reactor was counted as

unavailsble only 63 h (1.4% of the time) while repairs were being made to
the fuel-sampler drive. It was actually critical for 97.8% of the time.

The fingl phase of the MSRE operation, with 233y fuel, was not aimed 
Primarily at demonstrating reliability,21 but it turned out that the availa-
bility of the plant was still remerkably high. This is shown in Teble 4-6.
Here available time includes (in addition to the critical time) subcriti-
cal intervals during the zero-power experiments, time spent in changing
the pump power supply during varidble—épeed;tests, changing the core speci-
mens, gamma-scanning the drained salt systems, and’ experiments on behavior
of gas in flush salt. |

4.3 'Interrfiptions of Operations

FiSUre 4-1 provides a broad view of the MSRE operation and major ifiter—
ruptions. This figure was prepared for inclusion (together with similar
graphs from many other reactors) in the USAEC's annual presentation to the

JCAE and, according to the rules, assigns & brief, descriptive reason for

. each shutdown of 5 days or more. - A great deal more detail as to the nature .

and cause of these and shorter interruptions in operation is given in the

) figures and 11 tables which follow.

Figure L-2 covers 1966, the first year of po#er operation;- It shows
vhen salt was in the fuel loop, when the reactor was at power, and assigns
& number to each interruption in either. A number in a circle means the
interruption was due to some experiment. A number in a square meané that
the'interruption was necessary because of trouble with some system or com-
ponent. Figures b-3, 4-h, and 4-5 display the same kind of information
for 1967, 1968, and 1969 respectively. |

In Tabies L-7 through 4-10, the interruptions in operation during
each year are.categorized as to the type of interruption, the cause, and
other work that was accomplished during the interruption. Each interrup-
tion listed in these four tables is described more fully in Tables L4-11
through U-1k.
 

(:W

23

Teble 4-6 Time Critical and Time Available During 233U Phase

Of MSRE Program (September 10, 1968 - December 12, 1969)

Hours

Percentage

 

 

 

 Overall (9/10/68 - 12/12/69) 10959

 

 

| Periéd; | 7: N Elapsed Crit. Avail. Crit. Avail.
Zero-power expts.’(9/10/68; 11/28/68) 1895 649 - 1852 3h.2  9T.7
Interim (11/28/68 - 1/12/69) 1092 8 110 7.9 10.1
First Power Runs (1/12/69 - 6/1/69) 3355 -~ 3175 3268 9.6  97.k
Shutdown (6/1/69) - 8/9/69) 16k 0 8 0 24,8
" Later Power Runs (8/10/69 - 12/12/69) 297h ~ 2230 2971 T75.0 . 99.9
6140 8609 56.0  78.6

 

 
 

 

120

110

100

70

CUMULATIVE THERMAL MEGAWATT—HOURS (in thousands)
3

20

10

 

ORNL-DWG 73-691

 

AN =

oo, h

10.

11.
12,

13.

14.

15.

18,

17.

. Remove plugged capillary, check valve and filter

. Repair electrical wiring short in fusl sample enricher,
. Leak—test reactor cell.
. Remove and replace core—sample array; replace ons

. Clear ges line plug caused by salt overfill; replace
. Clear salt plug from gas line; repair air valve in

. Replace leaking sir-line disconnects in reactor cell,

MAJOR SHUTDOWN PERIODS

 

Low power testing.

in offgas system.

Identify plugging material and install larger
valves and newly designed filter in offgas system,

 

 

blower.

second blower,

reactor cell.

MSRE -

Initial Criticafity
Critical on 233y

Installed Capacity
Reactor Thermal Mw

Power Generation .
1989 Thermal Mwh

TYOTAL TMwh

6/1/65
10/2/68

8
33,283

105,737

 

 

s/

 

 

Remove and replace core—sample array; annual tests
of containment and controls.

131_4_/

 

 

Replace fuel sampler drive; retrieve sample fatch,

Remove and replace core—sample array. Annual
tests. Shake down 9533"""' plant. Process salt
to remove U. Load U.

 

Mix sait. Repair component cooling blower.
Repair fuel sampler drive. Service control—rod
drive.

Remove and replace cors—sample array, Replace

 

control rod and drive. Clear offgas line. Annual
tests. :

Measure fission product deposition.
Conclude nuclear operations 12/12/69.

 

10 A

 

 

I .
'
t .
‘
¢
'
i
4

i - : r7/'-

 

 

_ /...fi..l
K 2 3 ,/?/

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1965 Dec 1966 Dec

1987

 

Dec

1968

Dec 1960 Dec

Fig. 4.1 1Integrated Power and Major Shutdowns of the MSRE

C

we

 
 

 

25

INVESTIGATE OFF-GAS
PLUGGING — CHANGE
FILTERS AND VALVES

LOW-POWER
DYNAMICS TESTS

  

 

 

 

 

 

   

   
      
 

  

        
   

 

 

 

 

 

 

 

 

 

 

GO TO

- FULL POWER

REPAIR
SAMPLER

 

 

 

 

 

CHECK

OPERATE -
AT POWER

CONTAINMENT

 

 

 

 

 

 

et e e o s
1

 

 

 

 

REMOVE CORE SPECIMENS
REPLACE MAIN BLOWER

THAW FROZEN LINES
TEST CONTAINMENT

 

 

 

 

 

OPERATE

AT POWER

 

 

REMOVE
SALT PLUG

CHECK

CONTAINMET

 

OPERATE

 

 

T POWER

 

ORNL-DWG 73-592

“.POWER (MW)

BRFUEL o
~ JFusw .

 

 

 

o |
o
z3 |
c pEEEl R
, 20:0:0:0:0:0:0.0 2
15
31 15 ‘ |
1966 JAN FEB MAR APR | MAY JUN JUL AUG SEP ocCT
Fig. 4.2 1Interruptions of Operations — Year 1966

 

   

 

 
ORNL—-DWG 73-593

 

 

 

 

 

 

 

 

 

 

 

HSN14
13Nnd

(M) 43mod

 

 

xQ 0 < ~N
. n‘-- T iF

 
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

    

 

  

   

 

 

 

 

 

 

  

 

 

  

 

 

    

 

  

 

 

 

 

   

  

 

                           

 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

                                         
                          

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NI LIVS

 

 

 

 

 

 

 

 

1.
™M
. QO
- o
O
™
&
o w0 3
xcHo - >
<o =
o=
W = v
ey ™
T
i Q
- o
o
™
a ~
O w K
- W -
9
]
- O
— ™M ™
& |
R_h_.._._ "
$2 o 9
Q. -
EA —— = H..
e w. m
3 — Q
™ &
°
= o ”
2 3 3
L
N o
Q.
e ) o =
N ™ ".... m
: o O =) - |
[ N = 3 w
& & -
. M 0 . ™)
y  Z 8! —= ‘
o I )
Cm [ H :
82 S 2 %
o -— .,
P 01[ =
BEQ = . w
® F & g2
(2 53
Peteds
155 o
OO o _
@ 50X b ;
e e O ;
< 00582
— 500K
05
o XK :
-1 [
o™N ’ v“’"’ :
— G .m
= . «
Q =
1‘
N
———— |8
=) = S
. o
) 2
0o
-
_M () ™
N
wen . i
=u
mwm mn =
ATL ™ - <
B8 3 )
222 ®
H3IMOd 4001 13nd Q
POWER

FUEL LOOP

SALT IN

1968

XENON STRIPPING
EXPERIMENTS

 

 

 

 

 

 

 

 

 

 

 

27

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

INSPECTION AND MAINTENANCE
REPLACE CORE SPECIMENS

TEST AND MODIFY FLUORINE
DISPOSAL SYSTEM

LOAD URANIUM-233

e i t———

ORNL—DWG 73-594

U—233 ZERO POWER PHYSICS EXPERIMENT

1
i
l
§
i

C
(

PROCESS FLUSH SALT —~ PROCESS FUEL SALT

 

 

 

 

INVESTIGATE BEHAV

R
C

 

 

 

 

IOR FUEL SALT -

CLEAR OFF GAS LINES

EPAIR SAMPLER AND
DNTROL ROD DRIVE

 

 

 

 

L
|
: _—6—3-
| =
—a <
3
O.
2
0
15
;
I | | | | | o | | | |
5 B 3 B g B g 15 g 1B g 15 g 05 5 B
APR MAY JUN JuL AUG SEP ocT NOV DEC

Fig. 4.4 Interrti;lptions' of Operations — Year 1968

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FUEL

FLUSH

 

 
28

ORNL—DWG 73—5095

- -~ HIGH POWER REPLACE CORE SPECIMENS | INVESTIGATE COVER GAS, MEASURE
DYNAMICS  INVESTIGATE - OVERATIONS REPAIR ROD DRIVES XENON, AND FISSION TRITIUM
| CLEAR OFF GAS LINES PRODUCT BEHAVIOR
TESTS GAS IN FUEL LOOP SRR
GAMMA REMOVE CORE
SCAN SPECIMENS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘ @)
17 1819@ (24)

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

|—8
e () .
=
s
‘g
=
@
9 & -
ooy Nttt gl sassneeegoonre frrscnetetatonieieler ] Hoed Jonisleisluielvieieletd JRoelt e gt .'E':::.::':: 0 . T
5 16] 18 27) - 3
z3 | * - 2 T
35 ' | i Al
32 | | | |
| | I l | I ' I | !5 | | | 1I5 | I 1|5 ' 1l5 I ‘.)1|5 | I 115 1|5 15 |
15 31 15 28 15 31 18 30 1 - 31 . -30 | 3 31 30 31 30 31

 

1969 JAN FEB MAR APR MAY JUN JUL - AUG SEP oCT NOV | DEC_

Fig. 4.5 Interruptions of Operations — Year 1969

 

 
29

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i ; " Table k-7, INTERRUPTIONS OF (¥ERATION DURIRG 1966
What Happened ) Cause and Related Activities
| l%, 5 | 8= $lals 2
» 1. » m &
ol ol 22l 8] (58 31 EIEE]L) | Ele8 58y [adley 3] B 4], |5 .
Gl 813 LR T sl 5 g a5 a8 ael 5 | B0 g2]2e R 08 2] B ) 8| 3|2,
e we (ES| 2| XA 8 A |Bs| &) B 3A)5s) 58] 2 B8] 8k g5 du[as] 5 5 Bl B 3|5k
1 1/51fe0f M Pl _
2 | 1= A" : P
3 1/22 M| P
v | 123 At : _
5 1/23 M ) P-W
3 1/23 A*
7| 1/ak . A
'8 1/es il P _
9 lifes-a/2d m Miou : - ow| pw| w i w | ow
w0 | 216 M i M v | rw v
n | & | 7 X v
12 {.b422 M| ¥ ' P '
13 h/1h M P
1h h/16 a* 8
15 4f20 M M P .
16 L/20 A 8
17 | b/ M . P
18 Lfe2 A A A M P-W
1g Lfes A P-W
20 | w25 : AL 2 '8 P
4f28 Foll B P
22 bfo8 il I ) 8
23 | ufe9 ] x
2 kf2g ¥ ' W ¥
25 | s/m | ou BV
: . 2 | s5/12 A*l a s]| »
- ‘ J zr | 5516 K ' _ P
' 28 5/19 A A : : : P
29 | s/e3 A i 8
0 | s/25s | M P :
: 31 5/26 N P
32 5/26 M P P
33 5/26 M P
W | 5/ W N _
35 5/28 A ' v | pw| 8w v
% e | ou :
37 | 6 M W
8 1 6/19 X ' _ . P
39 6/26 ' A | A 8 1 P
w | 621 &0 o] afoa E Pew
w7/ il Y P
w2 | 7715 Ao : . _ P
| 1t A ' M} M i E L Hew] w | wi w
| 10/0 | ‘ -4l
ks | 1010 M . P _ )
s | 10/12 1 oaF - S : P
ar |06 || A oa b
K | 10/01 M P-W
k9 | 10/31 X M W
so | n/m o : W
51 | 11/12 M P
s2 |ups | oW P
.53 | AT M _ : P
5% | 11/20 |- M : W W | P-w
55 | 12/e3 A oa 2} i :
56 |o/n | a b e e
57 { 12/25 Al oo e . P
" *Reportable Unscheduled Rod Screm
A = Automatic
K = Manual
P = Primary
8 = feconlary - :
¥ = Worked on During Period Following the Intaermptio__n; Includes Preventive Maintenance

 
30

INTERRUPITONS OF OPERATICN DURING 1967

Table 4-8.

 

susmgoedg
az0)

P-W

 

9194], J0MNTV}
=Uo) poyLad

 

fasrdmey

P-¥

v

 

P0ATIA PZVAIT

 

spoy TOXU0)

 

UOT} W} TAHLYY §UT

P

P-¥

P-w

¥

P-¥

 

wn kg ITy
J08En.Ty euT|

P-w

 

I ey pus
TeOTReT

PW

 

© 7 afmng)
J9nod VA

 

RO WTTIUSA}
¥ JTeuRIpNU0D

 

wnefg JueToo)
usuodmo))

v

 

Cause and Related Activities

wnelg X039,

 

8w

 

iy
"eiz30

P-v

 

#ISNOTY 30
03 WPV

P-W

W

 

saneig 31vg
JUTOO) N TN

 

aozxy wewng]

 

"7 WO

 

*3dxg 03 enp
anq “peucwrduy)

 

PR

 

What Happened

QW JAETO0)

 

upmay Ty

 

wmIdg PuoT

 

X3 poy

 

og3enpaY
JM0g

 

 

Date

1/12

1/14

-

1/28
1/wr2f2

31
3/6
3/7

.

3/13

o
&
-

§

5/8

6/21

6/23

6/25
6/3
T/12
8/

g/1x

9/18

'9/18

1/1%
11/16

 

 

Xo.

 

123‘56789.”.“.2

~
~

A8

M1
16 | b/28-%/30
5/5

17
18
19
20
21
22
23

24

25| 8/8-8/11

2%
27
28

29| 109
30! /20
3| 10/23

32

33

 

12/13
Reportsble Unscheduled Rod Scram

A = Aitomatic
X = Mapual

P =

Priwary

W = Worked on During Period Following the Interruption; Includes Preventive Malntsnance

3 11 /o2
8 = Secondary

3B

 
31

 

 

 

suswgoedg »
8x0) ]
B, JUARITE} =
-00) potasd .
gaedmeg W M

 

SOATBA 9Z9axd

 

‘spoy tommen| >

 

WOT3 9 USIIG SUT

 

welg ItV
4usmmI}sul

 

sxayesl pue i
Te0Ta399TE - =

 

adeno
xem0d VAL

 

UOTHUTTIUSA
% FUSTUTBIUCY

 

mayehg JUBTOOD| _ :
.unm_noaaoo : ) x

 

 

umeds IaEy|

 

- Cause and Related Activities

wy 8hg -
883330 o - =

 

sasmoTg 10 ‘
_10795p%Y : ‘ =

 

sma18L5 3188
juBTo0) % Yend

 

INTERRUPTICNS OF OPERATION DURING 1968

JOJIXH flflgm

 

38T ¥0ou)

 

Table 4-9.

*3dxg 03 snp
anq ‘psuuerdun

 

 

pomusta] A A A B Aoma B RoRe R

 

uysaq 3ueTO0D|

 

 

 

 

 

 

s upsaq Tend : , xEE XX
ws1og peot| - ‘

= ‘ ol

v o

8 wexos POy o w = %
uopjonpay|

sonoq| X E R T E

 

 

*Report'a'bie' Unscheduled Rod Scram -

= Automatic. -

Primary

 

 

 

 

| 2 | 898 -NES8A5LIE8S
| 2| SRSSRSRSRRSIaaY
No.. H M MmS INO DO NG Ay MmN

 

W = Worked on During Period Following the Interruption; Includes Preventive Maintenance

M = mnual
8 = Becondary

A
P =

 

 

 
32

Table k-10. INTERRUPTIONS OF OPERATION DURING 1969

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

What Bappened Couse and Related Activities
g 2.0 g 5 Kl g 3 gf
- 1 » -
MARHAEHALH R RN T R
sd) 20l 0 T 28 5 | g e dh ae] s 0] s dal a2 2 2 | | ()
o) ~e (BE) 3| 3] 3 8|4 8e) 3| 035 350880 5 HLEEER sl kel k| A B BB a
1 113 | u W
2 /23 | w P
3 1/26 M P
A 1/28 | w P
S 2/7 “ P
6 2/ M P.
T 2/16 | M P
8 2/ M P-W
9 221 | 'm P
] 3n A oa P 5
1 3/8 N P v
12 3/20 x P
13 3/25 N P
s | 3/2s A*] s P 8
15 k3 u P
16 | &/10 Ml a|afa 8 v
17 2 | x A - P
18 | a1 A"l A a 8 P-¥ 8
19 5/% N U
20 5/5 N W
21 5/13 ¥ P
sf19 | x P
23 s/2s N P
b 5/26 N P
5 5/26 A"l P 8
s/er ¥ A P 8
6/1 » I'n | x |x r 8V W v v v |Pvw
8/0 | ) 1 e 8 '
29 8/z0 |. P 8
30 8/20 N P 8 .
31 8/20 X P s
2 8/20 N P 8
33 8/21 M P 8
3% 8/21 Y ? 8
35 | 6/28 A" P 8
6 | 8/29 A" P 8
BT | 8/29 | m P ’
B8 9T u P
B9 {  9/9 X P
o 9/16 N P
b1 9/18 u -V
h2 9/23 | m W
h3 9/30 A P-W
kb | 9/30 A P
hs |~ 9/3% A W
b | 9/3% A PV
h7 9/30 A P-W
b8 | lo/1 N P
ho | 10/3 M P
pO 10/15 M P
51 | 10/17 M P
b2 | 10/23 | a C 8 P
53 | 10/2h A 8 P
by | 11/2 ¥ | |xw x|
b5 | 12/3 n , P
b6 | 12712 |- w tn |nw {x|e v W
*Reportable Unacheduled Rod Scram
A = Automatic
M = Manual .
P = Primary
8 a Becondary Y
¥V = Worked on Durinz Pericd Following the Interruntion; Includes Preventive Naintenance
 

 

Teble 4.11

33

Description of Interruptions of Operation
" During 1966 | | o

 

No. - Date

" Description and Related Activities

 

1° 1/5-1/20

2 1/21

3 1/22
Y 1/23
5 1/23

6 1/23 | A

T 1/24
8 1/25

9 1/25-1/26

10.  2/16

-;Low—power tests to callbrate nuclear instruments,

check radiation levels, collect data for &ynamlcsl‘
and noise analysis, obtain reactivity balances,--
and check out the control circuits.

Rod scram due to ac01dental shortlng of the 32-¥ dc'
power supply while 1nvest1gat1ng a No. 3 safety
channel trlp

Rod and load scram to test ecircuitry.

Rod scram due to a voltage sag. - .

Pcwer-reduétion to attempt to blow out plugs in off-
gas line, equalizer 1ines,'and.main-charcoalfbeds.

Rod scram.due to 1nstrument malfunctlon when SW1tch1ng
linear channel ranges.

Load scram due to false low coolant flow indication.

Rod scram due to inadvertent actuation of the manual
scraméswitch

Fuel and coolant drain due- to plugging in the offgas:
system and to insulate the coolent flow transmitter

1ines.‘ While shut down, the restrictor in the equa-.

lizer line (521) and the check valve in the vent
line (533) to the auxiliary charcosal bed were re-
moved. The filter in the main offgas line was re-
placed and-the pressure control valve (PCV-522) was
replaced with a hand velve with larger ports. Work
was done.on No. 3 control rod drive, component
coolant pump No. 1, radiator door seals, and motor
generator No. 4. The thermal shleld air lock
= prdb1em.was investigated.

Fuel drain due to a fallure of the draln tank cell
‘space cooler motor, . While shut down, the inlet.

~ valves to the charcoal beds were ‘replaced and the
offges line at the fuel pump was reamed out. Re-
pairs were made on a damaged electrical penetration
on the’ sampler-enrlcher and a8 1eak on reactor cell

 

% - ,
Reporteble unscheduled rod scram.

 
 

 

 

34

 

 

- 20

Teble 4.11 (continued) . | - B | kifi

No. Date Description and Related Activities

10 (con't) space cooler No. 2_wds‘fixed. A static inverter was
installed to replace motor generator No. 4. The
treated water corrosion inhibitor was changed from

_ potassium to lithium nitrite-tetrsborate.
11 L/6 Fuel drain due to a plug at the auxiliary charcoal bed
‘ which turned out to be a poppet from a drain tank
vent check valve.
12 /12 Rod scram andiload_SCram to test performance of the rods
' and radiator doors.
13 L/14 Power reduction to collect data for noise analysis.
* )

14 4/16 Rod scram due to the fuel pump stopping which was
caused by a false low salt level 1nd1cat10n.

15 L/20 Manual rod scram and. load scram to end the 2 S=Mwr
operation.

¥ _ _ | o

16 L/20 Rod scram due to a false signal on safety channel No. 1 -
while safety channel No. 2 was being tested. ‘i;

17 Lh/21 Power reduction due to charcoal beds plugging.

* ' | o | ,

18 y/22 Rod scram, load scram, fuel drain and coolant drain
initiated by a faulty pump pressure relay which
tripped instrument power breskers. - Some salt froze
in radiator because the cooclant pump could not dbe
started due to a lubricating oil flow switch

| failure (FS-T54).

. _ - ,
19 h/25 Rod scram due to inverter failure.

. , _

L/25 Rod scram and load scram due to false low fuel-pump
level indication whlle doing periodic instrument
| check llsts.

. |

21  4/28 Rod scram and load scram due to TVA power outage
| caused by & storm.

# ' _ .

22 4/28 Rod scram and load scram due to spurious safety chan-
nel trips possibly caused by an electrical storm.

23 h/29_ ‘Manual rod scram and load scram to check effect of

cold doors on radiator temperatures. Shutdown was
necessary to repair the latch on the sampler- o
enricher. - b’ '

 

®
Reporteble unscheduled rod scram.
 

35

Table 4.11 (continued)

 

 

132

3

No. Date Description and Related Activities

2k L/29 Fuel drain necessitated by a failure of the samfiler-
enricher cable drive motor. While shut down, the
belts were tightened on component cooling pump No. 1
and attempts were made to blow out the plugs in the
offgas lines.

25 5/11 Power reduction due to offgas line filter plugging.

% ' >

26 5/12 Rod scram and load scram due to operator failing to in-
sert a control circuit jumper while doing perlodlcal
instrument check llsts.

27 5/16 Power reduetlon to increase pitch on main blower blades
to increase possible maximum power.

# .

28 5/19 Rod seram and load scram due to a TVA power outage.
While shut down, measured offgas pressure drop and
attempted to blow out plugs

¥

29 5/23 Rod scram caused by instrument malfunctlon end operator
error. Operator had switched to servo with a LO°F
difference between actual temperature and temperature
demand setp01nt. '

30 - 5/25 Power reduction to check the power coefficient of Te—
act1v1ty. o :

31 5/26 Power reduction to obtain reactivity balance data.

5/26 Power reductlon due to main blower No. 1 stopplng and
falllng to start.

33 . 5/26 | Power"reduction to collect data'for_anoise analysis.

F * . el T . .

34 s5/26 Manual rod screm &nd load scram to investigate a

R false flre alarm.

5/28 Fuel draln due to & failure of component c¢oolant pump

- No. 1 while component coolant pump No. 2 wes valved
off. - A planned drain had been scheduled due to the

~ high indicated cell leek rate. This was found to be

due to,arleak from the thermocouple headers into the
cell. :While shut down, a 20-psig leak test was made

 

¥_ — RCITE ,
Reporteble unscheduled rod scram.

 
36

Table 4.11 (continued)

 

 

No. Date Description'and Related Activities

35 (con't) on the cells, repairs were made on both 6omponen£
coolant pumps, the electrical load was redistributed
and attempts were made to locate the water leak
into the cells.

36  6/1h Power reduction due to & failure in the electrical
supply to the nuclear instruments.

37 6/1k Power reduction due to failure of a flexible coupling

: on main blower No. 1.

38 6/19 Power reduction to empty the overflow tank.

€ - - :

39  6/20 Rod scram and load scram due to the operator inadver-
tently inserting a test source at the wrong process
radiation detector (RE-528 instead of RE-565) while

‘ doing periodic instrument check lists.
* - ' ‘

40 6/27 Rod scram, load scram, fuel drain, and coolant drain
due to the electrical supply to component coolant
pump No. 1 shorting out at the penetration.

. | | ,

L1 T/14 Rod scram and load scram due to a TVA power outage.

¥

42 T/15 Rod scram and load scram due to a TVA power outage.
caused by a storm.

43 T/17 Power reduction and coolant drain due to & catastrophic

failure of main blower No, 1 hub. The fuel salt was
drained on T/24/66. While filling the fuel system -

- with flush salt, the fuel-pump bowl was accidentally
overfilled and some salt froze in the sampler and
offgas lines. Heat had to be gpplied remotely to
remove the plugs. While shut down, the core speci-
mens were replaced, and a new particle trap was in-
stalled in the offgas line. Lesks in reactor cell
space cooler No. 1 and in the treated water heat ex-
‘changer were repaired and the thermal shield degas~-
sing tank was installed. Both component coolant
pumps were repaired and piping was installed to
drain the condensed wsater from them. Repairs were
made on the radiator door seals, radiator heaters,
and the main blower motors. All the electrical
switch gear breakers were calibrated. Prior to
startup, the fuel and coolant pumps lube oil was
changed, the periodic check lists were completed and
the cells were leak-tested at 10 psig.

 

¥
Reportable unscheduled rod scranm,
 

37

Table 4.11 (continued)

 

No.

Date

 Description and Related Activities

 

bl
45

#

Lo

h7¥
48

ko

50

5

22

53

54

%
55

10/10

10/10

10/12

10/16

10/31 -

10/31

11/11

11/12

11/15

11/17

11/20

12/23.

.- Power reduction due. to a false signal from a process

radiation detector (RE-528).

Power reduction due to an experiment when the fuel-pump

pressure was rapidly vented from 15 to 3 psig.

Rod scram due to inverter trouble.

‘Rod scram and load scram due tb'false signal from a

process radiation detector (RE-528).

Power reduction to install & new rotor on main

"blower No. 1.

'Fuel drain due to upper offgas line (524) being

plugged. While shut down, the flow restrictor in
this line was replaced and heat was applied to main .
. offgas line at the fuel pump to unplug it.
Power reduction to check main blower vibration.
Power‘fEduction‘to empty the overflow tank.

Power reduction to empty the overflow tank.

Power reduction to attempt to blow the plug out of the
main offgas line at the fuel pump. )

Fuel drain due to a high indicated cell lesk rate which

proved to be due to lesking air line disconhects on
.. in-cell valve operators. Rotemeters were installed
on these air lines. While shut down, the plug in
the offgas line at the fuel pump was reamed out and
' repasirs were made on the component coolant pumps and
“the component coolant system pressure control valve
(PACV-960). Hesters were instelled on the inlets of
- main charcoal beds 1A and B. The cells were leak-
tested at 10 psig. = -

Rod scram and load scram by accidentally opening a cir-

- cuit while installing jumpers for an experiment.

 

% o
Reporteble unscheduled rod scram.

 
 

38

Teble 4.11 (continued)

 

No. Date  Description and Related Activities

 

56 12/2h - Power reduction caused by a pressure release experi- -
: - ment. The reactivity decreased, the regulating
rod withdrew to its upper limit and the operator

failed to change the limit.

57 12/26 Power reduction when venting the fuel pump after an
attempt to blow out the plug in the mein offgas
line. The regulating rod reached its upper limit
and the operator failed to change the limit.

 

& , .
Reportable unscheduled rod scram.
 

39

Table h 12 Descrlptlon of Interruptions of Operatlon Lo

Durlng 1967

 

No.

Date

Description and Belated.Acfi#ities_

 

10

1/12

1/1k

1/16

1/28

1/30f2/2j

2/5

o
2/26-2/2T
3
ii'3/6

3/7

Rod scram end load scram due to a freeze valve (FvV-108)
relay tripping the reactor out of RUN during an in-
vestigation of the plastic keepers on the relays.

Power reduction to cbserve xenon decey rate and ob-
tain zero-power reactivity balance data. -

" Fuel drain due to leskage from the in-cell air line

disconnects. The measured leakage was so high that
_ reasongble cell lesk rate calculations were not pos-
‘sible. The disconnects were replaced. While shut
"down, the main offgas fllter and: control valve
(PCV-522) were removed and a new filter was installed.
~ The treated water heat exchanger was replaced and
. maintenance was done on the component coolant pumps.

~ The cells were,leakftested at 10 psig.

Power”feducbiofi!to.feet automatic load control.

','Power reductlon to take plctures of the radiator and

1nvestigate heat transfer of the radiator and heat
exchanger. “Also made dynamic tests and checked the
temperature servo system. :

: Power'reductlon due to the'failure of a freeze valve

module (FV-107) which trlpped the reactor out of
RUN mod.e. B

: Power reduction due to main blower No. 1 vlbratlon.

cPower reductlon to-dbtaln,heat—transfer data.;

Power reduction due to coolant system offgas filter
plugglns -

“l Pcwer reductlon due to electrlcal power supply trouble

to nuclear safety and rod servo. .

'lPower reductlon due to 8 bad bearing on main blower

- No. 3. Remained at low pover to observe effect of
xenon decay on resctivity balance

 

* (
Reportable unscheduled rod scram.

 
 

 

 

 

Tsble 4.12 (con't)

40

 

No.

Date

Descriptibh and Related Activities

 

*
12

13

1k

15

16

17

18

19

20

21

22

3/13

3/19

3/23

411

“h/28-4/30

5/5

5/8

6/21
6/23

6/25 -

6/30

Rod scram and load scram due to false signal on safety

channel 3 while doing perlodlc 1nstrument check lists.

Rod reverse and load scram due to false low fuel-pump
speed indication while doing periodic instrument
check lists. '

Power reduction due to vibration on main blower No. 3.

Power reduction to obtain a special fuel-pump gas
sample. Also obtained heat-transfer data.

Power reduction to cdllect date for noise analysis.

Power reduction to test reactor response to load |
changes W1thout rod changes.

Rod scram, load scram, fuel drain and coolant drain to
remove reactor core specimens. While shut down,

. gamma-scan data were taken, a new source was in-
stalled, and heaters were installed on the inlets to
main charcoal beds 2A and 2B. Reactor cell space
cooler No. 2 was replaced and repairs were made on
control rod No. 2, radiator door brakes, component
coolant pumps and the main blowers. A boot was re-
placed on the sampler-enricher. Prior to startup,
the periodic check lists were completed, the cells
wvere leak-tested at 20 psig.

Power reduction to'obtain heat-transfer data.

Power reduction to repair an oil pressure switch on
component coolant pump No. 1. A

Rod scram and load scram due to a TVA power outage
caused by a storm. While shut down, repairs were
made on component coolant pump No. 1 oil pressure
switch, a treated water rupture disc (84h) was re-
placed, and a heavier base was installed on main .
blower No. 3.

‘Rod scram and load scram due to an operator resetting

the wrong channel while doing periodic instrument
check lists.

 

¥ |
Reporteble unscheduled rod scram.
 

 

Table L4.12 (con't)

41

 

 

No. Date Description and Related Activities
o

23 T/12 Rod scram and load scram due to a TVA pover outage
) | caused. by a storm.

oh  8/T Power reduction to obtain heat-transfer data.

25 8/8-9/11 Power reduction, fuel drain and coolant drain dué to
the sampler-enricher drive cable being severed.
During shutdown, the latch was retrieved but the
capsule remained in the fuel pump. Prior to startup,

| ' maintenance was done on compohent coolant_pump No. 1.

26 9/11 Coolant drain due to the radlator door blndlng in the

' guides.

27 9/18 - Power reductlon due to an 011 lesgk in component coolant
pump No. 2. |

28 9/18 Fuel drain due to a leak through the discharge valve of
component coolent pump No. 2 which prevented repair
of the oil leak.

29 10/1 Power reduction to collect date for noise analysis.

% o ’ -

30 10/20 Rod scram and load scram due to a power failure due to
‘an arc between a switch actuator rod and the main
power llne to the building.

31 :10/23  Power reduction to observe the effect of xenon decay

: on the reactivity balance. .

32 11/1k Power reduction to collect data for noise analysis.

33 kéll/l6 'Pofier reductlon to repalr sampler—enrlcher wiring.

3 11/22- Rod scram.whlle malntenance was being performed on the

B rod servo instrumentatlnn.-

- 35 ~f 12[13 Power reductlon to Observe the - efféct of xenon'p01son—

ing on the reactivity balance and to collect data
for noise analy31s._

 

# L Lo .
Reportable unscheduled rod scram.

 
 

 

42

Table 4.13 Description of Interruptions of Operation

During 1968

 

Description and Related Activities

 

No. Date
1l 1/22

2 2/21
3 2/29

Y 3/7

5  3/22
6  3/25

T  3/25

8  3/26

9  9/1k

Power reduction to observe the effect of xenon poison-
ing on the reactivity balance and to collect data
for noise- analysis.

Power reduction to test the rod servo instrumentation
"under simulated 233y conditions.

Power reduction to observe the effect of xenon poison-
ing on the reactivity balance and to collect data
for noise ansalysis.

Power reduction to observe the effect of xenon poison-
ing on the reactivity balance, to make dynamics tests

and take heat-transfer data.

Power reduction to meke dynamics tests.

Manual rod scram end load scram to observe xenon
poisoning effect on reactivity balance.:

Manusal rod scram to end_zasU operation.'

Fuel drain to end 23°U operation. Shutdown was emmi-
nent due to & tangled drive cable in the sampler-
enricher. After the drain, the capsule was acci-
dentally dropped into the fuel pump when trying to
untangle the cable. Unsuccessful attempts were
made to recover this. While shut down, gammas-scan
data was teken, the core specimens were removed, and
the main offgas line . gt the fuel pump was unplugged.

- The flush and fuel salts were transferred to the
‘£Uel processing plant where the 235U was removed.

33 was added to the fuel salt. Repairs were made
on two heat exchanger heaters, control rod No. 2,
both component coolant pumps, the inverter, and the
radiator doors. Meain blowers No. 1 and No. 3 bear-
ings were replaced and the coolant offges filter
wes replaced. Prior to startup, the periodic check
lists were completed and the cells were leek~tested"
at 20 psig.

Fuel drain for adding more 233U as part of the criti-
cality experiment.

 
 

 

 

Table 4.13 (con't)

43

 

Déécription and Related Activities

 

No. Date
10 9/17
11 9/21
12 10/21
, ¥ '
13 11/27
1k 11/28
15  12/17

Fuel drain for adding more 233y as part -0f - the criti-
cality experiment. :

Fuel drain for adding more 233y as part of the criti-
cality experlment.

Fuel drain due to inverter fallure. Operatof’falled to
reset reactor cell radietion monitor in time to pre-
vent drain.

Rod screm and load scram due to operator failing to in-
sert a Jumper around fuel pump level contacts. The
fuel-pump pressure was being vented after attempt-
ing to blow out a plug in the main offgas line.

Fuel drain and coolant drain to mix the fuel salt in
the loop and the drain tank. The offgas line at the
fuel pump was reamed out, the coolant offgas piping
was cleaned, and repairs were made on the component
coolant pumps.

Fuel drain due to difficulty with the latch and cable
drive of the sampler-enricher. While shut down,
control rod drive No. 3 was inspected and a weight
added to decrease its drop time.

 

 

. | | R
Reportable unscheduled rod scram,

 
 

- 44

 

 

Teble L.14 Description of Interruptions of Operations
| During 1969 '
No. Date Deseription and Related Activities
1 1/15 Power reduction to repair lower limit switches on
radiator doors.
2 1/23 Power reduction to investigate'effect of power on
"blips".

_3 -1[26 Power reduction to infiestigate effect of stopping the
fuel pump on "blips".

k1728 Power reduction to investigate "blips".

5 /7 Power reduction to collect data for noise analysis.
6 2/11 Power reduction to run tests on natursal ponvection;
T 2/16 Power reduction to run dynamics tests.
8 2/23 Power reduction due to failure of belts on stack
f&n NO. lo ’ .

9 2/27 Power reduction to connect the fuel pump to a variable
, speed motor generator set.

* ! -

10 3/4 Manual rod scram and load scram due to failure of the
veriable speed motor generator set.

11 3/8 Power reduction to -allow xenon to decay in preparation
to making void fraction tests. While shut down, the
coolant offgas filter was replaced. '

12 3/20 Power reduction to connect the fuel pump to the normal
electrical supply.

13 3/25 Power reduction to connect the fuel pump to the varisble
frequency motor generator set. .

1 3/25 Manual rod scram and load scram due to failure of the
variable speed motor generator set.

15 L4/3 | Power reduction to connect the fuel pump to the normal

electrical supply.

 

# _
- Reporteble unscheduled rod scram.
 

 

 

Teble L4-14 (continued)

45

 

 

6/1

 

- No. Date Description and Related Activities.
¥ ’ . ’ . ’ L )

16 L/10 Rod scram, load scram, fuel drain, and coolant drain

due to a burned-out fuse on one phase of the main
~transformer. Drains could have been averted by
operating alternate equipment.

17 L/12 Load scram due to a disturbance in the 48-v dec power
supply. = '

18 /15 Rod scram, load scram, and fuel drain due to the

' drain freeze valve thawing. The setpoint on a tem-
perature switch had drifted off the setpoint and an

| inadequate plug had been established. ,

19 5/4 Power reductlon due to a failure of the coolant stack
berylllum monitor.

20 5/5 Power reduction due to an error in calculating the
reactiv1ty balance :

21 5/13 Power reductlon to connect the fuel pump to the
variaeble speed motor generator set.

22 5/19 Power reduction to connect the fuel pump to the normal

' electrical supply.

23 5/25 .Power reduction due to plug in overflow tank vent
line vhlch caused dlfflculty in emptylng the overflcw
tank. '

2k 5/26 Power reduction to connect the fuel pump to. the varlable
speed motor generator set.

o

25 - 5/26 Manual rod.scram and load scram due to. fallure of the

: varlable speed motor generator set.
26 5/2T Manual rod scram snd load scram due to fallure of the
o variable speed motor generator set.
27 - . Manual rod scram, load scram, fuel drain, and coolant

~ drain to remove core spec1mens and due to plugging
in the offgas system. Control Rod No. 3 did not
scram. It was replaced during the shutdown. Ex-

: ten51ve gamma-scan date was tasken. A heater was in-
stelled on the mein offgas line near the fuel pump
which aided in removing the plug. The overflow tank

 

*
Reportable unscheduled rod scram.
 

46

Table L4.14 (con't)

 

No. - Date | Description and Related Activities:

 

2T {con't) offgas line was replaced and the coolant offgas sys-

o tem was unplugged. Repairs were made on the sampler-
enricher, and on a leaky in-cell air line disconnect.
Before startup, the periodic check lists (modified)
were completed and the cells were leak-tested at
20 psig. Full power was delayed while tests were
made at various fuel-pump speeds to 1nvest1gate
bubbles 1n the loop.

28 8/20_.  Power reduction due to failure of the variable speed
motor genersator set.

29 8/20 Power reduction due to failure of the varisble speed
motor generator set.

30  8/20 Power reduction due to failure of the variable speed
motor generator set.

31 8/20 Power reduction due to failure of the variable speed
motor generator set.

32 8/20 Pover reduction due to feilure of the variable speed
motor generator set. -

33 8/21 Power reduction due to failure of the variable speed
motor generator set.

34 8/21 Power reduction due to failiure of the variable speed
: motor generator set.

35 8/28 Manual rod scram and load scram due to failure of the

variable speed motor generator set. >
* . _ : '

36  8/29 - Manual rod scram and load scram due to failure of the
variable speed motor generator set.

37 8/29 Power reduction for xénon decay in preparatibn for
argon cover-gas experiments.

38  9/1T Power reduction due to an error in calculating the
reactivity balance.

39 9/9 Power reduction to collect data for noise analysis, to
meke gamma scans and run rod drop tests.

Lo 9/16  Power reduction to connect the fuel pump to the normal

electrical supply.

 

* .
Reportable unscheduled rod scram.

l}r\
 

 

 

Table 4.1k (con't)

47

 

 

No. Date  Description and Related Activities
41 9/18 Power reductlon to replace 8 bearlng on Main Blower
i No. 1.
b2 -~ 9/23 Power reduction to replece a bearing on Main Blower
No. 1. While shut down, connected the fuel pump to
the variable speed motor generator set.
43 9/30 Load scram due to dirty contacts on relays.
L)y 9/30' | Load scram due to dirty contacts on relsays.
L5 9/30 'fLoad scram due to dirty contacts on relays.
h6 9/30 Load scram due to dirty contacts on relays.
L7 9/30 Loed scram due to dirty contacts on relays.
L8 10/1 Power reduction’fo connect tfie fuel pump to the normal
- electrical supply. b -
.h9 10/3 Power ?eduction due to door being insdvertently left
' ’ " open in an exclusion area which caused inadequate
containment ventilation. -
50 . 10/15 Power redfiction to take 8 special fuel-pump gas sample.
51 10/17 Power.reductien to take special ffiel—pump offgas sam-
: ples and to run tests with the fuel pump off.
52 10/23 Load scram due to a false trip of a coolant salt flow
' B reley while doing periodic instrument check lists.
53 10/2k | Laad'Seram'while7testing'fe1ays after they were re-
paired. 'Operator failed to reset'the relays.
5k 1/2. Manual rod ‘scram, load scram, fuel draln and coolant
: - draln to ‘teke gamma-scan data,
55  12/3 Power reductlon to take gamme-gcan data.
56 12/12 ' Manual rod scram, load. scram, fuel drain, and coolant

drain. tQ end operation of the reactor.

 

 
 

 

48

The next three tables summarize the foregoing mass of 1nformatlon on
interruptions, Table 4-15 is a summary by type; Tsble h—16 by cause and
work done during the interruption. Because unscheduled scrams of the con-
trol rods are of special interest, a separate breskdown of these is given
in Teble 4-17. It should be noted that never were the rods seremmed because
7 of the reactor power, the #eactqr peribd, or the fuel temperature going out

of limits.
4,4 Time Required for Operstional Tasks

, One aSbect of the opefdbility of a plant (and ohe which affects plant

. availdbilify) is how long it takes to do various tasks required in the opere-
tion. The purpose of this section is to summarize the MSRE experience in
this regard. '

Manpower and the approkimate number of working hours reqpired.for
verious tasks are listed in Teble 4-18. One must recognize that some of
these figures are subjeet to considerable vaE}ation. The time required to
do a particular task'depended on chance difficulties that were encountered,
the experience of the crew, other jJobs being done concurrently and the
urgency of speedily completlng the task, i.e., whether or not it was given
a hlgh frlorlty. Furthermore, sometimes extra requirements such as special
data-teking caused the Job to take longer than usual. The figures in Tsable
4-18 apply to fairly normel situstions; not the best nor the worst. Very

brief descriptions of the tasks.are given below, with references to the

‘section of the MSRE Operating Procedures2? followed in each ease,rif applicable,

L.4.1 Auxiliary Systems Startup Check Lists

, During a shutdown, much of the equipment remained in operstion. However,
after long shutdowns, the auxiliary systems startup check lists (Section 4)
were completed to assure that nothing had been overlooked. All valves and
switches were checked, eqfiipment was put into operation and standby equipment'
was tested. These were done as late in the shutdown as practical.

LA — Electrical Startup Check List assured that all main breskers were
closed so that equlpment and heaters could be started from the control
boards. '
 

 

 

49

Teble 4-15 = Summary of Interruptions by Type

 

Number of Interruptions

 

 

Type of_Iptérifiptiohf:_ - . 1966 1967 1968 1969  Total
Power Reduction: Automatic 3 -0 0 0 -3
| Manual 23 23 D 39 90
Rod Scram:  Automatic 20 T 1 T 35
| : Manual . 6 1 2 3 12
Load Scram: ~ Automatic 13 8 1 1h 36
| Manual 6 1 1 3 11
Fuel Drain Automatic 3 1 2 6
| Manual T L 6 3 20
Coolant Drain Automatic 1 0 0 1 2
- Manual 3 3 1 3 10

Total - 85 Yy o 18 75 225

 

 
 

 

50

. |
Table 4-16. Summary of Interruptions by Cause and
Work Done During Interruption

 

 

 

 

 

 

 

 

 

 

 

- Number of Interruptions -
1066 - | __1067 ]_1968 |__ 1969 . Total
Category P S W|P s wlp S WI|P S WI|P 8 W
Planned 8 o o1k 0 0fl2 o 02k 0 0|58 O O.

Unplanned |
But Due to an . :

Experiment 3 00 O 0 0OjJO0O O0.0J13 0 0116 0 ©
Check Lists o 50| 120|000}010}|1 8 ,_
Human Error T 1 0 21 011 0fj2 3 011 6 3
Coolant Salt | o | .

Systems '3 00| 000]0O0OO0O]|]0OO11f3 01
Radiator or , _ fl

Blowers 6 1 5 L o 5|0 0 12 0 3|12 113
Offgas Bsytem {10 o012 0 0 0 2 1 2 |12 118
Water System 1 o 4! 01 3[000 ofl 1 7
Component Coolant ? | §

System 2 0 6 2 0 210 0 210 0 O{ 4 010
Containment eand =~ -~ = - i ' | _

Ventilation 1l 1 2 O 0 0/0 0 O 1 0 12 1 3

|TVA Power Outage 3 1 0 2 0 0|0 0 0 0 0 0|5 1 O
Electrical and | .

Heaters 4 o-4:; 1 0 1{0 01 ;213 1713 T
Instrument Air ! § ' g

System 1 01{ 101|000;001[2 03
Instrumentation 11 o0 3| 70 7|1 o0 o01{11 0 930 019
Control Rods O 02, 002/002;001|00T
Freeze Valves : O 00: 00 0jO O O}O0 10O 10
Samplers 2 03 2. 03{112}001(5 19
Periodic Con- * -
| tainment Tests 0 03] 002]0oo011i00 112|007

|Core specimens | 0 o0 1| 1 0 ©01j10 2 0 4

 

 

 

 

NOTE: P = Primary cause of interruption.
S - = Secondary cause of interruption.
W = Worked on during period following the interruption (includes

preventive maintenance).

% _
No interruption of operations during this entire period was caused
by the cover-gas system, lube-oll system, or leak-detector system.
 

 

 

Table 4.17 MBRE Cumulative Cycle History

_ Thaw Usage
' . ‘ ' Quench & Factor
" Component. Heat/Cool|Fill/Drain|Pover |Quench | Time | On/Off| Thaw| Trans.: &

 

*These figures are based on the original calculations, If they were based on freeze flange thermal

cycle tests, the usage factors would be 23.04% and 13.37%.

- TI6

 
 

 

Table L4-18,

APPROXIMATE TIMES REQUIRED TO PERFORM VARIOUS OPERATIONS

 

 

 

 

‘ _MANPOWER
Operating ' E= Engineer
Procedures Time to T= Technician

Section Complete P= Pipefitter

Number Operation (hrs) I= Instrument Mechanic
LA Electrical startup check 1list 10 - 15 1-T*‘-
4B Instrument air system startup check list L -8 1-7"
ke Water system startup check list 4y -8 -7
4D Component cooling system startup check list 1 -2 l-T*

Ll ~ Shield and Containment startup check lists . N o
(1) Ieak test of containment valves ‘130 - 160  1-E, 1-7, 1-2*
(2 Pressurize the cells (to 20 psig) 15 - 18 e
(3 .Cell leak rate at pressure >60 - e
(4 Evacuate cells and purge with nitrogen 18 - 24 e
(5) First cell leak rate at -2 psig S170 W
LF Ventilation system startup check list L - 8 1-1
hg Leak detector system startup check list 2 -4 1-7"
LH ‘Instrumentation startup check list. 160 - 200 1-E, 1-T, 1-I
SA Purging oxygen and moisture from the fuel circulation system 30 - ko L
5B Startup of cover-gas and offgas systems - ‘ 2 -k l-T*
. *
5E Startup of lube oll systems 2-4 1-T
5F Heatup of fuel and coolant systems 4o - 50 SR
S5H Routine pressure test 20 - 30 NN
*
. Done on shift,
*fMost of this was done on day shift with some assistance from the shifts,
***This d1d not require much attention.
This required the attention of most of the shift operating crev,

cs

 
 

 

 Table L4-18, APPROXIMATE TIMES REQUIRED TO PERFORM VARIOUS OPERATIONS

 

 

 

 

(continued)

- - MANPOWER
Operating o o : : E= Engineer
Procedures . | Time to T= Technician

Section ‘ ' Complete P= Pipefitter
~ Number ‘ Operation ' (hrs) I= Instrument Mechanic
51 Filling the fuel and coolant systemé

© (1) Filling the coolant system ' - 9-12 XK

(2 Filling the fuel system with flush salt 9-12 FHHX

(3) Drain of flush salt and secure - 3-5 *hRx

() Prepare for fuel f£111 - . 8-9 RN
- (5) 'Filling the fuel system with fuel salt | | 10 - 12 RN
5 | Criticality and pOWer operation o |

‘f(l)V | Preparation for. ‘eriticality and power operation : 1-2 Ty

(2)  Suberitical to full power. , | 1 1-7
6A3 or I | o j . . ‘

‘ 6Ah Fuel salt sampling or enriching 3-4 2-7
6B Coolant salt sampling 1 1-T
8 . Periodic instrument checks

8A '~ Neutron level instruments 5 -7 2-1

8B Process -radiation monitors _ 1 3-7

- 8D Safety eircuit checks 2 =L 2 to 3-T
o Emptying the overflow tank 1/2 -1 e
124 Logs |

- 12A-2A  First control room log on each shift 3/4 - 1-1/4 'l-T*

*

Done on shift.

Most of this was done on day shift with some assistance from the shifts,
This did not require much attention, :

This required the attention of most of the shift operating crew,

¥

i

 

£q

 
 

 

Table 4-18, APPROXIMATE TIMES REQUIRED TO PERFORM VARIOUS OPERATIONS

 

 

 

 

(continued)
MANPOWER
Operating . B= Engineer
Procedures Time to T= Technician
Section ' Complete P= Pipefitter
Number Operation (hrs) I= Instrument Mechanic
12A Logs (continued) '
12A-2A Second control room log on each shift 1/6 l-T:
12A-2B First bullding log each day 1-1/2 - 2 1-T,
12A-2B  First bullding log on other shifts | 1/2 - 1 1-T,
12A-2B  Second building log on each shift 1/3 - 1/2 1-T
10 'Shutdown of the Reactor |
(1 Full pover to subcritical 0-1 1-T,
2 Fuel salt drain and secure _ 3~-5 1-T*
3 Coolant salt drain and secure 2-14 1-7
L - Cool down to LOO°F L8 - 60 6k
OA-1 Restarting equipment after an electrical power outage 1/6 N
, *
5J-1 Return to power after a load and rod scram : 1 1-T
114, sI, | |
& 5J Return to powver after a drain
(1) Drain to both drain tanks (transfer through transfer lines) 50 - 60 WHHH
(2 Drain to both drain tanks (transfer through £ill lines) 20 - 30 R
(3 Drain to one drain tank _ _ _ 10 - 15 RN
.
Done on shift. ,
**Host of this was done on day shift with some assistance from the shifts.
**¥IMis d1d not require much attention,
O s required the attention of the shift operating crew,

1S
 

 

 

35

4B — Instrument Air Systeni Startup Check List assured that the com-
pressors and driers were opereting properly, that the standby compressor
would automaticallj start if needed, and that emergency nitrogen cylinders
were installed. o .

LUC — Water System Startup Check List assured that all pumps and tower
fans were operable and that proper water flow was established on all
equipment. 7 | - ‘ )

4D — Component Cooling Systems StartupdCheck List assured that all
vaiving was set properly and that all blowers were operable.

UYE — Shield snd Containment Startup Check List is divided into
5 parts. Part (1) checked that the primary and secondary block and check
valves, containment enclosures, and llnes were leak-tight to assure ade-
quate containment. Removable' shieldlng was also checked to assure that it
‘ had been reinstalled. This check list was assigned to an operatlng engi-
neer on day shift with technic1an and plpefltter help as needed.
 Part (2) — this is the time required to pressurlze the cells to 20 psig
by adding instrument air. Part (3) involved taking pressure and tempera-
ture data periodically to establish the cell leak rate. Some containment
valves and lines were checked while atrrressure. The time required for
checking these is included in Part (1). Part (L) is the time required to
vent the cells, evacuate tcf—2.psig, and purge with nitrogen. Part (5)
alsc,involved taking periodic data. Since there was a small wster leak in
the cell,'it took about T days for the atmosphere to reach equilibrium

hF —-Ventilation System Startup Check List assured that a stack fan
was in operation, that the other one would automatlcally start and that
ventllatlon was adequate in all aress.

o hG —-Leak Detector. Systems Startup Check List assured that all valves
were open to leak-detected flanges.and that the overall leak-rate was
satisfactory. e | i

' hH —-Instrumentatlon Startup Check List assured that each instrument
'and circuit functioned properly Abnormal conditions.were similated and
where possible the control acticn'was'ellowed to occur. One operating
engineer on day shift wasraSSigned respohsibility-for completing this
- check list with technicien and instrument mechanic assistance as required.

Some of the tests were done on shift.

 
 

56

4.,4.2 Reactor Startup

 

The manipulations involved in actual startup of the reactor
(Section 5) are described below. ‘ |
| SA — Purging Oxygen and Moisture from the Fuel Circulating System in-
volved setting the helium purge at maximum rate and opera&ing the fuel
pump to circulate it. The time of purge was a calculatedvnumber.

5B — Startup of Cover Gas and Offgas System involved setting the
valvgs'and checking the helium treatipg station. o
_ ’5E_—-Start of Lube.dil Sysfems involved checking the supply tank,
draining the o0il catch tank, éetting valfes and checking that the standby
pumps would start sutomatically. | |

' SF — Heatup of Fuel and Coolent Systems involved raising the heater

settingé end maintaining a reésonéble temperature distribution. The fuel
- and coolant systems could be heated simfilfaneously. The time given is
for heating the coolant system from room temperature and the-reactor sys-
tem from 400°F (normal shutdown condition) to 1200°F ahd reaching
equilibrium.

5H — Routine Pressure Test involved raising the fuel system pressure
to 60 psig (usually with flush salt circulating), then lowering the fuel
system pressure an@ increésing'the coolant system pressure to 60 psig.
All pressure switches and their control or alarm actions were checked.

5I — Filling the Fuel and Coolant Systems is divided into 4 parts.
Part (1) involved thawing the freeze valves, filling the coolant loop,
.checking thaw times of the freeze valves, refilling and starting the
coolant pump. Part (2) involved thawing the freeze valve, filling the
fuel loop with flush salt and starting the fuel pump. - Part (3) involved
-dreining the flush salt, emptying the overflow tank and freezing the
freeze valve. Part'(4) involved thawing the freeze valve, adjusting tem-
peratures, running rod drop times and various other safety checks and
~taking base line countrate data. Part (5) involved filling the fuel
loop with fuel salt, stopping at six levels to teke count rate deta,
freezing the drain freeze valve and starting the fuel pump. -

53 —-Criticality‘and Power Operation is divided into 2 parts.
Part (1) involved running rod drop and other tests. The time does not
 

- maintaining containment. qu_samples could be delivered to the laboratory

57

include check lists 84, 8B, or 8D which sometimes had to be done before
criticality (see 4.4.3). Part (2) involved menipulating the rqu‘and

heat-removal system componentsrto reach full power.

L. 4,3 Operation

Various Jobs were done while the reactor was operating. These are
described below. _
6A3 or 6Alk — Fuel System Sampling or Addition of Enriching Capsules

involved a series of;manipulations to insert and withdraw a capsule while

" per shift if no difficulties were encountered. This rate could not be
sustained due to decontamination bf carriers, etc. |

| 6B — Coolant Salt Sampling involved a less complicated series of

- manipulations to insert and withdraw a sample from the coolant pump.

8 — Periodic Instrument_Checks involved testing as much of the

 

eritical equipment as possible without interrupting operation of the
reactor. | ' : |
91 — Emptying the Overflcw Tank involved pressurizing it with helium
and venting when it was nearly empty. This took about twice as long when
the FP offgas line was plugged(
12A — Control Room and'Building Logs were taken tfiiéé per shift.
Some adjustments were usually required, water treatment was added and

other odd Jobs were done as part of the logs.
k.%.4 Shutdown of the Reactor

This (Sectlon 10) is descrlbed below. - It is divided into L parts.
Part (1) involved reduc1ng the power. This could be done by scramming
| the rods and the load or by 1nsert1ng the rods and lowering'the radiator

 doors. Part (2) involved-draining the fuel system and freezing the freeze

"f,valves. Part (3)'ifivblved drsining the coolant system and'ffeezing the

 

 

freeze vaIVes; Part (L) 1nvolved coollng the systems to hOO°F
" '4,4.5 Recovery from Unplanned Shutdowns

A numberrof times durlng operatlon, power outages and other unex-
pected events caused shutdowns. Often there was a desire to return to

operation as soon as possiblé. Some of these are described below.
ORNL-DWG 73596

 

  

DAYS DAYS
1-30,~26 -20,~15 -10,-5 | 1 | 2 , 3 , 4 .8 , 8 , 7 4 B ; 9 , 0 ; M, 12 , B , 4 4 15
| i 1 | [ I | o i } I I I i
Repairs, modifications, preventative maintenance
and core specimen replacement.
Instrumentation startup check list. E

Containment startup check list,

Other systems startup check lists.
Last closure of primary system,
Purge of primary system,

Wald membrane and instal] blocks on cells.
Pressurize cefls to 20 paig.

Leak—test cells at 20 psig.

Elvacum cells lnd pruge with nitrogen.
Heat up fuel and coolant loops.
Moisture reaching equitibrium in cells,
First cell lsak rate at vacuum.

Fill coolant system.

Filf fuel systém with flush salt.

Sample coolant salt,

Sample flush nit.

Drain flush salt.

Prapare for fuel salt fill.
VFIII fuel system with fuel salt.

Sampie fuel ult.l

Preparation for criticality afid powor operation.

Criticality and taking the reacter to full power.

Fig.

 

4.6

 

 

 

 

Return to Full-Power Operation (Typical Schedule)

8%

-
 

 

 

59

. 9A-1 — Restart Equlpment After a Power Outage involved startlng the
dlesels, restarting equlpment and resetting tripped instruments. '

5J-1 — Return to Power After a Load and Rod Seram involved msnipu-
‘lating the rods and heat~removal system components to return to power. If
‘no instrument or equipment tests were necessary, thls ‘could- be done in
less than an hour. |

11A, 5I, and 57 — Return to Power After a Drain is divided into
3 parts. Part (1) assumes a drain to both drain tanks followed by a
transfer through the transfer lines, refill and return to power (thls is
the design method of transfer). Part (2) assumes a drain to both drain
tanks followed by a transfer through the fill lines, refill, and return to
power. (This method of transfer was used to save time. Jumpering of
interlocks was necessary.) Paft (3)'assumes a drain to only one drain

tank followed by a fill and return to power.
h 4.6 Reactor Startup After a Major Shutdown

‘ Each startup was dlfferent, requiring scheduling.between the com-
pletion of maintenance and start of operation. The cell membrane was
usually installed shortly'after the final closure of the primary.system.
Assuming this, and that maintenance on out-of-cell components was com- -
pleted by the time they were needed, a somewhat typical startup‘would be

as shown in Fig. L4-6.
4.5 Changes Made in the Plant

The number and klnds of changes made in a plant after it beglns
ioperatlon are influenced by two different things: original design and
":changlhg activities. 1In some_51ngle-purpose systems the number of changes
may simply reflect how well'the eriginal design met its goals. In an ex-
perimental plant sfich as'the}MSRE “most of the changes may be required for
experimental purposes or changes in the mode of operatlon. Ih any event,
'_tlme spent in meking changes affects the availability of the plant.
| Durlng the course of the MSBE operation many changes were made, but

only a few caused significant deley in the program.

 
 

60

After the end of construction and the non-nuclear checkout of the
MSRE, it was required that a change request be formally initiated, re-
viewed and approved before any modification was made. (Section 13B of -
rReférencg 22,7which prescribes the procedure, defines a modifigatioh as
"a change in the physical plant which produces a significently different
characteristic or function in any component or system.") In the 55 months
from June 1965 through December 1969, a total of 633 requests for changes
in the reactor system were initiasted, of which 512 were_approved. For the
chemical processing plant, 113 change requests were initiated of which 87
were epproved. _ ' 7 |

Table h—lQrsummarizes for each 6-month period the nufiber of reqfiests
initiated and approved fbr the reactor. The table also categorizes the
approved changes as to the reason for making the change and the type of
change that it was. As indicated, most of the changes were aimed at filling
the needs of normal 0peration and over half were changes in either instru-
mentetion or setpoints. | |

Table 4-20 surmarizes the change requests for the chemical processing | ci;
plant. ‘ ' '

Although we know how many changes were made, we cannot accurately sum

up the times required for meking the chenges. We can say, hc#ever, that in.
- the case of the reactor change requests, the vast majority either took
very little time to execute or were made while other work was going on.
Exceptions include the work on the fuel offgas system in the spring of
1966. In the -processing plant, the summer of 1968 was spent in teéting
and modifications, particularly of the fluorine disposal syétefi.
 

Summary of Reactor System Requests

Table L4-19

61

 

Type Change

I3Y30

10

29

 

muafiompom

12

1k

12

10

75

 

uo e jusundysuy

10 |

61

60

34

25

13

234

 

TeoTI309TH

17

46

 

Bupdid

7

14

18

100

 

TeoTUBYOSN

11

11

11

21

 

moleq

 

Reason for Change Requests

13430

19

 

serdureg TeTo9dg

14

 

anofiw#omxfi

26

 

90U TUIAUOCY

 

uoigexadp
TBWION

13
- 86

10k

59

25

11

12

12

373

 

TouuoOsIad puy

Juswd by JO

UO0F30930Id

12

37

 

Aq9383

‘5

14

 

" Number of

pasoxddy

gy

137

T6

25
30

23

7

21

512

 

- Requests

huonwvfivmm

Ea

126 | 115

173

103

69
3h

32

‘18

21

20

633

 

 

Period

JTeH |

'lsfa

ond

1st

2nd

2nd

1st

2nd

1st

2nd

 

 

Te9x

 

1965

1966

1967 | 1st

1968

1960

Total

After institution of change request procedure on June 21, 1965.

 

 

a
Summary of Chem Plant Change Requests

Table L4-20

62

 

Type Change

I5Y30

 

sjutodjag

12

 

UOTFBIUSTNI SUT

16

33

 

TBoTX309TH

 

Jurdig

17
10

37

 

T8O TUBLUISK

 

" nofeT

 

Reason for Change Requests

13430

 

soTdueg TeIoadg

 

- guswixadxy

 

~ 90U TUSAUO)

 

uotgerady
uotjexadp TewION

b2

25

'89

 

TouuOsIag puy

jusud by Jo
uotT19930ad

 

£q93eg

 

Number of

pasoxddy

k-

25

gl

 

Requests

_wwummzdum

56
30

113

 

 

JTBH

1st

2nd

1st

2nd

lst

2nd

1st

2nd

1st
2nd

 

Period

 

JIB3X

 

1965

1966

1967

1968

1969‘

Total'

 

 
 

 

 

 

63
4.6 Tabnlation'of:ReéordeduVariables athull power

It seemed desirable to record a set of readings of all of the various
| reactdr-variables'at normal conditions. However,‘even thbngh the MSRE ran

~ at full power much of the time, conditions were constantly changing due
mainly to the varilous- experiments which were performed ‘Also some of

the variables were not recorded regularly thus making it difficult to
seleet the best time. After consideration of these factors, the 12 to 8
shift on 10/12/69 was selected as a period at full nOWet when conditions
were fairly normal and considerable information was available. Values

were taken from the various recorder charts, routine control room and
building logs and the weekly, monthly, ‘and other check lists. These are
tabulated in Tables 4.2l and 4.22. A snapsbot listing all of the computer‘
inputs was also retrieved from the computer tapes for 0400 on 10/12/69.,
These values are listed in Table 4.23. The location of the sensing element
is deseribed briefly in the description column. The number and letters in
tbe'identification column correepond to those used in the design drawings m
and other design documents. ' They can be used to further\identify the

variable.

 
 

64

Table 4,21, Tabulation of recorded variables

A

 

Identification o | Description _?g?i;?gg
TE-100-A5 Line 100 (Reactor butlet) temperature | _120$°F_
TE—lOD—A6. Line 100 temperature | © 1169°F
TE-lOl-ZA Line 101 temperature ; | 12005F
TE-101-3A " Line 101 temperature | - 1215°F
TE-102-3A Line 102 temperature | - o 1160°F
TE~102-B4A Line 102 temperature o | | 11526F
TE-102=-5A Line 102 (Reactor 1n1et) temperature | 1173°F
TE-103-A1B Line 103 temperature ) | " 908°F
TE-103-B1 Line 103 temperature | - 798°F
TE-103-6 Line 103 temperature | | ~ 1150°F
TE~-103-8 .; , Line 103 temperature o | | | 1223°F
TE-103-B11 - Line 103 temperature — | | | | 983°F
TE-103-14B Line 103 temperature '1200°F
TE-104-5B Line 104 temperature : 918°F; |
' TE-106-5B Line 106 temperature  1080°F
TE-108-7 Line 108 temperature _ 222°F
TE-109-7 Line 109 temperature 227°F
TE-~200-B7B Line 200 temperature 1018°F
TE~200-D7B- Line 200 temperature | | 1025°F
TE-200-A8E Line 200 temperature 1035°F
TE-200-B8B Line 200 temperature ‘ 1010°F
TE-200-C8B Line 200 temperature - 1015°F
TE-200-A9B  Line 200 temperature 1010°F
TE-200-B9B Line 200 temperature . 1025°F
TE-200-14A Line 200 temperature 1018°F
TE-200-16B Line 200 temperature - 1015°F
TE-200-19A Line 200 temperature ' 1023°F
TE-200-AS~A1B Line 200 penetration temperature 965°F

TE-200-AS-B1A Line 200 penetration temperature 642°F
 

65

Table 4.21. '(continued)

 

 

 

Identifiqation Description. §87i§7§9
TE-200-AS-C1A Line 200 penetration temperature - 332°F
TE-200-AS—-2A Line 200 penetration temperature 363°F
TE-201-2B Line 201 temperature 1072°F
TE-201-5A Line 201 temperature 1062°F
TE-201~7B .. Line 201 temperature 1068°F
TE-201-A9B Line 201 temperature 1078°F
TE-201-B9B Line 201 tempé:ature : 1045°F
TE-201-~A10B Line 201 temperature 1073°F
TE-201-B10B Line 201 temperature 1059°F
TE-201-C10B Line 201 temperature 1086°F
- TE-201-A11B Line 201 temperature 1078°F
TE-201-C11B Line 201 temperature 1083°F
TE-201-D11B Line 201 temperature 1075°F
TE-201-AS-A1B  Line 201 penetration temperature 706°F
TE-201-AS-B1A Line 201 penetration temperature 772°F
TE-201-AS-C1B Line 201 penetration temperature 350°F
TE-201-AS-2A Line 201 penetration temperature 350°F
TdI-201A ‘Radiator AT (salt) 59°F
Xpr 201 Radiator powerrrflk 7.1 MW
 FR=201 @olant salt flow 860 gpm
TE-203-1 Line 203 temperature 870°F
TE—203¥2 . - Line 203 temperature 107°F
. TE-204-1A Line 204 temperature 1135°F
TE-204-2A Line 204 temperature - 1155°F
TE-204-3A Line 204 temperature 1205°F ..
| TE-204-4A Line 204 temperature  1200°F
TE-204-5A° - Line 204 temperature 1193°F
TE-204-6A Line 204 temperature 1170°F
TE-204-A7B Line 204 temperature 1178°F
TE-204-B7B Line 204 temperature 1170°F
TE-204-8B Line 204 temperature 1130°F

 
 

66

Table 4.21 (continued)

 

 

Identification - Description '§g7i§7§9
TE~-206-1A Line 206 temperature - 1073°F
TE~206-2A Line 206 temperature 1124°F
TE~-206-3A Line 206 temperature | 1147°F
TE-206~4A Line 206 temperature 1163°F
TE~206-5B Line 206 temperature . 1150°F
TE~-206-6A Line 206 temperature 1073°F
PI-407 Leak detector pressure 100 psig
RM-500 Cover-gas supply radiation 0.3 mr/hr
PI-500A Helium header pressure 223 psig

- FIC-500A Helium flow rate 6 liters/min.
PI-500G2 Helium treating station pressure 250 psig
PI-500M Reduced helium pressure 35 psig
PI-501A Reduced helium header pressure 34.5 psig
PI-510A 01l tank No. 2 pressure 7.5 psig
PR-511D Coolant drain tank pressure 4.8 psig
TE-512-1 Line 512 temperature 96°F
FI-512A Coolant pump purge gas flow 0.6 liters/min
PI-~513A 01l tank No. 1 pressure - 8 psig
TE-516-1 .Line 516 temperature 107°F
PR-516 Fuel pump pressufe 5 psig
FI-516B ~ Fuel pump purge gas flow 2.4 liters/min,
PR-522A Fuel pump pressure 5.2 peig
TE-524-2 Line 524 temperature 93°F
FI-524B Fuel pump upper off-gas flow rate 477
LI-524C 0il catch tank No. 1 level 147
FI-526C Coolant pump upper off-gas flow rate 0.04 liters/min.
LI-526C 0il catch tank No. 2 level 18%
RM-528 Coolant gas supply radiation 1.5 mr/hr
PR-528A Coolant pump pressure 5.1 psig

O
 

67

- Table 4,21 (continued)

 

 

TE-702-1A

Line 702 ;emperatufe

Ident}figetion Description ‘§§7i;7§9
A 0,I 548 Helium oxygen-centent' 0.4 ppm
A H 0. T 548 Helium moisfiuréftontent 2 ppm
FI-548A Helium oxygen anelyzer flow 100 cc/min
FI-548B Helium moisture analyzer flow 100 cc/min
PdI-556A Main eharcoal bed AP | 3.4 psi
RM-557 Charcoal beds off-gas radiation. 0.1 mr/hr .
'RM-565 -. Cell air radiation 1.5 mr/hr
FI-566 Reactor cell ai:”oxygen.analyzer flow 100 cc¢/min
A0z 1 566 Reactor cell aifaoxygen content 2.7%
PR-572B Fuel drain tankrNoL 1 pressure 5.9 psig
PR~574B - Fuel drain tank”No. 2 pressure 5.5 psig
PR-576B Fuel flush tank pressure - 5.0 psig
PI-589 Orerflow tank pressure : 5.1 psig'
FI-589 Overflow tank bubbler flow rate . 27 psig
PI-592 . Fuel pump pressure 5.4 psig
FI1-592 Fuel pump bubbler flow rate_r- 25 psig
FI-593 Fuel pump bubbler flow rate . 23 psig
LR-593C Fuel pump level 53%
FI-594 Coolant pump. bubbler flow rate 24.8 psig
FI&SSSe:"' Coolang pump bubbler flow rate 25.5 psig
',LR-SQSC . Coolant pump,leve}r, 57%
_ 7F14596 L Fuel pump bubblef flow rate - 25.2 psig
RM-596 Fuel pump gas supply radiation 0.1 mr/hr
7Fi—598tf; : Coolant'pump“Bubbiefiflow'ratel 24.0 psig
 FI-599 . Overflow tank bubbler flow rate - 27.0 psig
1—599B5. Overflow tank level | 3 - 23%
FI—600A~e Overflow tank bubbler flow rate - 26.5 psig -
| LI-600B . Overflow tank level | 247
PI-701A - Fuel oil pump No. 1 pressure C 10 psig
PI-702A Fuel oil pump No. 2 pressure 64 psig
* 135°F

 
AR it b e b e

 

 

 

 

68

Table 4.21 (continued)

 

 

Identificétion Description --?g?i;?gg
FI-703 Fuel pump lube o0il flow rate 3.8 gpm
FI-704 Fuel pump coolant oil flow rate 8.2’gpm
PI-751A Coolant oil pump No. 1 pressure 8 psig
PI-752A Coolant oil pump No. 2 pressure 56 psig
TE-752-1A Line 752 temperature 130°F
FI-753 Coolant pump lube oil flow rate 4.0 gpm
FI-754 Coolant pump coolant oil flow rate 6.7 gpm
LI-806A Steam dome water level (FD-1) 0%
LI-807A Steam dome water level (FD-2) 0%
FI-810 Condenser water flow rate (FD-1) 40 gpm
FI1-812 Condenser water flow rate (FD-2) 40 gpm
FI-817 Offgas particle trap water flow rate 4 gpm
TI-820-1 0T-1 water outlet temperature ° 87°F
TI-821-1 O0T-1 water inlet temperature 80°F
FI-821A 0T-1 water flow rate 8.4 gpm
TI-822-1 0T-2 water outlet temperature 81°F
TI-823-1 0T-2 water inlet temperature . 77°F
FI-823A 0T-2 water flow rate 8.6 gpm
TI-826 Treated water cooler (TW out) temperature 98°F
RM-827 Process water radiation 3 mr/hr
TI~829 Treated water cooler (TW in) temperature 108°F
PI-829A Treated water pump pressure 70 psig
FI-830 Fuel pump motor cooling water flow rate 4.6 gpm
FI-832 Coolant pump motor cooling water flow rate 4.8 gpm
FI-836A ~Drain tank cell space cooler water flow rate 63 gpm
FI-838A Reactor cell space cooler No. 1 water flow rate 53 gpm
FI-840A Reactor cell space cooler No. 2 water flow rate 59 gpm
FI-844A Thermal shield water flow rate 50 gpm
PI-851A Cooling tower water pump pressure 34 psig
FI-851C Cooling tower water to cooler flow rate

273 gpm
 

69

Table 4.21 (continued)

 

 

Identifieatidn  Description ?g?i;?gg
TI-854 Treated water cooler (CIW out) temperature 84°F

TI-858 Cooling tower water temperature 80°F

FI—859 - Thermal shield slide water flow rate 7 gpm
FI-862A Coolant cell‘spaee'eooler (west) flow rate 20 gpm
FI-864A Coolant cell space cooler (east) flow rate 20 gpm
F1-873 CCP gas cooler water flow rate 15 gpm |
FI1-875 CCP No. 1 and 2 oil cooler water flow rate 6.3 gpm .
TI-881-2A Air compresser ‘cooler outlet water temperature 86°F
TI-881~2B Air compresser head outlet water temperature 92°F
' PdI-900A CCP AP I o 8 psi |
PI-927A Ventilation filter suction 2.2 in. H;0
PdI-927B2 Ventilation stack filters AP 3.3 in. H,0
PI-927C .VentilationISteEkifan suction 5.5 in. H,0
PAI-937A TR to 840 level AP ~ | 0.05 in. H0
PAI-938A SESA to TR AP 0.04 in. H,0
RM-6000-1 Reactor cell radiation 50,000 R/hr
RM-6000-2 Reactor cell radiation 30,000 R/hr
RM-6000-3 Reactor cell radiation 50,000 R/hr
RM-6000-4 Drain tank celi'i-radietipn ' 10 R/hr
RM#GGOb—5  Drain tank~celi redietion' 60 R/hr
RM-6000-6 Drain tank cell radiation 500 R/hr

. RM-6010 Coolant cell radiation 100 R/hr

'RRr8100 | Reactor power'e" 8.5 MW
RR~8200 | Log reactor power 7 MW
FI-9000 Instrument 'air flow - 20%
PIC-9006-1 Emergency Ngrsetpbint7'” 65 psig
FI-9006 - Instrument air flow (emergency header) 447
PdI—AfiZQAZHI Radiator air pressure drop | 677

ZI-AD2 Bypass damper position 10%
FI-ADBAT_ "Radiator stack flow rate 70%

TI-AD Radiator air inlet temperature 71°F
TI-AD3-8A Radiator air outlet temperature 190°F

 
L

 

70

Tahle 4,21 (continued)

 

 

Identification .Description ' ?g?i;?gg
PI-AR-1 Instrufient'air receiver tank pressure 88 psig
TE-CC-1 . Coolant cell ambient temperature 88°F
TE-CC-2 ” Coolant cell ambient temperature . 90°F
TE-CC-3 Coolant cell ambient temperature 78°F
TE-CC-4 Coolant cell ambient temperature 80°F
TE-CC-5 Coolant cell ambient témperature 76°F
TE~-CC-~6 Coolant cell ambient temperature 115°F
TE-CC-7 Coolant .cell ambient temperature 105°F
TE-CC-8 Coolant cell ambient temperature 103°F
TE-CDT-1A Coolant drain tank top temperature 1195°F
TE-CDT-3A Coolant drain tank top temperature. 1182°F
TE-CDT-5A Coolant drain tank middle temperature 1200°F
WR-CDT-C1 Coolant drain tank weight 3%

Eil COP-1- - Coolant o0il pump No. 1 current. 0 amps
Eil COP-2 Coolant oil pump No. 2 current 8.7 amps
TE-CP-1B Coolant pump flange-neck temperature 245°F
TE-CP-BZA' Coolant pump bowl-neck temperature " 786°F
TE-CP-C2A Coolant pump bowl-neck temperature 795°F
TE-CP-3B Coolant pump bowl top temperature 957°F
TE-CP-4A Coolant pump bowl top temperature 965°F
TE‘C?PSA Coolant,pump bowl top temperature . ' 965°F
TE-CP-6A Coolant pump bowl bottom temperature 1042°F
TE-CP-7A Coolant pump bowl bottom temperature | 1020°F
TE-CP-8A - Coolant pump bowl flange top temperature 107°F
TE-CPM-1B Coolant pump motor temperature 100°F
SI-CP Coolant pump speed t 1780 rpm
E1I-CP Coolant pump current 51 amps
EwI-CP Coolant pump power | 56.5 kW
LR-CPA Coolant pump level (float) . 4.8 in.
TE-CPLE-A2 CP level element pot (lower) temperature 1140°F
 

71

‘Table 4.21 (continued)

 

 

Identification ~ Description §87i;7§9
TE-CPLE-A4 CP level element pot (upper) temperature 1097°F
TE-CPLE-A5 CP level element pipe (top) temperature 1102°F
TE-CPLE-A6 CP level element sensor (lower) temperature 515°F
TE-CPLE-A7 CP level element-sensor (upper) temperature 372°F
TE—CR~121V Coolant radiator inlet pipe temperature 1060°F
VTEQCRQIZS. Flow venturi pipe temperature 1020°F
LI-DC Decontamination- cell sump level 0 in.
LI-DTC),T Drain tank cell sump level 0 in; |
TE-FD-1-2A Fuel drain tank No. 1 (upper) temperature | 1115°F
TE-FD1-13A Fuel drain tank No. 1 (middle) temperature 1170°F
TE-FD1-16A Fuel drain tankwNo. 1-bayonet (top) temperature 1130°F
TE-FD1-17A Fuel drain'tank No. l_bayonet (upper) temperature 1146°F
TE—FD1418B ) - Fuel drain tank No. 1 bayonet (center) temperature 1158°F
TE-FD1-19A Fuel drain tank-No.nl bayonet (lower) temperaturej lléidF
TE~FD1-20A Fuel drain tank)No,ll bayonet (bottom) temperature 1161°F
WR-FD1C  Fuel drain tank No. 1 weight 0%
TE-FD2-2A Fuel drain tank No. 2 (upper) temperature 1140°F
TE-FD2-13A  Fuel drain tank No. 2 (middle) temperature - 1150°F
TE-FD2w16A Fuel drain tank,No. 2 bayonet (top) temperature 1115°F
E—FD2—17A Fuel drain tank_No._2_bayonet.(upper)'temperature 1130°F
TE—FDZ-lBB Fuel drainitank No;VZ ba&onet'(center) temperature 1140°F
TE-FD2-19A Fuelldrainrtank No. 2 bayonet (lower) temperature 1140°F |
TE-FD2-20A - Fuel drainltank No.(? bayonet (bottom) temperature 1140°F
WR-FD2C Fuel drain tank No. 2 weight o - 5%
TE—FF—lOO—Z Freeze flange 100 center temperature 885°F
TE—FF—102-2_ Freeze flange 102 center temperature 974°F_
TE—FF—ZOO-Z Freeze flange,ZQQ-center temperature:- 754°F
TE-FF—201-2 Freeze flange 201'center temperature 806°F
TE—FFTeSA - Fuel flush tank (top) temperature 1132°F
TE-FFTI-7B Fuel flush tank (middle) temperature 1174°F

 
 

72

Table 4.21 (continued)

 

 

‘ FT-201B-4A

_Ideptifieation Desertpt;on_ ?g?g;?%g
WR-FFT-C Fuel flush tank weight 69.6%
PAI-FI-Al Vent. system roughing filters AP 0.6 in. E;0
PdI-FI-A2 Vent. system ebsolute filters AP 1. 6 in. HzO
E{iI FOP-1 Fuel oil pump No, 1 current 0 amps
EiT FOPQZ Fuel ofl pump No, 2 current 8.2 amps
TE-FP-1B Fuel pump neck-flange temperature 305°F
TE~FP-2B Fuel pump neck temperature - 559°F
TE-FP-3B Fuel pump neck-bowl tempereture 962°F
TE-FP-4B Fuei‘pufip wal top temperatpre 992°F
TE-FP-5B Fuel pump neck-bowl temperature ~ 990°F
TE-FP-6A Ffiei pfimp bowl top temperature 1030°F
TE?FP+9B Fue1 pump neck-bofil temperature 955°
TE-FP-lOB " Fuel pump Bowl_top temperature 1002°F
TEwF?-liB Fuel pump flange top temperature 150°F
TE-FP-12B Feel‘pump bowl center,temperatere'- 998°F
TE-FPM-1B Fuel pump motor temperature - 120°F
SI-FP Fuel pump speed N 1176 rpm
Ei-FP Fuel pump current 44.5 amps
Ew-FP Fuel pump power | 34 kW
LI-FSC Fuel storage cell sump ievel | 1.1 in.
FT-201A-1A Flow element 20l1A pipe temperature 1130°F
FT“ZOlAFZA Flow element 201A top flahge temperature 1210°F
FT-201A-3A Flow element 201A bottom flange temperature 1160°F
FI-201A-4A Flow element 201A pipe temperature 1130°F
FT-201A-5A Flow element 201A top flange"temperature 1210°F
FI-201A-6A Flow element 201A bottom flange temperature '1250°F
FT-201B-1A Flow element 201B pipe temperature '970°F
FT-201B-2A Flow elefient”ZOIB top flange temperature | 1180°F
FT-201B-3A Flow element 201B bottom flange temperature ~1170°F

Flow element 201B pipe temperature 1130°F
 

 

 

 

73

s

 

 

TE-FV-204-1B

Freeze .

shoulder temperature

| Table 4,21 (continued)'
Ideutifieation ;rbeseription arig?ié?ES
FT-201B-5A Flow element 201B top flange temperature 1200°F
FTfZOLB—GA Flow element 2018 bottom flange temperature !1130°
“TE-FV-103-1B Freeze_val§e7103 shoulder temperature - 1000°F
TE—FV—lOBeZA Freeze ralVep103 center temperature .390°
TE-FVf103;SB ‘Freeze valve_losushoulder temperature _5@O°
- TE-FV-104-3B Freeze valve 104 shoulder temperature 450°F
_TE-FVf104-BA Freeze valve 104 adjacent’pipe,temperature 450°F
TE~-FV-104-5B Freeze valve;lO& pot temperature o 590°F |
TE~FV-104-6B Freeze valve 104 pipe temperature 650°F "__
TE—FV—105-2A Freeze'vaiue:idsécenter temperature . 1250°F_L'
TE—FV-lOS:AAA Freeae.valvelOSjadjacent pipe temperature: 1160°F;1,-
TE—FV<105?B4A Freeze valve-losmadjaeeut pipe temperature' 1190°F,
TE-FV-105-5B Freeze valvejibsgpot‘temperature o o 1080°
TE-FV~-105-6B Freeze valve 105 pipe,temperature.. 1120°F |
TE—FV-lOSéZA Freeze valve 106;ceuter.temperature 1215'F
TE-FV;106-A4A Freeze'value-106'adjaaent pipe temperature” 1140°F
TE-FV-106-B4A Freeae.valve106nadjaeent pipe'temperature _ 1190°F
TE-FV-106-5B Freeze valve ififipot temperature | | 1120°
TE-FV-107-1A Freeze_valve.1Q7rehbu1der,temperaturep 500°F
TE;FV?107;BB Freeze_value_;p7shouider temperature‘_‘_ 4909
TE—FVe107—A4 7 Freeze valve;;07madjaeeutpipe,temperature. S530°F
'TE-FV—107-SBr_ )Freeze'va}vei07"pottemperature; B 610°
TE-FV-107-6A -FreeZevalve-io?.pot temperature'--” - 1540° .
TE-FV=108-SB Freezemalye;iOBishoulder temperature,q:t,; 440°.
,VTE;FV-108-34 __FreEie,vaire ibé;adjacent pipe temperaturer ASOPV:gfl
_rTEéFV;108;SB Freeze valve 108 pot temperature - 640°F
TE—FVQiOB-GA ‘Freezemvalve ifiBepot temperature '5436?;
TE-FV-109-1B ~ Freeze valve-lOQfshoulder temperature o 470°F
| TE;FV-IOQfGA Freeze valve 109 pot temperature 595°F
valve 204 7909?

 
 

 

74

Table 4.21 (continued§

 

 

Identification Description‘ ?fi?i;?gg
TE~FV-204-24 Freeze valve 204 center temperature 235°F
TE-FV-204-3B Freeze valve 204 shoulder temperature 830°F
TE-FV-206-1B Freeze valve 206 shoulder temperature 710°F
TE-FV-206-2A Freeze valve 206 center temperature 300°F
TE-FV-206-3B Freeze valve 206 shoulder temperature 830°F
TE-H 103 fieater 103 ;éfiperature | 908°f
PI-HB-A Hiéh bay préésure‘ | - =0.19 in. H20
TE-HX-1A Heat exchahgércoolént out -temperature 1060°F
TE-HX-4A Heat exchanger\cbolant in temperature 1180°F
TE-HX-7A Héat-exchénéér shell (center) temperature 1165°F
ZI-ID-A Inlet radiator door position 82.5%
TE-OFT-1 Overflow tank pipe temperature 934°F
TE-OFT-2B Overflow tank top temperature 1202°F
TE~OFT-4 Overflow tank side temperature 1190°F
ZI-0D-A Outlet radiator door position 79.5%
LI~0T1A3 Fuel oil .supply tank level 647
RM-0T1 Fuel o0il supply tank radiation 1.7 mr/hr
LT-0T2A3 Coolant o0il.supply tank level 58%
RM-0T2 Coolant oil supply tank radiation 0.1 mr/hr
TIC-0; R1-1 Helium oxygen removal No. l‘temperature 790°F

TIC 0; R2-1 Helium oxygen removal No. 2 temperature . 1235°F
TIC 0, R1-2 Helium oxygen removal No. 1 wall temperature 513°F

TIC 0; R2-2 Helium oxygen removal No. 2 wall temperature 860°F

TIC PE 1 Helium preheater No. 1 temperature 790°F

TIC PH 2 Helium preheater No. 2 temperature 800°F

TE PT-1 Particle trap temperature 360°F

TE PT-2 Particle trap temperature " 360°F

TE PT-3 Particle trap temperature - 360°F

TE R-5A Reactor top temperature 1206°F
TE R-6A Reactor top temperature 1210°F
 

 

75

\"Table 4.21 (continued)

 

 

Iantification , | - Description | L v§g7i;7§9

TE R-7A ' Reactor neck'temperéture, - 799°F

TE BréA Reactor neck temperature | | 680°F

TE R-9 Reactor néék"temperatfire‘. o o | 600°F --..

TE R-10 | Reactbr‘neck temperature‘, - 534°F

TE er? B Reactor side temperature l . 1190°F

TE R-23A Reactor side temperature o | 1185°F

TEKRrBZAV - Reactor bottom temperature . 1170°F

TE R-34 . Reactor neck flénge temperature | 233°F

TE R-35 - Reactor neck‘flange:tempe;ature 198°F

TE R-36A | Reactor control rod No. 1 (upper) temperature 449°F

TE R-37A ) | Reactor cqntrp%rbd No,lz (upper) temperature 431°F

TE R-38A Reactor con;?blirbd‘Nq. 3 (upper) temperature 460°F

TE R-39A Reactor cofitrol;bd,No. 17(1ofier) tem@g;ature 855°F

TE RrQOA,E- Reactorlcontrol-rod No;HZI(lowér) temperature 786°F

TE R-41A | ‘Reactor control rodNo; 3_(lower) tempergture 689°F

TE R-43B Reactor graphite tubg”(;dfier)'fémperatfire- 1025°F

TE R-44A Reactor neck (bottom) temperature o 1218°F

TE R-46A ~ Reactor neck (upper) temperature | o ~ 250°F

TE fif47 i . Reactor neck (uppe:)”;emperatfire | - 215°F

TE'RtAS o Réactor neck (fiééér).temperature, _:__T: o 212°F
" TE R-52 Reactor thermal well temperature | L ~ 810°F

LI RC-C ‘Reactor cell sump level . - 0 in.
'._PITRC=A ' Reactor cell pressure - =25 psig

FI-1 Containment stack flow ~  15%
VRM—SIA_:f _:_ - Cdntainmeng_éfiéék;éléhé. ; S e 100 cpm
':BMngB. | Contaififiéflt?éfi#ck“beta gafima o 11 cpm
CRM-SIC _COntéinméht'é;g@k;iog;né_  :3- o ' ” - 530 cpm
r;LI;SC;A S 'Storage'éell.sfimp;igvel ' - 1.0 in. H0

‘LI TC-A ) Spare cell.gfifip'iéfiél | S 0.4 in. Hz0

PI VI-1- Vapor suppreséiqn tank pressure 0.4 psig

LI WI-A Waste tank level 107 in. H20

LI WTC-A Waste tank cell level | 2.6 in. H30

 
 

 

~ Table 4.

76

21 (continued)

 

 

 

l1dentification " Description ?g?i;?ég
Main blower vibration <1 mil

OACOT Official average coolant outlet temperature 1011.5fF

OAFOT Official average fuel outlet temperature 1208.3°F
Control rod No. 1 position 35.5 in.
Control rod No. 2 position 44 in.
- Control rod No. 3 position 43,1 in.
Fission chamber No., 1 position 60 in.
Fission chamber No. 2 position 65.7 in.
Fission chamber No. 1 count rate 10* cps
Fission chamber No. 2 count rate 10“cps
Control rod No. 1 clutch current 143 ma
Control rod No. 2 clutch current 144 ma

" Control rod No. 3 clutch current - 147 ma

Motor generator 2 current 28 amps
Motor generator 3 current 32 amps
Motor generator 2 voltage 52 volts
Motor generator 3 voltage 52 volts
Battery voltage ‘ 50.5 volts
Motor generator No. 1 voltage 260 volts
Motor generator No. 1 current 130 amps
Inverter voltage 206 V
Inverter current 89 amps
Main blower No. 1 current v260 amps
Main blower No. 3 current _ 280 amps
Component coolant pump No. 1 current 0 amps
Component coolant pump No. 2 current 88 amps
Instrument air dryer purge rate 12 cfm
Table 4,22,

77

(average of 3 phases)

Tabulation of recorded heater current on 10/12/69

 

 

H203-1

werrio oS crpein B e St
H-CR-1 15 H203-2 0 RCH-5 12,
H-CR-2 17 H204-1A 3 RCH-6 15
H-CR-3 23 LE-CP-1 4 RCH-7 8
'H-CR-4 18 LE~CP-2 7 H102-2 13
H-CR-5 18 FV204-1 2.5 R-1 18
H-CR~6 19 FV204-2 1 R-2 19
H-CR-7 15 FV206-1 2 R-3 19
H-CR-8 25 FV206-1A 1.5 HX-1 0
200-13 17 H204-1 15 | HX-2 16
' 201-12 13 H206-1 1 HX~3 .16
202-2 14 CDT-1 15 FP-1 6
200-14 6 CDT-2 11 FP-2 6
200-15 10 CDT-3 12 RAN-1 0
201-10 11 cp-1 13 RAN-2 0
201-11 6 cP-2 12 200-16 2
201-13 H200-1 10 201-14 1
202-1 7 | H200-11 12 102-1 3
204-2 10 _ H200-12 14 522 0
205-1 7 | H201-1 13 102-4 9
© 204-3 6 H201-2 10 102-5 1
. FT201A1 5  H201-9 16 103 26
 FT201A3 5 . H100-1 2 FV-103
 FT201A2 6 CRGH-L 16 H-104-1 8
- FT201Ar 4 RCH-2 13 FV-104-14 4
" FT-201BL 6  RGE-3 19 FV-105-1A 11
_ FT201B3 5 | ORCH-4 21 FFT-1 16
' FT201B2 6  H100-2 18 FFT-2 15
FT201B4 6 H101-2 11 FD-1-1 18
0 H101-3 13 FD-1-2 17
 

‘78

Table 4.22  (continued) -

 

 

Description ?g?i;?gg Descripti9n_. ?;7%;729 | f Déécription » §g7i;7§9
FD-2-1 18 FV-104-~3 11 FV-108-3 3
FD~2-2 13 FV-104-4 10 FV-108-1 1
FV-104-1 | FV-105-2 ' FV-108-3 5
FV-104-3 FV-105-3 FV-109-1 3
H~-104-5 10 FV=104-7 13 FV-109-2 4
H-104-6 8 FV-106-2 8 FV-109-3 2
FV-105-1 12 FV-106-3 10 FV-109-1 Sl
FV-105-3 11 FV-110-2 0 - FV-109-3 5
FV-105-1 7 FV-110-3 0 FV~110-1 0
FV-105-4 11 FV-107-1 T2

FV-106-1 11 FV-107-2 3

FV-106-3 9 FV-107-3 3

FV-106-1 6 FV-107-1 1

FV-106-4 FV-107-3 5

FV~106~1A 11 FV-108-1 3

FV-104-2 9 FV~-108-2 5
 

 

79

Table 4.23., Computer snapshot taken
at 0400 on 10/12/69

 

 

228 Main charcoal bed pressure drop

Scan
Identification Seq. o : Description Reading
No. ' 10/12/69

" EWM-CP-D 256 Coolant pump power 31.8 kW
EWM-FP-D 255  Fuel pump power 34.8 kW
FqI-569 349  Reactor cell evacuation flow 1.27rlit2r/min
FT-AD3-A 40 Radiator stack flow | 195,000 cfm
FT481?A 242  Containment stack flow 20,200 cfm
FT-201-A 15 Coolant salt flow 849 gpm
FT-201-B 28  Coolant salt flow 840 gpm
FT—SIZ—A' 236 Coolant pump purge gas flow 0.65 liter/min
FT-516-E 234 = Fuel pump purge"gas flow 2.39 1liter/min
FT-524-B 235  Fuel pump upper offgas flow 133 cc/min
FT-526-C 237 Coolant pump upper offgas flow 83 cc/min
FT-703-A 238  Fuel pump lube oil flow ‘ 3.75 gpm

| FT-704-A 239  Fuel pump coolant oil flow 8.04 gpm
FT-‘753-A- 20 Coolant pump lube oil flow 3.87 gpm
FT-754-A 241 Coolant pump coolant oil flow 6.54 gpm
LE-CP-A 65 Coolant pump level 4.6 in. salt
LT-0T-1-A 248  Fuel oil itank level 12.4 in. oil
LT-0T-2-A 250  Coolant oil tank level 11.1 in. oil
LT-524-C 251 0il catch tank No. 1 level 11.3 in. oil
LT-526 252 01l catch tamk No, 2 level 16.0 in. oil
LT-593-C ' 50  Fuel pump level 6.1 in. salt
-LT—SQS;C-: 61  Coolant pump level 5.6 in. salt
LT-596-B 54  Fuel pump level 5.1 in. salt
LT-598-C 62 Coolant pump level 5.5 in. salt

- LT-599-B 57 Overflow tank 1evel - 5.6 in. salt
LT-GOO-B. 58 Overflow tank level 6.3 in. salt
PDT—ADZ-A 24  Radiator air pressure drop 9.1 in. Hz0
PDT-556-A 3.4 psi

 
 

80

Table 4.23 (continued)

 

 

Identification 222? ‘ Description Reading
| No. c 10/12/69
PDT-960-A 230 Component coolant pump AP 8.0 psi
PT-HB-A 233 High bay pressure -.07 in. H,0
PT~RC-A 348 Reactor cell pressure -1.98 psig
PT-500-A 223 Helium header pressure 220 psig
PT-510 27 0il tank No. 2 pressure 6.8 psig
PT-511-C 225 Coolant drain-ténk pressure -- psig
PT-511~D 218 Coolant drain tank pressure 5.7 psig
PT-513-A 226 0il tank No. 1 pressure- 7.7 psig
PT-516 347  Fuel pump pressure 5.0 psig
PT-517-A 224 Drain tank suppiy pressure 8.5 psig
PT-522-A 13  Fuel pump pressure 5.0 psig
PT-528 66 Coolant pump pressuré 4,6 psig
PT-572-B 219  Fuel drain tank No. l‘préssure 5.2 psig
PT-574-B 220  Fuel drain tank No. 2 pressure 5.1 psig
PT-576-B 221 Fuel flush tank pressure 5.3 psig
PT-589-A 53 Overflow tank pressure 7.2 psig
PT-592-8 33 Fuel pump pressure | 5.6 psig
PT-608-B 222 Fuel storage tank préssure ~1.2 psig
RE-NLC1-A 32 Reactor power 8.6 MW
RE-NLC2-A 36 Reactor power 8.6 MW
RE-0T-1-B 262 0il tank No. 1 radiation 1.8 mr/hr
- RE-0T-2-B 263 01l tank No. 2 radiation 0.09 mr/hr
RE-5C1-A1 9  Reactor Power 8.5 MW
RE-S1A 277 Containment stack alpha 14% scale
RE-S1B 278 Containment-stackbeta-gamfia 7Z scale
RE-SIC 270 Containment stack iodine 21% scale
RE-500-D 261 Cover.gas supply radiation 0.3 mr/hr
VRE-SZS-B 275 Coolant gas supply radiation 1.5 mr/hr
RE-528-C 276 Coolant gas supply radiation 2.1 mr/hr
RE-557-A 273 Offgas from charcoal beds radiation 0.1 mr/hr
 

 

81

" Table 4.23 (continued)

 

Scan

 

~ Identification Seq. . - ‘Description R . . Reading
B : No. .o 10/12/69
RE-557-B 274  Offgas. from charcoal beds radiation 0.1 mr/hr
RE-565-B - 271 © Cell air radiation L _ 1.0 mr/hr
RE-565-C 272 Cell air radiation ‘ - 1.4 mr/hr
RE-596fA' 280 Fuel pump gas supply radiation E | 0.1 mr/hr
RE-596-B 282  Fuel pump gas supply radiation o | 0.1 mr/hr:.
RE-596-C - 283 ~Fuel pump gas supply radiation- 0.1 mr/hr
RE-675-A 284  Sampler cold offgas - : 1.1 mr/hr
RE-675-B . 285 = Sampler cold offgas . : 4.2 mr/hr
RE-678-C 286 Sampler hot offgas \ : . 2600 mr/hr
RE-678-D 287  Sampler hot offgas - T 4000 mr/hr
RE-827-A - - - 264 - Process water radiation o 25 mr/hr
RE-827+B - 265 Process water radiation o 26 mr/hr
RE-827-C 266 -iProcess]fiater radiation _ , 34 mr/hr
RM-NCC1-A6 259 - Reactor count rate - 10,000 cps.
RM=NCC2-A6 260-l Reactor:cfiunt,ratg _ : - ; 10,000 cps”
RM-NCC1-A7 S Reaétorzpbwerf-- e S 9.6 MW
RM-NCC2~A7 - 42 . Réactorgpower,-: . - 9.2 MW . - :
RM-NCC1-A9 44 Reactor,§2riod - - - - -300 sec: -
RM-NCC2-A9 .45 - Reactor period - .. - - ~150 sec ..
SE~CP-Gl-A 52 ‘:Coolantipumfi'speed, SR L - 1775 rpm -
SE-FP-El-A 11 Fuel pump speed . - - 1190 rpm
.TEeADi—lA:. 184 - Radiatoisifilet air temperature . B7°F- |
_TE-AD3¥4 185 fiRadiatofébutiét duct wall temperature 107°F
TE-AD3-5A 186 Rédifitdr}ofitlet~duct wall temperature. -  120°F
TE-AD3-6 = 187. Radiatdrgbutlet?duct.wall temperatfire, | 129°F
TE-AD3-7A 188 - Radiétdrbfitlet duct,wall'tempefature 173°F.
TE-ADS-SA__ . 190 Radiatbfogtletjair-temperature S ©  178°F
TE-CDT=2A 181 . Coolant drain tank bottom temperature ~  1210°F
TE-CDT-8 182 ' Coolant drain tank middle temperature 1203°F -

TE-CP-A2B 127  Coolant pump bowl-neck temperature - | 804°F

 
 

82

 

Table 4.23,-(continuéd)

 

 

110

Freeze flange 101 middle temperature

Identification 22:? Description ~ Reading
. No. 10/12/69

TE-CP-1A 126 Coolant pump flange-neck temperature 247°F
TE-CP-3A 128 Coolant pump bowl top temperature 955°F
TE-CP-8B 125 Coolant pump flange top temperature 102°F
TE~CP~9A 129 Coolant pump bowl middle temperature - 1018°F
TE-CPM-1A 124 Coolant pump motor temperature | 96°F
TE~CIS-D 90 Surveillance rig top zone temperaturei 1200°F
TE-CTS-E 88 Surveillance rig mid zone temperature 1250°F
TE=-CIS-F 87 Surveillance rig bottom zone temperature | 1220°F
TE-DBE-1 267 Diesel house ambient temperature |  76°F
TE-DL-1 137 Computer room ambient temperature 71°F
TE-DL-2 138 Computér reference thermal plane temperature 69°F
TE~-DIC-1 303 Drainffank-cellwambient temperature 148?F
TE-DTC-2 304 Drain tank cell ambient temperature 144°F
TE-DTC-3 305 Drain tank cell ambient temperature 151°F

- TE~DTC-4 306 Drain tank cell ambient temperature 145°F
TE-DTC-5 307 Drain tank cell ambient temperature 149°F
TE~-DTC-6 309 Drain tank.-cell ambient temperature 150°F
TE-FD1-1A 166 Fuel drain tank No. 1. top temperature 1073°F
TE-FD1-3A 163  Fuel drain tank No. 1 bottom temperature '1149°F
TE-FD1-12A 164 Fuel drain tank No. 1 middle temperature '1178°F
TE-FD1-18A 167 Fuel drain tank No. 1 bayonet temperature 1158°F
TE-FD2-1A 171  Fuel drain tank No. 2 top temperature | 1062°F
TE~FD2~3A 168 Fuel -drain tank No. 2 bottom temperature 1130°F
TE-FD2-12A 169 Fuel drain tank No. 2 middle temperature =  1159°F
TE-FD2-18A 172  Fuel drain tank No. 2 bayonet temperature 1138°F
TE-FF100-4 106 Freeze flange 100 inner temperature | 904°F
TE-FF100-5 107 Freeze flange 100 middle temperature 648°F

- TE-FF100-6 108 Freeze flange 100 outer temperature - 559°F
TE~-FF101-4 109 Freeze flange 101 inner temperature 824°F
TE-FF101-5 585°F
 

 

 

83

Table 4.23 (continued)

 

 

Identification g;:? Description ‘Reading
B No. | 10/12/69
TE—FF101-6 111 Freeze,flemgeplfil_buter temperature ' 454°F
TE—FF102-4 112  Freeze flange 102 inner temperature 868°F
- TE-FF-102-5 113 Freeze_tlange 102 middle temperature 631°F
TE-FF-102-6 114 FreezetflangepIOZ outer temperature 541°F
TE—FFZOOfA 115 Freeze flamge 200 inner temperature - 769°F
TE-FF200-5 116 Freeze flange 200 middle temperature 546°F
TE—FFZOO—G 117 Freeze f;ange_2000uter temperature 477°F
‘TE-FFZOi—4, 118 Freezefflamge 201 inner temperature 796°F
TE-EF20175 120 Freeze flange_ZOl middle temperature 605°F
TE—FF26146‘ 121 Freeze_flange_201 outer temperature 495°F
'TE-FFTélA 175 Fuel flush tank top temperature 1134°F
TE-FFT-2A '173 .Fuel f;ush_tankrbottom temperature 1157°F
TE-FFT—lO-V 174 Fuel il&sh'tank middle temperature 1174°F
- TE-FP-1A 92  Fuel pumpneckeflange"temperature - 303°F
TE-FP-2A 93  Fuel pump neck tempereturev 7 518°F
TE;FP—3A; 94 Ffiei pmmp neck-bowl temperature 962?F
TE-FP-4A 98 Fuel pump_bqwl_top'temperature 1004’F
TE—FPQSA 95 Fuel pumprmeek-bowl,temperature :984°F,
|  TE~FP-7B 102 Fuel.pfimp~bqw1,;pwer temperature. {1212iF
TE—FP-SBt..] -103 Fuel pump;béwlybdttpm.temperature,»_ 1208°F
"TEeFP—QA:': 96  Fuel pumptfieek-bowl temperature - 950°F
' TEQFPQIOA 97 Fuel_pumpbowl_topetempereture 992°F
__TEfFP?ilA" 100 ?Feelfpumpffiefige'top temperature 146°F
TE-FPM-1A _104 fFuel-pfimpfmotor temperature | 117°F
7”TE-FST—10:' 217  Fuel storage ‘tank temperature S 81°F
':TE-FVIOB-ZB 39 _Ereeze_valve_103 center temperaturer 407°F
VTE-FV104elB | 30=i'FreeZe_ve1ve104Lshoulder_temperature 462°F -
TE-FVLDS%2B"* 47 _Freezeya;Yeplqszeenter temperetere_f”' 1236°F
TE~FV106-2B 51  Freeze velre 106_center temperature 1214°F .
TE-FV107-1B 144 Freeze valve 107 shoulder temperature 102°F

 
 

 

 

 

84

Table 4.23 (continued)

 

 

Particle trap temperature

Identifiéation ggz? Description Reading-
No. 10/12/69
TE-FV107-2B 145 Freeze valve 107 center temperature 490°F
TE-FV107-3B 146 Freeze valve 107 shoulder temperature 90°F
TE-FV108-1B 147 Freeze valve 108 shoulder temperature 77°F
TE~-FV108-2B 148 Freeze valve 108 center temperature 446°F
TE-FV108-3B 149 Freeze valve 108 shoulder temperature 76°F
TE-FV109-1B 150 Freeze valve 109 shoulder temperature 103°F
TE-FV109-23 151 Freézéryalve 109 éentervtemperature’ ‘ 478°F'
TE-FV109-38 152 TFreeze valve 109 shoulder temperature 108°F
TE-FV110-1B 153 Freeze valve 110 shoulder temperature 106°F
TE-FV110-2B 156 Freeze valve 110 center temperature 77°F
TE~FV110-3B 155 Freeze valve 110 shoulder temperafure 88°F
TE-FV111-1B 156 Freeze valve 111 shoulder temperature 88°F
TE-FV111-2B 157 Freeze vfilve 111 center temperature 80°F
TE-FV111-3B 158 Freeze valve 111 sfibfilder temperature 88°F
TE-FV112-1B | 160 Freeze valve 112 shoulder temperature 89°F
TE-FV112-2B 161 Freeze valve 112 center temperature 83°F
TE-FV112-3B 162 Freeze valve 112 shoulder temperature 89°F
TE-FV204-2B 17 Freeze valve 204 center temperature 254°F
TE-FV206-2B 43 Freeze valve 206 center temperature 303°F
TE-HB-1 176 High bay ambient temperature 83°F
TE-HX-2A 34 HX fuel outlet nozzle temperature 1170°F
TE-HX-3A 67 HX fuel inlet nozzle temperature 1207°F
TE-HX-9A 91 HX shell center temperature 1185°F
TE-MB1-1 134 Main blower No. 1 bearing temperature 86°F
TE-MB1-2 133 Main blower No. 1 bearing temperature 68°F
TE-MB3-1 132 Main blower No. 3,bearin§ temperature '93°F
TE~MB3-2 131 Main blower No. 3 bearing temperature 68°F
TE-NIP-2 345 Nuclear instrument penetrétion temperaffire 139°F
TE-OFT~-6B 105 Overflow tank bottom temperature 1172°F
TE-PT2YM 183 203°F
 

 

 

-~ ——

85

. Teble 4.23 (continued)

 

 

TE-SER-1

178

Speciel

equipment room ambient temperature

_Identification g::? b Description ““-" Reading
o . No. ] R -.10/12/69
TE—PT-ZEMV 165 Perticie trap-temperature 167°F °
| TE-PT-2FF 170 | Particle trap temperature 94°F
TE-R-2 | ~ 26 Reactor top:tempereture 1189°F
TE-R-4A 73  Reactor toputempereture '1209°F~
.TE—RrISA" 78  Reactor side tempereture - liBi’F
TE-R-18A | 79 Reactor side temperaturei 1196°F B
:_TE-R:ZQA; 80 - Reactor side temperatureTb 1182°F o
| IE&E525§" 82 Reactor side temperature .i,. 1178° ‘
.TE—REZQA, | 55 Reactor bottom temperature, 1182f
TEferZAp_' | 83 Reactor“bottom tempeneture 1182°F
TE;RrZBA - 84 | Reactor'bdttomtemperature' 1181°F . |
_'TE-REZQA 56 ReactorrbOttem temperature 1180°f .
TE-R-30A 85  Reactor bottom temperature 1181°F
eTE-RrBlA,l 59 Reactor Bottom temperature ’ iiSOfF
| TE—RrézAn 77  Reactor graphite tube temperature - é31fF’ "
_TE;R,45Afp 76 Reactor neck upper temperature | 282°F‘._
TE-R-49 8 Reactor top temperature:, | 1185?Fp .
.TE—R950'_'. 10 Reactorftop temperature | 11§1°F'Afl"
TE-R-51 '_ 12 Reactor top temperature 1185°F
_TE-RC§i£ o 290 Reaetor_cell embient temperature” '139°F
TE-RC-2 291 Reector?eeiiiembient tempereture. 143°F
TE-RC&Bp . 292 Reactor ¢éii afibient temperaturep 1495F .
-TEfRC+4“:;' 294 Reectur cell ambient temperature:_-- _ 135°E
| TE-RC-5 295 Reactor cell ambient temperature  138°F
| TE;RC§6_;E | 296 Reactor cell ambient témperaturerw | - 127°F
TE;RC#7p._ | 297 Reactor cell ambient temperature::,b-'":‘ | 1§2°F |
 TE-RC-8 - 298 V'Reeeter cell ambient temperature‘ft- ~ 149°F
VTE-RC*Q | 299 . Reactor cell ambient temperaturebflf_ 139°F
TE-RC-10 301 Reector cell ambient temperatureii:. 153°F
101°F

 
 

86

Table 4.23 (continued)

 

 

Identification g::? | Description . : . Reading.
_ . " No. , ' 10/12/69
TE-TRM-1 - ‘177 Transmitter room ambient températureé | ~ 81°F
TE-VH-1 180 Vent house’gfib{ent temperature - 76°F
TE-VI-1 143 Vapor tank .vi’ai.ter“ temperature | . 63°F
TE-VI-2 | 289 Vapor tank aif temperatfiré _‘. 66°F
TE-100-A1 5 Line 100 temperature | | 1207°F
TE-100-A2 25 Line 100 temperature 1207°F
TE-100-A3 46 Line 100 temperature 1208°F
TE-100-1A 68 Line 100 temperature 1209°F
TE-100-3A 70 Line 100 temperature  1208°F
TE~101-2A 60 Line 101 temperature 1210°F
TE-102-A4A 72 Line 102 temperature 1167°F
TE-102-1A 71  Line 102 temperature 1165°F
TE-102-5D 6 Line 102 temperature 1167°F
TE-200-C7A 122 Line 200 temperature 1014°F
TE-200-20A 64 Line 200 temperature 1022°F
TE-201-A1B 22 Line 201 temperature 1068°F
TE-201-A1C 20 Line 201 temperature 1068°F
TE-201-A2B 18 Line 202 temperature iOll°F |
TE-201-A2C 16 Line 202 temperature io10°F
TE-201-B11B 123 Line 201 temperature 1068°F
TE-201~1B 63 Line 201 temperature lo72°r
TE~202-A1 7 Line 202 (well) temperature 997°F
TE-202-B1 27 Line 202 (well) temperature 1007°F
TE-202-D1 48 Line 202 (well) temperature 1007°F
TE-522-1 135 Line 522 témggrature 84°F
TE-524-1 136 LineA524 température : 101°F.
TE-556-1A - 201 Line 556 temperature 94°F
TE-702-1B 192 Line 702 temperature 132°F
TE-705-1A 193 Line 705 temperature 142°F
TE-707-1A 194 Line 707 tempefature: 140°F
TE-752-1B 195-' tine 752-temperaturé' | 123°F!
 

87

» _Tab1g_4,23‘ (continued)

 

~

 

Cdmpensated ion chamber No. 1 position

 

Identification gg:? Description L Rgadingg
ent2 Mol ‘ 10/12/69_

TE-755-1A 196 Line 755 temperature 126°F -
TE-575-1A 197 Line 575 temperature 127°F
TE-791-1 140 Line 791 temperature . - 102°F
TE-795-1 141 Line 795 t¢épé;éturé 146°F
TE-804-1 215 Line 804 temperature 100°F

- TE-805-1 216 Line 805 temperature 104°F
TE-811-1 211 Line 811 temperature 81°F
TE-813-1 212 Line 813 temperature 79°F
TE-826-1 202 Line 826 temperature 99°F
TE-831-1 206 = Line 831 temperature 102°F
TE-833-1 213 Line 833 temperature 99°F
TE-837-1 203 Line 837'temperature‘ 102°F
TE-841—1 205 Line 841 femperature 106°F
TE-845-1 207 Line 845 temperature 124°F
TE-846-1 204 Line 846 temperature 107°F
TE-851-1 210 Line 851 temperature 79°F

| TE-874-1 208 Line 874 temperature 118°F
TE-876-1 214  Line 876 temperature 106°F
TE-916 200 Line 916 temperature 335°F
TE-917 198 Line 917 temperature o 126°F
TE-922° 199 Line 922 temperature | 118°F

 WM-CDT 246 Coolant drain tank welght 152 1bs
_WM&FDL 243 Fuelrdrain:tank:No;'1_Weight | 0 1bs
WM-FD2 244  Fuel drain tank No. 2 weight 468 1bs
WM-FFT 245 Fuel flush tank weight 8800 1bs
wMEEST 8 247 ;Fué1 étor€ge.tank.wéight 0 1bs
XPM-201 268 Reactor power. T.6Mi
ZM-FC1 257  Fission chamber No. 1 position 61 in.
ZMEFCZ_ 258 - Fission chamber No. 2 position 66 in,
ZM-NCR1 19 36 4n.
 

 

88

Tdbié?4.23 \(continued)

 

 

Radiator outlet door position

85 in.

Idenfificatipn g::? Description '-'Réaéing

No. C : 10/12/69
'ZM%NCRé 21 Compensated ion chamber No. 2 position éhhfh.‘
ZM-NCR3 23 Cofipensated ion chamber No. 3 position - 44 in,  —
ZT-1ID 37 Radiator inlet door position 89 ih.
ZT-0D . 38
 

 

89

5. FUEL SYSTEM

J. L. Crowley  C. H. Gabbard
R. H. Guymon J. K. Franzreb

5.1 - Descfiption

 

The fuel—c1rculat1ng loop consisted of a graphlte—moderated reactor,
"a centrifugal type fuel pump with an overflow tank, and a shell-and-tube
"heat exchanger, all connected'by 5- 1nch Hastelloy-N piping. The normal
'hbperating temperature was about 1200°F and fuel or flush saltrwas'circulated
at 1200 gpm. When the reactor was not:in\operation, the selt was drained
to one or both of the fuel drain tanks or to the fuel flush tank. Inter-
connecting'salt piping and‘freeze valves permit filling the reactor or
transferrlng salt between tanks by manipulatlng valves in the hellum sup-
ply and vent lines.

5.2 Purginngbisture and'Oxygen from the SyStem

In the fall of 196k, before salt was charged into the drain tanks,

oxygen and moisture were purged from the fuel c1rculat1ng system and the

© drain tanks. ThlS was accompllshed'by pressurizing the system with helium

to assure that it was lesk-tight followed by a combination of evacuation
and purglng with dry hellum“before and during the heatup. Details of the

. heatup are. glven in Chapter 17. The hellum'was 1ntroduced at the fuel

- pump whlch was in operatlon to provide clrculatlon in the loop. The system
wes vented or evacuated . at the normal fuel pump offgas line (518) or at the
' salt transfer line (llne 110) to the fuel processing system. The latter
.provided a 1onger flow path and thus a more effectlve purge.

- : Flgure 5-1 gives the sequence of 0perat10ns used. Pe&ks of moisture
_;in the effluent gas were observed at sbout 250°F and again at about. 650°F.

The system'was evacuated-after,eacn gf;these_mo;sture pesks.'.Further heat-

. ing tb:1130°F'didqn6t'eause”eny'Significent addf%ional moisture releases.

 

rLater -analyses of salt samples 1nd1cated that the purging had been very
effective.
H20 {ppm}

FLOW (liters/min)

'TEMPERATURE (°F)

1000

800

QFF-GAS MOISTURE

  

 

ORNL - DWG 65-4469

      
 
 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

600
400
® READINGS AT LINE 110
A READINGS AT LINE 510
200
OUT OF SERVICE
0
- o EVAC. TO 18-in, Hg
- 88| E e Ty Dpeia EVAC, TO Yo
- : 1 o} (FFT AND VAC. .
z?l’T‘l REACTOR * Spsig W" S psig % Spsig ————‘l
D! 25inHg OPENED) Hg |
10 .
HELIUM INFLOW
5t ]
" - ' - 1 5
o T 11, ! L
1250
1000
750 REACTOR VESSEL TEMPERATURE
500
250
0
20 22 24 2 28 30 1 .3 5 7 9 T 13 5
OCT 1964 NOV 1964 ?

Fig. 5.1 Initial Removal of Moisture from Fuel System

06

 

 
 

“ times were deemed acceptable.

91

5.3 Fuel—Circulating-System Volume Calibration

After the flush salt had been added to the drain tanks and transfers
made to fill the freeze valves (see Section 5.11), the fuel circulating
system was filled and operated for 8 days. It was then drained and filled

several times to check the drain times and calibrate the -system. The cali-

rbration was done by increasing the drain tank pressure in increments and

recording the drain tank weight and the differential pressure: between the

~drain tank and the fuel pump.r This is plotted in Fig. 5-2. The system

was purposely_overfilléd-to determine the position of the fuel-pump over-

~ flow line and to test the Overf16w-tank level indicators.. Overflow oc-

curred at readings,on,the twoéfuel—pnmn bubblers of 89.5 and 91%Z. A total

-0of 105 1b of salt. transferred;to the oVerflow tank prodnced a.reading on .
the level instrument of 11, 5% or 4.2 inches of salt in acceptable agree-

‘ment with the calculated response.

i:5.43 Drain Times

The temperature'distributiOn in the drain valve (FV-103) was con-

trolled so that it would thaw in-9 to 11 minutes after an emergency drain

' signal., (See Chapter 20 ) The time required thereafter for the salt to

drain from the loop- depended on which valves were open in the salt and gas

~ lines. ‘Normally the freeze valves ‘to both drain tanks were kept thawed

but for a while after a salt fill one valve would still be frozen. An

1_emergency drain signal_acted;to,thaw the freeze valves and to open valves

in drain tank vent lines and'in'the equalizer lines between'the gas in the
'fuel pump and in the drain tanks but it was considered possible that one

or more of these valves. might not open. Tests were conducted, therefore,

| hto measure ‘drain- times for various conceivable combinations of valving.

On a normal emergency drain the loop.drained completely in 9~11 min.

- after FV-103 thawed, With the equalizer valves open but one freeze valve
':7kept'frozen, the drain time. wasraboutr30 minutes, regardless of whether or
,'not the vent valves were open. With the equalizers closed, one freeze

~ valve frozen and the vents open, the drain time was 41 min. These drain

 
 

 

AP BETWEEN DRAIN TANK AND PUMP BOWL (PSI)

92

ORNL-DWG 73-597 -

 

 

 

 

 

 

 

 

- | O

 

 

 

 

 

 

 

 

 

 

 

24
22
—SALT IN BOTTOM
20 J1__ OF PuMP BOWL
18 /
16
/d—SALT AT TOP OF REACTOR
14 i
12 /
10 /
8
Z—— SALT AT BOTTOM OF REACTOR
6 | ] |
0 2000 4000 6000 8000 10,000
POUNDS OF SALT TRANSFERRED
Fig. 5.2 Fuel Circulating System Calibration
 

 

 

93

5,5 Mixing of Fuel and Flush Salts

During 235y fuel‘operetion, the uranium éoncentration of the flush
selt increased an average of about 215 ppm each time it was used aftér
fuel salt operation; 'Duringfz33U'0peration, with:a'loWer uranium concen-
tration in the fuel, theVécrresponding increase should have been only

- 39 Pbm. The actual 1ncreases ‘during the three flushes after 233U fuel cir-

culation were 36, 42, and 39 ppm. Using these figures, approximately
L0 1bs of fuel salt mixed with the flush selt during each flushing opera-
tion. (For more details on salt analysis and 1nterpretat10n of results,

see. Reference 23.)
5.6 - Primary System Leak

During operatien‘the-eexi-air activity was continuously monitored
to detect any leaks from_thetprimarysystem. ‘No leaks were detected until
after the final fuel salt drain in December, 1969. At this time the cell-
air activity didlincrease which'indicated a leak. Subsequent tests showed<
that the lesk was at or near one of the drain-tank freeze valves (FV-105).

The activity was mostly xenon W1th some iodine, krypton, and noble metals.

 

'Four days after the first release there was less than 25 curies of xenon
. and less than 50 mlllicuries of 1od1ne in the cell. This was released to

the atmosphere W1thout exceedlng the release rate permltted by the MSRE
,_safety limits.? o R S o , B

- The prdbabillty of the leak resulting from corrosion seems. remote. A
;,rrev1ew of the operatlon of the freeze valve does not. 1nd1cate any excessive
'.jthermal stresses.- No abnormalities were found upon rev1ew1ng the . constructlon

xhrays and other 1nspection reports. It was noted, however, that the weld
_*between the air shroud and the salt plplng at the freeze valve was not
- fspeclfled as a full penetratlon weld This led to the suspicion-that the

| - leak may be a crack that started at this p01nt and was propagated.by stress
""" cycllng. Determination of the exact locatlon and nature of the leak will
be attempted during the post-operatlon examlnatlons., More informatlon

~on the prelimlnary evaluatlon of the leek is given in Reférence 25.
 

 

S

94

5.T Operation

The operation of the fuel system was satisfactory. Difficulties en-
countered with the components.are described in the following sections. The
rate of transfer of salt to the overflow tank was higher than expected.
Details on this and the loop void fraction and xenon poisoning are given in

Reference 18.
5.8 Fuel Pump and Overflow Tank

5.8.1 Description

The fuel salt circulating pump was a centrifugael sump-type pump with
an overhung impeller, developed at ORNL expressly for circulating molten
fluoride salts of the type used st the MSRE.26 (See Fig. 5-3 and 5-4.)
At the normal operating speed of 1189 rpm, it had an output of about 1250
gpfi at a 48,5-ft salt head. About 50 gpm of the pump output was circulated
internally to the pump bowl vie & spray ring to promote the relesse of
entrained or dissolved gaseoué fission products. The gas space in
the pump bowl was purged with helium to sweep these to the offgas disposal
system. The helium was introduced Jjust below. the shaft seal in the béaring
housing. Most of the gas flowed downward through the lasbyrinth between the
shaft and the shield block to prevent radioactive gas and salt mist from
reaching the seal. The remainder flowed upward to prevent any 0il which
leeked through the seal from getting into the pump bowl. 0il for lubri-
cating‘the bearings and cooling the shield plug was recirculated by an
external pumping system. Helium bubbler type instruments were used to

measure the liquid level as a means of determining the inventory of salt

in the fuel system. Smell capsules were lowered into the bowl to teke

safiples for analysis or to add fuel salt.

The meximum height of the liquid in the pump bowl was limited by the top
of an overflow line (5-1/2 in. ebove the center line of the volute) which
connected to aS:Sfft3 overflow tank located beneath the pump. Helium bub-
bler type instruments'were also used for measuring the liquid level in the
overflGW'tank ' Slnce the overflow line extended to the bottom of the over-
flow tank c1051ng & valve in the overflow tank vent line forced the salt
back to the fuel pump.

(fi;
 

 

 

 

N

 

‘_/\\

 

 

 

 

 

 

J@\

 

S

 

 

 

s

7
IO
220220

 

 

 

 

 

 

 

 

_

s et

 

 

SAMPLER ENRICHER

LEVEL INDICATOR

SHAFT
COUPL!

SHAFT SEAL
{ See Inset)

LEAK

LUBE OIL
BALL BEARINGS

{Face to Face)

{Bock to Back)
L_UBE OIL OUT

SEAL 0IL LEAKAGE
DRAIN

LEAK DETECTOR_
{Out of Section)
{See Inset)

BUBBLE TYPE

OPERATING
LEVEL

To Overflow' Tank

Fig. 5.3

BALL BEARINGS ~,

   
 
 
 
 
 
 
  
 
      
  
 
 
 
  
  
  
 
 
    
 
  
   
   
   
 
 

ORNL-LR-DWG-58043-B

 

 

 

 

 

 

 

 

BE OIL BREATHER

 

 

 

NG HOUSING

- GAS PURGE IN

- SHAFT SEAL (See Inset)

SHIELD COOLANT PASSAGES
(in Porallel With Lube 0il)

SHIELD PLUG

GAS PURGE OUT (See Inset)
GAS FILLED EXPANSION
 SPACE

STRIPPER
(Spray Ring)

SPRAY

 

MSRE Fuel Pump

 

c6

 
ORNL-DWG 69- 10172

 

96

  

  

            

 

 

 

 

              

 

  

SUCTION

m

     

[

 

 

SAMPLE
CAPSULE
CAGE

 

Cross Section'ofiMSRE Fuel Pump-Shbfiing Flow‘Paths;

Fig, 5.4
 

lation no signlflcant dlffieultles were encountered durlng hellum circula-

97

5.8.2 EBarly Operationrwrd

The fuel pump had been loop—tested with molten salt for 100 hrs at
1200°F before it was installed at the MSRE in October, 1964, After instal-

tion while the fuel system.was belng purged, during early flush salt opera-
tion which started on January 12, 1965, or during the criticality experiments.

 

A continuous but very slow, ‘accumulation of salt in the overflow tank was |

observed throughout the early»operatlon. (See 5.8.9.)
5.8. 3 Examination of Fuel Pump after Cr1t1ca11ty Exper1ment27

The fuel-pump rotary element was removed for inspectlon in September 1965 |

~...at the end of Run 3.  The pump ‘had been used for circulating helium for 1410

 

hours and it had'been fllled:with salt at the MSRE for 2120 hours, 1895 hours
of which the salt was belng clrculated _

- The pump was generally in. good condition ‘and appeared ready to be used
for full-power operation. The only dimen31ona1 change was a 0.006~in. grovwth
of the pump tank bore diameter where the upper O-rlng mated with the pump
tank.

The most 51gnif1cant dlscovery was ev1dence of 8 small 011 leak through
_the gasketed Joint at the catch basin for the lower Oll seal. This oil had

o run down the surface of the. shield plug, where it had become coked by the

“higher temperatures near the bottom. Some of the oil had reached the upper
O—ring'grooye at thepbottom of ‘the shield plug and,had'become coked_ln the

| v:groove5?but-none*appeared}tofhawepleaked past the O-ring during high-

' temperature operation;'3Sofie7fresh>oil was observed below the ring after

“ the rotary element had been moved to the decontamlnatlon cell, but it was

_believed that thls oil drlpped from the open llnes durlng the transfer to
'gfthe cell.r- ’ — Lo ‘

 

All durlng subsequent operatlons, considerable dlfflculty‘was en-

mfi“f&ountered from plugging of the mein offgas system, due largely to decom-

- position products of ‘0il that had leaked into the pump bowl. (See
" Chepter 8.) Because of this trouble -the gasketed Joint was seal-welded
‘on the spare rotary element._ However, the spare never had to be 1nstalled

at the MSRE.

 
 

 

I

 

A layer of flush salt about 3/8-inch deep containing about 40 in.3
was trapped and frozen on top of the labyrinth flange. Apparently the
salt had drained through the 1/8-inch diameter-holes_ih the flange'ufitil
surface tension effects balanced the hydrostatic head. This. layer can be

seen in Flg. 5-5, which is a photograph taken during the 1nspect10n. This

photograph also shows the contrast between the surfaces exposed to the
salt, which 'were bright but not corroded, and those above the salt, which
were distinctly darker. There was a coating of fine sait mist paerticles

on the lower face of the shield plug and in an "O"-ring groove around the
shield plug there were small amounts of flush and fuel salt that must have -
been transported as mist.

The pump was reinstalled using remote maintenance techniquee so that
these techniques could be evaluated. Four universal joints on the flenge
bolts that had been fofind'broken during the disassembly were repaired prior
Fo the.reinstallation of the rotary element. These failures resulted from
excessive bolt torqfie that had been used earlier to obtain an initial seal
on the flenge. (It turned out that the jack screws had not been fully

_backed off before the flange bolts were tightened.)

S,B;h " Pump and Pipe Support Problem

~ A problem related to the overall fuel-pump installation became evident
during the post-criticality shutdown in the fall of 1965. The fuel pump

could move in the horizontal plane, but was fixed against vertical movement.

The heat exchanger could move‘pprizontally; the heat exchanger support frame
was fixed against vertical movement at the north end, but was mounted on
spring supports at the south end. The south end of the heat exehanger,wes
coupled to the fuel-pump bowl by & short length of S-in. Sched.-40 pipe.
The heat exchanger was supported from below, and the pump bowl was supported
frofi above. The connecting piping was supposed to move upward at the heat
exchanger and downward at the pump when the system was heated. |

Because of the physical arrangement of the piping and equipment, stress
ranges in the piping due to thermal cycling were calculated to‘reech a;

maximum of 20,000 psi, which is acceptable. Uncertainties eiisted in

- critical parts of the heat exchanger, however, particfilarly in the nozzle
0
O

 

Fig. 5.5 Fuel Pump Rotary Element

 

 
 

100

where estimates were as high as 125,000 psi during cyecling from 150 and

1300°F. The end of the exchanger toward the pump bowl was therefore

mounted on springs. This should have reduced stresses in the nozzle to the

range of 20,000 to TO, 000 psi.
~ Careful observations during a. heating cycle to 1200°F showed that the
equipment did not move as expected, however. Because of the complex equip-
ment eonfiguretion and the inevitable uncertainty of the celculations, it
was decided to make strain gage measurements with the equipment cold and
moting the piping by mechanical meeans for measured distances with measur-
gble loaeds. The highest stresses were fbun& to be in the'filletrwhere'the
nozzle was welded to the hesat exehanger head when & spring force of 2000
poundS'was exerted to raise the end of the heat exchanger 3/16 in. The
measured stress in the fillet was 13,000 psi and was a factor of sbout
5 greater than the stress in the nearby piping. Dye—check'oftthe'nozzle
to the head of the heat exchanger showed no indication of cracks. |

The conclusion from these tests’was that the mounting arrangement was
adequate to allow the system to go through more then the 50.thermai cycles
required without a fatigue faiiure. The system was then put to use for

power operation.28

5.8.5 Effect of Bubbler Flow Rates on Indicated Fuel-Pump Level

The fuel-pump level was determined by bubbling helium.through.the salt
and measuring the differential preSSure between this line and & reference
line which connected to the gas space of the pump (see Fig. 5-6). The end
of one of the bubbler dip tubes (596) was 1.87hk in. lower than the other
(593). A common reference line {592) was used for the two bubblers. The
level readout instrument had a full scale (0-100%) range of 10 inches of
selt. The centerline of the volute'was_at 35%. 'Compensation'was_previded
in the instrumentation for chenges of density between flush end fuel salt
and for differences in the lengths of the bubbler tubes. L

Since the d/p cells used to indicate level were located outside the
reaetor,cell,_there ves some pressure drop in the lines Between these and
the'pump.‘ The smount of pressure drop was dependent upon flows through the
lines. Tests were made early in 1965 to estsblish the reletionship between

these flow rates and the indicated levels.
 

 

 

101

ORNL-DWG 70-5193

- 593

 

 

 

 

 

 

™1.636in.
R | { —1.874 in.
---—-—VOLUTE ¢ — o ,

L .
"’/////////77/’///// 7y /- ///W 4 o

—DENSITY ZONE

 

 

 

 

 

 

Fig. 5.6 Schematic Representation of Fuel Pump Bowl Level and
' Density Indicators ' : L

 
 

 

 

- 102

- Data for the upper probe are,presented'in Fig. 5-T. The curves for
the lower probe were similar. It can be seen that the indicated level de-
creased*with increasing pump-bowl refereace leg flow and that the indicated
level increased with increased flow into the dip tubes. ' The actual salt
level was not changed during the tests. Normal operating conditions were
set at 25-psig forepressure on all three of the bubbler flow elements,

5 psig in the fuel—pump cover gas. '

5.8.6 Fuel-Pump Level Changes and Limitations

Based on the recommendetions of the pump .development group at ORNL, the
level alarm and interlocks were originally set as shown in Teble 5-1.

The differences in the level at'which the fuel pump could be started
(64%) end the level at which the pump would‘automatieally stop (55%) was
necessary because the 1nd1cated level decreased 10 to 12% 1mmed1ately upon
starting the pump. This was due to fllling the spray rlng and fountaan flow
chamber, :

The narrow dlfferences between the high and low’ alarm and control set-
points caused considerable operational difficulties. Prlor to starting the

pump after a fill ‘the average loop temperature could not be sccurately
determined. Since the fuel~pump level changed about 12 L% per 100°F change
in loop temperature, the selected fill point was not always satisfactory for
operation. In which case the freeze valve had to be thawed and the system

level adjusted. During operation the reactor outlet temperature was usually

~ held constant when the power was changed. Therefore, the bulk average tem-

perature changed with power changes and this caused changeS'in_the fuel-pump
level. In addition to this, experiments were run which required operation
at different reactor outlet temperatures. During load and rod scrams the
system cooled rapidly. ©Sometimes it was necessary to reheat the system
before restarting the pump. The eperating levels were further restricted
when it appeared that the offgas plugged more rapidly when the salt level

. was gbove 60 to 65% and gas entrainment in the fuel loop seemed to increase

below sgbout 50%. The problem was further compounded by the changing pump-
bowl level due to salt transfer to the overflow tank.l8
 

 

LR-593C (%) S—36 IN POSITION 3

70

- 69

103

ORNL-DWG 73-598

 

74

 

73

 

 

 

72

 

 

7

 

 

 

 

 

 

 

 

7,

0T A"
=
67 | - ‘

L
66‘.r___,.——"' |

 

 

 

 

 

 

 

 

 

 

 

65 : '
16 18 20 - 22 - 24 26 28 30

FI-593 (psig)

NOTES: ,

FI-592 and FI-593 flow rates were proportional to the difference between the
pressure upstream of the flow restrictors and the pressure in the fuel pump.
The fuel pump pressure was held constant at 5 psig during these tests.

- LR-593C was read from the bdttom of the inked space which was about 1.5% in

width. .

Fig, 5.7 Effect of Bubbler Flow on Fuel Pump Level Indicators

 

 
.ok

' 4T5%

104

- Table 5-1.. Original FP Level Interlocks

 

Level

; +78%

- 464%

T ¥55%

' 4539

Action and Reeson for Interlock

~Stops fill, gives a temperature setback.and rod reverse to

prevent overfilling of the fuel pump (overflow point is
about 90%). .

Annunclatlon

Pump cannot be started below thls level. This is to
prevent cavitation. ‘

Annunciation

Pump will stop below this level. Again, this is to
prevent cavitation. | .

 
 

105

Tests were therefore performed whereby the interlocks were bypassed and
the pump was started and operated at lower levels. Since there appeared to
be no caVitation'or;adverse”effeets_on the pump, the interlocks given in
Teble 5-1 were changed to 8%, T5%, 55%, 40%, and 38% late in 1965.

Normal level during operation was still maintained between 50% .and 60% due
to above,considerations._ HoWever,frecovery after a load and rod scram was

much easier.

5.8.7 Coolant Air £o Fuel Pump |

Tn the des1gn of the MSRE it was calculated that the upper portion of
the fuel-pump tank wouldrbejspbjeot to substantial heating from fission
.. products in the gas spaoe-above-the salt. Since the useful life of the
pump tank would be 1imited;BYTthermal—stress considerations at the junction
of the volute_stpport cylinder,with the spherical top of the tank, close
control was at‘first maintained over the temperatures in this region.
Design studies had indicatedethat the maximum lifetime would result if
the junction temperature were'kept about 100°F below the temperature on
the tank surface\6-in. out from the junction. Component cooling "air"
_(95% Nz) was prOV1ded to. malntain this temperature distribution. A secondary
~consideration in controlling the temperatures was a desire to keep as
Vmuoh of the pump tank as p0551b1e above the liquidus temperature of the
salt. o : e o =
In operating the reactor, it would.be 1dea1 if a fixed flow rate of
~air over the pump tank would.provide a. satisfactory temperature distribution

= for all conditions.' Early de51gn calculatlons 1nd1cated that this condition

 

;could'be met with an air flow of 200 cfm.29 ‘However, temperature measure-
R fments on the pump—test 1oop and during the 1nitial heatup of the MSRE indi-
'fircated that only ebout 50. cfm'would.be required, and that the air would have

',,:to be turned off when the pump ‘tank was empty.

To minimize the temperature effects when the coollng air'was turned on,

. air flow during power operation‘was ‘to have been set at the minimm that

'pjwould give the de31red temperatures. It was found thet an air flow of 30 cfm

.:fgave 8 satlsfactory temperature distribution at all power levels. In_order

to achieve and control;thisxrelatively low flow rate, a new valve, having

 
 

 

106

& lower C, had to be substituted for the original valve. Figure 5-8 |
shows the temperatures in the two regions of interest as a function of re-
actor power level with that air flow. The variations in the individual
temperatureS'were caused'by varietions in the pump—bowl level salt temperar
ture and air flow. Both the 1ndlv1dual temperatures and the temperature
differences increased linearly with power, as expected. It was felt during

early MSRE operation that, although temperature differences would exceed

- 100°F at full power, the reactor could be so operated with the 30-cfm air

fIOW'w1thout significantly reducing the 1life of the pump tank.

When the reactor was started up for Run 8 in September 1966, an unex-
plained'shiff-downward in these temperatures was noted. 'Latef the cooling
air to the pump tenk was turned off during the attempts to melt out the salt
Plug in the 522 line, and although the temperatures on the pumpetank surface
were higher than with the cooling air, the temperature gradient was less.
Since the temperature distribution was as good or better than with the air

cooling, the use of air cooling was discontinued.

5.8.8 Salt Transfer to the Overflow Tank

Early operation of the reactor showed that by some unexplained mecha~
nism, salt gradually accumulated in the fuel pump overflow tank even when
the salt level in the pump bowl was well below the overflow point. The
transfer rate depended on salt level, and the transfer ceased when the
level was ebout 3 in. below the overflow point. This situation existed
until sbout April 1966, when transfer began to be observed at lower salt
levels. The rate appeared to incresse gradually as time went on until
it ieveled off in June and July at 0.57 1b of salt per hour, independent
of salt level as far down'ae 4,7 in. below the overflow point. The change

occurred at the time of the stepwise increase in power, but no mechanism

- connecting the two has been identified. This rate of transfer continued

through the 235y operatlon and requlred emptylng the overflow tank about
three times per week, _
Prior to operation with 233U fuel salt, the flush salt which had been
processed to remove the 235y was circulated for 42 hours. During this and
the first 16 hours of fuel carrier salt circulastion, the transfer rate and Qfi;
ORNL-DWG 66-14441

 

1200

 

 

1100 |-

 

 

 

 

 

 

 

1000

 

' TEMPERATURE (°F)

900

 

 

 

1

) i
| A
' A
a
A o
) ®
& * a

.

@

®

 

 

 

 

 

 

 

 

- Q0
o
O

| C)'——oio——-l

Fig. 5.8

2 3 4 5 6 7T 8

POWER (Mw)

" Variation of Fuel-Pump Tank Temperafures with Reactor
Power. Cooling-air flow, 30 cfm.

LOT
 

 

108

loop void fraction appeared to be unchanged from the previous periods. Two
hours after the start of a 12-hour exposure of & beryllium rod in the pump
bowl, the bubble fraction in the system increased from the normal 0.1 vol %
to about 0.6 vol % and.the rate of transfer to the overflow tank increased
from sbout 0.4 1b/hr to greater than 4 1b/hr. ,

During the remsinder of the 233y operation the transfer rate was high
(up to T2 1b/hr) and varisble. The. results of the investigation of the
overflow rate, the changlng bubble fraction, and subsequent power pertur-

bations are given in Reference 18

5.8.9  Burps of'the Overflow Tank

During early opersation of the MSRE, when the bvefflow_rate'was-very low

and the need to push salt from theroverfldw tank back to the fuel-pump bowl
was infrequent, the practice was to empty the overflow tank completely. The
sudden pressurization of the fuel pump at the end of the burp'géve false
level indications and stopped the pump. Also when power operation was
started, it was noted that gaseoué fissionlproducts fiere being pushed out
the o0il seal line (524) by the sudden pressurization. Therefore, the
procedures were changed such that the overflow tank was not completely
emptied of salt during operation.

In Fébruary 1969, the main offgas line plugged to the polnt where it
presented a 4-psi pressure drop to the normal 3.2 liter/m offgas flow.
During & burp of the overflow tank, this plug blew out with the reactor at
full power resulting in a complete burp of the overflow tank. On &t least
' three other occasions when the overflow tank was being emptied, the offgas

plug blew out causing more salt to be transferred then plenned.

5.8.10 Varisble-Speed Drive for the Fuel Pump

- Prior to February 1969 the fuel pump was always operated.w1th the
normal 60-HZ power supply at -~ 1189 rpm. Then, in order to investigate the
effect of fuel circulation rate on system'béha&ior (Bubble ingestion, xenon
stripping, transfer to overflow tank, ‘blips),18 a varlable-speed motor-
generator set was brought to the MSRE to supply. pover to the fuel pump

during experiments. As described in Chap. 16, ¢onsiderable effort was
 

109

expended in modifying ahd repairing the M-G set to obtain satisfactory
reliebility. The pump itself, however, operated with no difficulty for
substantial per;ode at speeds:between 50% and_los% of normal.

- 5.8.11 Flooding of the Pum;g Bowl with Flush Salt

. At the end of Run T in. July 1966 the fuel 1oop was filled with flush
salt to rinse out residual pockets of fuel salt and thus reduce the radi-
atlonrlevels for the scheduled work in the reactor cell, As the salt
wes overflowing intorthe overflow tenk to rinse it out, the level in.the
flueh_tank was lowered too far_fihich exposed the bottom of the dip tube.

A large bubble of helium'gee:et—a preesure of ebout'30 pSig entered the

- bottom of the loop.  As the bubble rose, its volume increased due to the
'decreesed pressure and the fuel-pump level increased faster than it could
overflow to the overflow tank. Salttflooded-the reference bubbler, the
ennulus around the shaft, the”sempler tube, and'the'main offgas line.

Severel factors contributed to the accident. Interlocks normally'pre—

:vented filling to the overflow 1ine.. These had been bypassed to allow
flushlng of the overflow tank. The flush-salt 1nventory was marginal at

' normal operating temperatures (1200°F) The flush-tank temperatures were
“1ow (ebout 1140°F) which made the salt more dense and thus there was an’

’1nsuff1c1ent volume.
;5.8 12 Plugging-of the Offgas Line at the Fuel Pump -
Intermittent problemS'were encountered‘with plugglng of the 1/2-in.

: SChedrhO mein offges line et a p01nt Just beyond'where it left the top of
" the fuel-pump bowl. In-the early years of operatlon, thls either melted

.',1tself free when the reactor was brought to power, or 1t was periodically

;reamed out vhen the reactor was shut down, through the use of a mechanlcal,

 flexible rotating snake";5 In July 1969, a specially built heater assembly

",'consisting of two 1000~w formed calrods wes 1nstalled remotely around the

 

7' 522 offges 1ine between the top of the pump bowl gnd’ the gas coollng shroud"
: of -the fuel pump This, together Wlth beck—bIOW1ng'w1th hellum, was suc-
CESSfUI in clearing the line. Mbre details on the offgas plugglng prdblems

are given in Chepter 8.

 
 

 

 

110

5.8.13 Conclusions and Recommendsations

The fuel pump was used to circulate salt for 21,788 hours at tempera-
tures near 1200°F with no perceptible change in performance and no failure.
Some plugging of the offgas line was encountered. This should be con-~
sidered in future designs. Perhaps dual lines could be used with heaters to
melt out plugs if they develop. Leakage of oil (1 to 2 cc/day) into the
pump bowl contributed to the offgas plugging prdblem. This possibility was
eliminated in the replacement rotary element by seal-welding a gasketed
joint, but because the plugging problem was managesble, theISPare element
- was never installed. "

The transfer of salt td the overflow tank was an operating nuisance in
that it had to be periodically transferred back to the pump. The Mark-IT
pump (which has operated for over.13,000 hours in a test loop) has more
height in the pump bowl, eliminéting the need for an overflow tank.

5.9 Primary Heat Exchanger

The primary heat exchanger was a horizontal shell and U-tube tjpe.
During early power operation the heat transfer capability #as'foundlto be
considerably lower than expected. A reevaluation of the physical properties
‘showed that the thermal conductivity of both the fuel and coolant salts was
sufficiently below the value used in design to account for the overestimate
of the -overall coefficient of heat transfer. Table 5-2 shows a comparison
.of the physical property data used in the original design fo the current
data. The heat transfer coefflclents calculated by the conventional design
procedures using these two sets of data are also shown. o

The heat transfer did not ‘change during subsequent operation of the
reactor end no other d&fficulties occurred. The measured overall heat trans-
fer coefficients ranged from 646 to 675 with an average of 656 Btu/(hr-ft2 °F)
for 8 measurements. These measurements were made on the basis of nominal
full power at 8.0 Mv and a coolant salt flow of 850 gpm. If the actual
coolant flow rate were 770 gpm, which would be the flow consistent w1th 8
power level of T.34 Mw, the measured overall coefficient would be 59h as
compared to a calculated value of 599 Btu/(hr-ft2-°F). !The performance is

covered in detail in References 30 and 31.
 

 

 

111

Teble 5-2. Phy51cal Properties of Fuel and Coolant Salts
| Used in MSRE Heat Exchanger Des1gn and Evaluatlon

 

Original - o Current

 

" Fuel  Coolant Fuel wCoqlant

 

Thermal Conductivity, Btu/(hr-ft-°F) 2.75 3.5  0.832  0.659

Viscosity, 1b/(ft-br)  17.9  20.0  18.7 23.6
Density, b/ft3 . 15L.3 1200 1h1.2 - 123.1
 Specific Heat, Btu/(lb-°F) 0.6 0.57  0.4735  0.57T7
- Film Coefficient, _Btu/(hr-ft2-°F) 3523 5643 19T - 1989

Overall Coefficient,'Btu/(hr-ft2-°F) | 1186 - 618

 

 
 

 

 

 

112

5.10 Reactor Vessel and Reactor Access Nozzle

-5.10.1 General Description

 

 

| Reactor Vessel and Core — The reactor vessel was a 5-ft-diameter by
8-ft-high cylinder which contained a 55-in.-dismeter graphite core. A
cutaway drawing of the reactor vessel and core is shown in Figure 5-9.

The vessel design pressure and temperature were 50 psig and 1300°F with an
ellowable stress»of 2750 psi. _ _ , , ' ,

- The fuel salt entered the flow distributor where it was directed down-
wvard around the circumference of the vessel. It flowed in a spiral path
‘through & l-in. annulus between the vessel wall and-the core can for cooling
purposes. Anti-swirl vanes in the lower head of the vessel straightened the
flow path before it entered the graphite moderator core. Fiow“paSSageS5
formed by grooves in the sides of the graphite moderstor bars;'directed the
laminar salt flow to the upper head plenum. The salt flow then left the re-
actor vessel through the side outlet of the-reactor access nozzle which is
described in the following section. _ : L

During operation the fuel salt was held in the circulating system by
a freeze valve (FV-103) attached to the lower head of the reactor vessel,
The drain line and freeze valve was a 1-1/2 in. Sch.-40 pipe which was
_flattened for sbout 2 in. to give a flow cross section about 1/2-in. wide.
Cooling air in a shroud surrounding the flattened 'section maintained & frozen
salt plug for a "closed" valve. The 1-1/2 in. pipe extended 2-3/k in. into
the lower head of the vessel and was covered with a hood for protection
against sediments collecting on top of the frozen plug meking it inoperative.
The 1-1/2 in. hoodéd drain thus would remove all but a small puddle of salt
from the lower head even if there had begn heavy sedimentation. To provide
for drainage of the remaining puddle, a 1/2-in. tube was mounted through
the wall of the portion of the drain protruding inside the vessel and ex-
tended dowvnward through thé freeze valve below. This, in effect, formed
parallel freeze valves operated by the same controls.

The reéctor core consisted of 617 full and fractional size graphite |
elements 2 x 2 in. cross section and about 67 in. long. Salt flow channels

were formed by machined half channels on each of four faceé of each element.

o
 

 

 

 

 

 

 

 

 

 

113

ORNL-LR-DWG 6109TRIA

FLEXIBLE CONDUIT TO
CONTROL ROD DRIVES

 
   
   
   
 
  
 
 
 
 
   
  
     

GRAPHITE SAMPLE ACCESS PORT

S ACCESS PORT COOLING JACKETS
-~ .

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 -/ g | _ .
| N ~~— CORE WALL COOLING ANNULUS
REACTOR CORE CAN ~ 1. _ ‘

 REACTOR VESSEL —i:

 

 

 

ANTI-SWIRL VANES —

o < MODERATOR
'VESSEL DRAIN LINE

- SUPPORT GRID -

. Fig. 5.9  MSRE Reactor Vessel

 
 

 

114

When not buo&ed.by being immersed in salt, the vertical graphite elements
were supported by & lattice of graphite blocks which‘in turn are supported
by a Hastelloy-N grid fastened to the bottom of the core can. The ‘core can
was thus free to.expand dounward relative to its top attachment to the're—
actor vessel while the graphite was free to expand upward relative to 1ts
support at the lower end of the core can. The grephite moderator elements
were restrained from floating out of the core by a HastelloynN rod through
holes in the dowel section &t the bottom of each graphite element. In
addition a'W1re pessing through the upper graph1te elements prevented the
upper portlon of & broken element from floating out of the core.

To prevent possible overheating in an otherwise stagnant -region, & ‘
small portion of salt entering the resactor was_dlverted 1nto.the.reglon
Just above the core can support flange in the annulus}betueen the vessel and .
the .core can. . Vessel wall temperatures in this region and also the lower
head were monltored during operation for p0351b1e deviations.

, At the center. of the core were located sample specimens, three control
rod thimbles, and f1ve removable graphite elements. . The- sample specimens
were located in one corner of & U-in. square with. the three control rods
occupylng the other three corners. The five,removable elements then occupied.
the remaining spaces between these four,positions. ‘The controls rods are
described elsewhere in this report (Chapter 19). | ‘

The sampleispecimen assemblies mentioned sbove were made of various com-
binations and arrangements of graphite andrmetal.‘.They were exposed to much
the same salt velocity, temperature and nuclear flux. as the core matrix.
These are described in general elsewhere 1n thls section. ' )

The reactor vessel was supported from thettop removable cover of the
‘thermal shield by twelve hanger—rod.assemblies. These hanger-rod.assemblles
were pinned to lugs welded to the reactor vessel just above the.flos.dis—
tributor. Thetsupport arrangement was such that the reactor,ressel could be
considered\to be anchored at, the support lugs. The portion of the vessel
below the lugs was free to expand downward with no restraint.

Reactor Access Nozzle (RAN) — Attached to the upper head of the reactor
vessel vas a hofln._long 10-in.-diam nozzle. rThe nozzle had & 5-in.-diam '

{
side outlet for the leaving salt located about 10 in. above the reactor
 

115

vessel upper head. The remaining extension of‘the 10-in.-diam nozzle pro-
vided a pocket of trapped gas during the fllling of the reactor. However,
most of the volume of this exten51on was occupied by the nozzle plug which
is a removable support for the,three control rod thimbles, the 2-1/2 in.
graphite sample access pipe,iand'for]the discharge screen. - See Fig. 5-10.
Conventional_leak—deteeted metalyoval-ring—Joint3flanges were used on both
sample‘access'openings"to provide the necessary containment of the primary
system. The radia1 clearance between .the removablegnozsle plug and the -
nozzle was 1/8-in. at the top and 1/h-in. at the bottom to provide a tap-
ered annulus. It was 1ntended that -salt be frozen 1n this tapered annulus
providing. a salt seal to prevent molten salt from contacting the metal seal-
ing surfaces. However, maintaining a frozen salt seal was found to be not
possible. This is dlscussed;in more detail later. Cooling air was pro-
‘vided on both inside of the plugland outside of the lO—in.Vnozzle but only
to the inside of the plug of ‘the 2-1/2 in. graphlte sample access. Heaters

were provided to thaw any frozen salt remainlng in the 10-in. annulus after .

'a salt drain. For thaw1ng the 2-1/2 in. annulus the cooling tube was re-~
moved from the plug as part of the sampling procedure and temporarily re-
placed with a metal sheathed heater.

Some twenty thermocouples (not including spares) were 1nstalled at
'varlous locations on the RAN. to monitor temperatures of the nozzle, both
plugs and. control rod thimbles. A thermocouple well attached to the
, graphite—sample access plug extended into the. flowzng salt stream to indi-
- cate reactor outlet temperature. o o '

- ~ Strainers were provided et the reactor outlet to prevent passage of

~ graphite chips lerger then 1/16 in. A strainer basket vas attached to the

"”Wlower end of the 10-1n. nozzle‘plug and extended downward 1nto the upper

';-head reglon of the reactor vessel.@ “The three control rod thimbles and the

graphlte sample assembly passed through the strainer basket. Since the

 five graphlte elements were not pinned as were the remaining moderator ele-

'ments, a cross-shaped extension of the basket assembly projected beneath

n the basket to provide a hold—down ~ See Fig. 5-11.

‘The. core sample spec1mens “were removed and replaced only: when the re-

actor was drained of salt and partially cooled. The spec1mens were removed

 
116

ORNL -DWG 68-5507

- (coon.me AlR

 

 

 

19% in.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ST A

COOLING AIR

 

 

 

 

 

 

FUEL SALT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e L L L

 

 

10-in. ANNULYUS —

~ 2Y5-in. ANNULUS ——

Fig. 5.10 Reactor Access Nozzle Showing 10- and 2-1/2 in. Annuli
 

 

 

 

 

 

 

117

ORNL~ DWG 636475

PLUG -

   
       

FUEL OUTLET

ACCESS
NOZZLE

TOP HEAD OF REACTOR
VESSEL

       
  
  
 
 
 

STRAINER —
(16-goge PLATE WITH ¥-in. -

~diam. HOLES ON J4-~in.
CENTERS)

CONTROL ROD THIMBLE- GRAPHITE SAMPLES

2-in. GRAPHITE
- - :-MODERATOR

Fig.75.11 fR¢gctot,F¢e1 Outlet Strainer

 
 

118

through the smaller of the two access nozzles described earlier. A stain-
less steel standyipe was left attached to the small graphite-sample access
nozzle via a bellows. The upper end of the standpipe was bolted to a liner
set into the lower concrete roof plug. During normal operation the liner
opening was closed with a concrete plug. When semples were to be removed,
this plug was replacéd with a work shield which contained openings for
tools, lights, etc. All joints fiere moderately leak-tight and the stand~-
~pipe was provided with a nitrogen purge and offgas connections. A Roots
blower located in the service tunnel provided a lightly negative standpipe
pressure to assure inward leaskage. :

Core samples were takeh by removing the graphite sample access flange,
withdrawing the sample into the standpipe, placing the samplé in a special
carrier, inserting a replacement sample specimen into the core, and re-

placing the access flange.
5.10.2 Hest Treatment of Reactor Vessel

After the reactor. vessel was 1nstalled tests of the particular heat
of Hastelloy~N used in the vessel showed that the closure weld between the
top head and the flow distributor ring could have poor mechanical proper-
ties in the as-welded condition. Theréfbre in the fall of.1965, the‘reac-
tor vessel was heat-treated in place, using installed heaters, for 90 hours

at 1400°F.
5.10.3 Reactor Access Nozzle Freeze Tests and Effect of Circulating Bubbles

The main purpose for a frozen salt seal in the'annuli of the 10-in. and
2-1/2-in. diam. RAN plugs was to prevent contact of the sealing surfaces
with salt. The freeze flanges used as piping dlsconnects have a similar
function. The maein difference being thet the freeze flanges incorporated
a radial freeze joint while the RAN employed a longitudinal freeze joint.

During development on the Engineering Test Loop freeze Joint, it was
noted that a frozen salt seal could not be maintained relisbly if in con-
,fact with molten salt., It was intended to operate the MSRE RAN freeze
‘joint with & ring of frozen salt above the normal molten—salt level. This
would be accomplished by pressurizing the fuel system above normsl operat-

iné pressure soon after filling the reactor with salt, Cooling air on the
119

RAN would freeze & ring of salt in the annuli befbre the pressure was low-
ered to its normal value, leaV1ng a gas void between the frozen and liquid
salt. The gas trapped in the RAN annuli was thus the principal barrier
between the molten salt andjthe containment sealing flange. The ring of
frozen salt wouid then“serve'only as & backup protection against sudden
pressure increases fcrcing salt up into thelannuli. |

The first attempts to form this backup seal in this manner on the MSRE
RAN were unsuccessful due to 1nsuff1c1ent cooling air although the availa-
ble flow met the requirements of the design calculatlons.46 The flow was
permanently,increased from aboutj3'cfm:to’the§range of 15 to 20 cfm by
modifying the control-valveitrim. ‘Even with the increased air flow, there
was sufficient movement of salt in the annuli due to the turbulent flow
below to prevent forming a good frozen-salt seal, However, gas trapped
above. the ‘molten salt kept the liquid level well below the flange. The
‘liquid gas interface changed 1n height due to any change in volume of the
gas pockets., During early operation, the interface was at least a foot
‘below the flange and the flange temperature was about 200°F During later
,operations, there were periods when the salt level was higher and the tem-
perature of the flange approached 300°F.

There were two mechanisms which caused gas to be transferred to and
from this pocket. Gastasgtransferred from the pocket’aS“a_solute and
added by entrapment of circulating”bubbles from the salt stream below,

The equilibrium difference'between these two transfer rates determined the
" galt level in the annuli and thus the temperature readings of the wall.
- Any change in reactor operating condltions which disturbed the balance'
rd:between these two transfer mechanisms would thus change the height of salt
_ in the ennuli. It vas noted that the salt level in the RAN annuli would
‘be increased by any of the following changes 1n operating conditions
fhlowering the system pressure (cover gas in the fuel pump) increa51ng cir-

culating salt temperature or decrea51ng the amount ofjcirculating bubbles

o (i.e. decreasing pump speed).,; .

 

Examples of all three of these causes are shown in Figs. 5-12 and 5-=13.

Please refer to Fig. 5—10 fbr location of the thermocouples used in these

two graphs.

 
 

 

 

 

120

ORNL-DWG 73—599

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1400
EACTOR OUTLET - -
'-.---———-_ N
12m —— F — ___jPJ _—\ '\ . )
TE R—438
c
2. 1000
m .
S U TE R—7B
%
o0
|17}
a.
£ 800
P
600 —J
400
5 10
&
o
3 J
gollll | 1 1 | . | 1 1|

 

 

 

 

 

 

3 4 56 7 8 8 10 11213 14 1516 17 18 19 2021 22~
DATE BEGINNING FEBRUARY 3, 1968

Fig. 5.12 Effect of Reactor Outlet Temperature and Fuel Pump Cover-
Gas Pressure on Reactor-Access Nozzle Salt Level '
(as indicated by wall temperatures)
 

121

ORNL~DWG 73-600

 

1400

1200

 

   

- FUEL PUMP SPEED

——1_——— A G T S S

 

TE R-43

 

 

" TEMPERATURE (°F)
FUEL PUMP SPEED (rpm)

 

TE R—7

 

 

 

 

 

 

 

400

0 5 10 15 2
 TIME, MARCH 11, 1969 (hrs)

.

Fig. 5 13 Effect of Fuel Pump Speed (i.e. circulating bubbles) on
'Reactor-Access" Nozzle Salt Level (as 1ndicated by wall
temperatures) : ,

 

 

 

 
 

\

122

The effect of temperature and pressure can be seen in Fig. 5-12. At
the beginning of the dbsérved time period, the salt level in the RAN was
relatively low with the pump pressure higher than its normal 5 psig, After
a pressure drop on February 6, the salt level .overcame the immediate effects
of pressure and began to rise as indicated by the R-U43 thermocouple. Sev-
eral days laster the reactor outlet temperature was changed in two small
steps of 30° and 15° yet note the very large effect of almost 600°F on the
R-T7 thermocouple due to the salt level increase. Later by raising the pump
pressure from 3 to 5 psig the salt level is again lowered in the RAN annulus.

The effect of a direct change in the émount of circulating bubbles cén
be seen in Fig. 5-13. At the beginning of this pefiod the fuel-pump speed
vas lover than normal and it can be seen that the RAN salt level was very
high.' Other indications such gsthé reactivity balance led to the conclu-
sion there were no bubbles circulating at this qdndition. When”the fuel-
pump speed was increased, with no other change in operating temperatfire or
pressure, the RAN salt level dropped very rapidiy.}.(the the abscissa in
Fig. 5-13 is now hours instead of days.)

Circulating bubbles were involved in more dramatic effects than tem-
-perature‘changes in the RAN, of course. Some more important variables were
involved such as the effects of circulating voids end xenon poisoning on
 the reactivity balance. The effects seen in the RAN temperatures only
helped explain the cause of some,dther;events.

One phenomenom noted in which it is thought the RAN trapped gas had a
direct relationship is that of power perturbations in the MSRE in January,
1969. There was an unusually large amount of circulating bubbles at this
time as verified by several indications including low RAN temperatures.
Itris thought some gas, clinging to the core, was suddently swept from the
core causing a momentary increase in reactivity. The triggering mechanism
for release of gas clifiging in the core was presumabiy a8 very small pertur-
bation in flow or pressure. An occasional release of a burst of gas from
the RAN which was suddenly compressed in the pump could have heen the cause
of such perturbations.r There were other indications to connect the RAN

trapped gas with the power perturbétions.“7
 

 

123

5.10.4 Core Specimen Installation and Removal

Throughout the operation of the MSRE with salt in the primary loop
there was a sample array of one kind or another in the core. The arrays
that were exposed between September, 1965 and June, 1969 were of the de-
sign shown in Fig 5-14. The array that was in the core from the time of
construction until August 1965 contained similar amounts of graphite and
INOR-8 (to have the same nuclear reactivity effect) but differed in inter-
nal configuration., In 1969, during the last fivermonths of operation, a
.‘ different array, designed'tc stndy the effects ofksalt velocity on fission
product deposition, was exposed in the core.

A core specimen assembly of the type shown in Fig. 5 14 consisted of
three separable stringers (designated R, L, and S). Whenever an assembly
was removed from the core, it'was taken to a hot cell, the’stringers were
taken out of the basket, and a new assembly was prepared, usually including
one or two of the previously'exposed stringers. Sometimes the old basket
was reused, sometimes not. The history of exposure of INOR-8 specimens in
this kind of array is outlined in Fig. 5f15. The numbers indicate the
heats of INOR-8 from which the:rods in-each stringer were made.

The following section describes the experience with installation and
removal of core specimen assemblies, nith emphasis on the procedures, tools,
~ and systems involved. Descrlpticns of the materials that were exposed and
the results of their post-operation examinations are given in detail in

references 32—36. (

- Pre-Power Array —-The specimen array that was in the core during the

 

. prenuclear testing and nuclear startup experiments differed internally from
the later surveillance assemblies, but externally_it was similar and its
installation and remcval'emplqyed the equipment and procedures prcposed for
later use. The original installation was in 1964, before salt had been cir-
- ‘culated, and,-although"dcne.remotely, was closely cbserveg to determine
needs for modificatiqns in the tools and procedures. The assembly was re-
 moved in August, lQGSrby theiremote procedure. (At that time the salt was
slightly radioactive from the *®°U startup experiments.) Only minor diffi-

culties were encountered with some of the tools and fixtures.

 
 

©INOR-8 ROD OF TENSILE SPECMENS~,__

  
  
   

| GRAPHITE (CGB) SPECIMENS—__ .
" BINDING STRAP~_
o AT e Y

 

LN

o

A ‘/’/&\\\\\3/

 

 

 

 

 

Fig., 5.14 MSRE Surveillance Facility Inside Reactor Vessel

C \ ‘ C

el

 
_ STRINGER L.
 STRINGER R’
STRNGER'S'

" CORE TEMPERATURE -
HIGH
_SALT IN CORE

RUN NO,

 

 

 

 

 

 

5085
- 50814

5085
5065

 

 

 

 

 

 

5085
5081

67-5514
7320

 

 

 

 

 

 

 

5085 .
5081

 

 

21545. || e7-502°

. 67-551

 

 

 

21554 || 67-504

 

 

7320

 

 

 

 

 

1 0

 

 

 

4

 

 
 

ICIT1 10

 

 

1546 17 18 19 20

       

| Fig;‘S.ls Outline of MSRE Core Surveillance Pfogramf

 

ORNL-DWG 72-7495

6C1
 

 

126

While the core access was.open for the sample removal, a visual in-
spection was made of the core with a 7/8-in.~diam. scope. The inspection
revealed that pieces were broken from the horizontal graphite bar at the |
bottom of the core that was supposed to support the sample'assembly; The
broken piecesrwere recovered, using a vacuum cleaner. To circumvent this
damage, the new sample assembly to be installed was modified to be sus—
pended from above, using a special fitting installed in the strainer bas-
ket . above the core. This fitting (the "basket lock" in Fig. 5-14) had
spring fingers that locked into the strainer screen when it was. pushed
into place. The. lower end of the sample basket was extended to reach the
hole in the lower horizontal graphite bar to provide lateral support..

Array 1. Installation of the basket lock and the first standard ar-
ary in the core in September, 1965 was uneventful. The array was removed
in July, 1966 after about 1087 equivalent full-power hours.(EF?H) of opera-
tion. When the array was examined in the hot cell, portions of the string-
ers were found to have been damaged. Some obstruction’(thought to be salt
in the annulus) had been encountered in 1ifting the assembly through the
access nozzle. The damage was not caused by handling, however, but by con-
traction during cooldown. When the core was drained some salt was trapped
between the ends of graphite speeimens'where it froze and interfered with
the differential contraction of the parallel graphite and metal columns
during cooldown. As a result the graphite columns buckled.

Array 2. Because of the damage, none of the three Stfingers from the
first array were included in the second. The new stringers were modified
to,preventrtrapping of sait, but otherwise the array was like the first,
The second array was installed on Sept. 16, 1966, near the end of the 2-
month shutdown to replace the main’blowers.

The second array was removed in May, 1967. By this time the reactor
had operated- 4510 EFPH and the samples were removed -sooner after the end
of power operation, so the salt was more highly radioactive than before,
resulting in troublesome contamination on the tools and in the standpipe.
This slowed the operation somewhat and it was necessary to install a char-
coal filter in the standpipe vent line to limit iodine release to the stack.
Difficulty was encountered in obtaining a satisfactory seal on the reactor
access flange, which required feplacement of four bolts. Hot-cell examina-

tion of the array showed that the basket lock had pulledGOut of the strainer
127

screen and was stuck on the baSket; A new basket 1ock was therefore fabri-
rcated and was installed in the screen., _

Array 3. The third array consisted of one new stringer and two
stringers previously exposed in Array 2, so it was highly radioactive at
the time of installation. No- particular difficulty was encountered in
handling the array, however, ‘since the equipment and procedure were de- |
.signed for this condition,, A satisfactory seal on the core access flange
_was'not obtained on the first try, and inspection_showed~the O-ring gasket
was dented -and scratched..'Therbolt.holes'were cleaned and retapped and - a
,'new~gold—plated?ga9ketVWaslinstalled. An'acceptably tight joint was then
‘obtained. = 'lf--f o | - o

 Array 3 waS'removed on April 2, 1968 after the conclusion ofrzssu
operation (9005 EFPH). Some inconvenience resulted when the closure on
- the transport containment liner refused to operate properly, but the re-
-moval and transfer were safely accomplished without the" liner. While the
array was out. of the core the flange was sealed with a rubber—gasketed
blind flange to avold possible damage to. the permanent sealing surfaces.-

Array 4. Onme of the old stringers and two new stringers made up
-_Array 4, which-was installed uneventfully on April 18, 1968 | This array
was removed in June 1969 at 11 555 EFPH. The removal was delayed for
_.several days -due to problems with the removable heater that was required
to melt the salt from the tapered annulus between the removable plug and
the 2—1/2—in. access nozzle.f For the first time, flush salt was not cir-
: culated just prior to the sample removal The usual measures were taken
'gqto control the radiation and contamination during the sample removal opera-
:ytion. There was some contamination of the immediate work area (mostly

:nobledmetal fission produets) during removal of tools, which required
,mopping.'y _ o . ' '
e Special Array. The final array was all new and quite different in-
'f}ternally from previous arrays.r Externally, however, it was practically
rthe same and standard tools and procedures were used in its installation

5(Ju1y, 1969) and-removalg(December, 1969) _ Both operations were uneventful.

 
 

 

 

 

128

5.10.5 Radiation Heating

Radiation-produced heat in the reactor vessel walls caused the outer
surfaces of the vessel to be hotter than tfieladjacent-salt by;an amount
that was proportional to the reactor power. Any deposition of solids
would cause even higher surface temperatures. There were two locations
in the reactor vessel where solids could tend to accumulate. These loca-
tions were the lower head and the lugs located just above the inlet volute
and which supported the core matrix. The differences between thelreactor
vessel temperatures and the salt iniet temperature were carefully moni-
tored during power operation so that this condition would be detected had
it occurred. _ W ,

There was no indication of sedimentation during the operating life
of the reactor. The temperature difference did increase with use of **°U
fuel; this was expected dué to the higher neutron leakage associated with
this fuel. |

5.10.6 Discussion and Recommendations

The reactor vessel and core satisfactorily performed all functions ‘
for which it was designed. Originally the operating life of the fuel sys-
tem was to be'determified by the number of thermal cycles on the freeze
flanges} However, the design life of the freeze flanges was extended on
the basis of development tests. The stress-rupture life of therreaqtor
vessel was then reviewed to determine if the operatidn of the MSRE would
be limited_by the possibility of stress-rupture cracking. The reevalua-
tion indicated the reactor could be operated at least-another'year longer
than its originally predicted 20,000 hours s%ress—rupture life.’°!s’

The reactor access nozzle also adequately performed all of its func-
tions. It was obvious, however, that some amount of bubbles circulating
with the salt was'neéessary.to replenish the gas inventofy of the RAN an-
nuli and keep the salt level down to the desirable level. The RAN design

was thus inadequate from the standpoint of prdviding a frozen salt-seal

such as was the case with the freeze flanges. The»ifiability to freeze a
 

 

 

129

salt seal was the result of a larger than anticipated salt flow in the
RAN annuli which increased the heat load beyond the capability of the

cooling air-provlded., Additional thermocouples‘in»the area of the RAN
would have allowed better definition of the salt level.

If an access flange such as the RAN is to be used again in a molten
salt system, there are three possible methsds of obtaining its main func-
tion — that of preventing.contact of salt with ‘sealing surfaces:

(1) Provide an external source of gas to the annulus. A salt level
determination would also be'neeessary. _

(2) Provide an internal source of gas in the form of circulating
bubbles which would constantly replenish the gas inventory in the annulus.
‘A trapped gas pocket at a higher pressure than the pump cover pressure
 will not remain indefinitely. .

(3) ‘Reduce the access of flowing salt to the annulus by baffles,
‘labyrinth or such, so thatia frozen salt seal can be maintained.‘

If bubbles are to be purposely injected into the fuel salt for the
removal of xenon and krypton, then item (2) appears to be the best solu-
tion. The joint could be designed to remain essentially full of gas at
all times. |

5.11 Fuel and Flush Salt Drain Tanks

5.11.1" Description

, Two fuel saltrdrsin tanks (FDFIVand FD-2) were installed‘in-the drain

- tank cell for the safe storage ‘of the fuel salt during shutdown periods.

“?A third tank (FFT), also located in the drain-tank cell, was provided for
,storing the flush salt which was ‘used for cleaning up the fuel circulating

system before and after maintenance.

 
 

130.

A fuel drain tank is shown in Figure 5-16. The INOR-8 tank was 50 in.
in diameter by 86 in. high, and had a volume of 80 ft3, sufficient to hold
in non-critical geometry all the salt from the fuel ¢circulating system.

The tank was provided with a cooling system designed to remove 100 kW of
fission product decay heat. The cooling was accomplished by boiling water
in 32 double contained bayonet tubes and thimbles in each of the tanks.

The fuel flush tank was similar to the fuel drain tanks, except that
it had no cooling system. Since the flush salt did not contain fissile
material,ffhe only decay heating was from the small quantity of fission
products that it removed from the fuel system during the flushing operations.

The weight of salt in each of the three tanks was indicated by forced
balance pneumatic weigh cells.. The weights were recorded on strip charts
but could be read more accuratély from instélle&,mercury manometers. Two
‘conductivity type level probes (one near thé]bottom'of the tank and the
other near the top elevation of the salt) were provided for use as reference

points.

~S;11'2 - Calibration of the Drain Tank Weigh Cells Using Lead Wéights

During early testing in the fall of 196k the weigh cells were cali-
brated by loading the tank supporting rings with lead blllets. The drain
tanks were at ambient temperature during this calibration. Drifts in-indi-i
cated weight of up to 39 1lbs were'noted when no.changes were being made.
Repairs of air leaks in the instruments caused shifts of up to 65 lbs. The
weigh cells were further calibrated during the addltlon of flush and 235U
fuel salt. (See below.)

5.11.3 Addition of Flush.Salt and.Further.Calibration..of Weigh Cells

Later in 1964, thirty-six batches of flush salt.(v250 1bs per batch)
were added from cans in a portsble furfiace directly into drain tank FD-2
througha'heated line and flanged dip tube attached atethe‘inspectibn
flénge'oh the tank. Each batch of salt. added was weighed on accurately
callbrated scales and the tare welght of the container was checked to ob-
tain the net weight of salt’ added. All temperatures were maintained as
- near 1200°F as possible. During additions some difficulty was encountered

with salt freezing in sections of the addition line. On one occasion the
 

 

 

 

 

 

131

ORNL-LR-DWG 61719

INSPECTION, SAMPLER, AND
LEVEL PROBE ACCESS

   
 
  
  
  
   
   
  
 
  
  
    
  
 
     

STEAM DROME

WATER DOWNCOMER INLETS

CORRUGATED FLEXIBLE HOSE
STEAM RISER

BAYONET SUPPORT PLATE

STRIP WOUND FLEXIBLE % ' =1 ¥\ — GAYONET SUPPORT PLATE
HOSE WATER DOWNCOMER : —a i P4

GAS PRESSURIZATION
AND VENT LINES

FUEL SALT SYSTEM
_FILL AND DRAIN LINE

SUPPORT RING

VORI

)

BAYONET HEAT EXCHANGER
THIMBLES (32)

32\

 

 

 

 

 

 

 

 

 

 

 

 

 

=i s \ N THIMBLE POSITIONING RINGS
FUEL SALT SYSTEM _
FILL ANDDRAINLINE =/~ _ TANK FILL LINE - .

'Fig. 5.16 Fuel Salt Drain Tank

 
 

 

132

drain-tank vent line plugged. To locate and remove the plug, the vent line

was heated with a tdrch, starting at the tank and working down the pipe
until gas could be blown through the line into the drain tank. The plug
blew loose while heating the line several feet from the ténk. The plug
location, the relative low temperature required to remove the plug and some
oil found in the nozzle to the inspection flange on the tank led to the
conclusion that the restriction was csused by an oil residue. '

During the addition of the first two batches of salt, weigh cell read-
ings were taken every five minutes to determine when the probe light came
on, Then this salt was transferred thiough the transfer lines to FFT and
FDél and back to FD-2 through the reactor fill lines. This operation left
salt in all freeze valves which were then frozen to isolate the tanks for
the first time. - |

While the salt was in the FFT, two more cans of flush sélt were added
to FD-2 to recheck the location of the lower'probe. In the two checks the
probe light came on when 469.6 and 492 1bs hed been added to the tank.
These figures corrésponaed to weigh cell readings of 419 and 316 1bs. Af-
ter all of the flush salt (9230 1bs) hed been added to FD-2, it was trans-
'férred between the three drain tenks. By using the weigh cells on.both
tanks involved dnring each transfer, the weigh cells on FD-1 and;FFT were
calibrated. The indicated weight at the upper and lower probés on FD-1
and FD-2 were also noted. The lower probe of FFT was not functioning (See
2,15}, The weight between the probes was 7353 + 176 1bs for FD-1 and
T545 + 191 1lbs for FD-2. _

From the sbove data and sdbsequent operation, it was concluded that
the weigh cells are~inadequate for precise inventory work. In general,
the longer the time required for a given operation, the larger the error

introduced. Transfers of small.qnantities of salt over short time inter-

vals appeared to give fairly good data. Using the tank weight at the lower

probe light as a fixed reference was useful in preventing the transfer of
too much sslt to the fuel system and as an indication of the afiount of salt
remaining in the tank after a'fill.- The implication is not that the weigh
cells were at fault. Pipe loading on the tanks caused by temperafiure
changes of the tanks and adjoining pipes probably caused the varistions.

Errors as high as 200 to 300 lbs have been noted during transfers.
 

 

133

5,11, 4  U-235 Fuel Salt Addition

The carrier salt and larger additions of 235U enriching salt were ad-
ded by the same general method used for adding the flush salt. Details of

these additions and the 235y criticality experiment are given in Ref. LO.
5.11,5 Salt Transfers .

Salt is transferred between drain tanks by pressurizing the supply
tank and venting the receivér-tank. - Originally transfers were made only
through'the 1/2-in. transfer lines (107-110). In the event of a drain to
both drain tanks followed immediately by & refill of the circulating loop,
. considerable time was spent freezing the fill freeze valves (105 and 106),
thawing the . transfer freeZe"valves (108 and 109), refreezing them after the
transfer and then rethawing the £i11 freeze valves (105 and 106). After
careful” cons1derat10n procedures were modified to permit jumpering inter-
locks and transferring through the fill lines. No difficulties were en-
.countered, however, all of salt could not be transferred this way since

the fill lines did not go completely to the bottom of the tanks.
5.11.6 After Heat Removal -

On 2/17/64 and 2/2L4/6kL, tests were made of the heat-removal capacity
ofrthe-steam domes. Flush salt uas put into FD-2 and the salt was heated
to ¢1200°F. With 40 gallons of water in the feedwater tank (FWT-2), a
supply valve (LCV-80T) was opened to admit this water to all 32 of the
bayonet cooling tubes. The”water was refluxed for approximately two hours
“withrcooling tower'water-flow7to:the drain-tank coudefiserlkept coustant at
4o gal/min. The inlet and ex1t temperatures of the coollng tower water
" were monitored, as were the drain tank temperatures and feedwater tank

levels. At equilibrium condltlons, FD-2 temperatures dropped linearly at

- 'a rate of 86°F/hr. The heat—removal rate calculated from condenser cooling

 

.,water flows and temperatures was 139 kW/hr. The heat-removal rate calcu-
-~ lated from the temperature—decay rate of ¥FD-2 and the heat capac1ty of FD-2
~ was 132 kW/hr. S o | e
A second similar test wes made bj'oyening the other supply valve
(ESV-807) and keeping LCV-80T closed. The flow through this valve was

 
 

134

~marginal, the level in the FWT decreased slowly and the heat-removal rate
vas only 110 kW/hr based on a water heat balance and 98 kW/hr from the salt
temperature decay rate. , ,

These data showed that the bayonet coollng tubes were more than ade—
quate to remove the design decay heat rate of 100 kW/hr (for IO—MW opera-
tion). Since the reactor was run at only approximately 8 MW, the margin
of safety was enhanced.

Because of leaks through the seats . of the water supply valves, the
reactor was operated some of the time with the feedwater tanks empty and
procedures were written to manually valve water to these tanks if neces-
sary to remove afterheat. When the salt was divided about equally between
the two drain»tarks, &s would occur on an emergency drain, no heat removal
was necessary. Turning off the drain tank heaters was sufficient to prevent
excessive temperature increases. For instence, on November 2, 1970, the
reactor was drained while operating at full powér. 4380 1bs of salt drained
to FD-1 and 5130 1bs drained to FD-2. With the drain tank heaters turned
off, FD-1 temperature increased from 1060 to 1155°F in sbout 6-1/2 hours
and then started decreasing. FD-2 temperatures increased from 1075 to
1195°F in sbout '9-1/2 hours and then started decreasing.

5.11.7 Plugging of Helium Lines.at.the Top..of. the.Fuel.Drain-Tanks

The helium supply, vent and equalizer lines enter the top of the fuel
drain tenks via a common 1/2-in. Schedule-40pipe. In June 1969, this line
on FD-2 became almost totally plugged. The plug first appeared when the
main 4.2 liter/min He offgas flow was routed past the drain tanks due to
plugging in the regular offgas routes. The plug was at least partially
cleared by first heating the tank to 51275°F overnight and then backblowing
with helium at 42 psig. This technique was not successful when FD-1 line
plugged in August 1969. Due to this plug, fuel drained preferentially to
FD-2 during a planned drain on November 2, 1969. During the last drain on
December 12, 1969, the fuel was made to drain more equally by.deliberately.
closing the vent from FD-2 for part of the time during the draining period.
The result was that the totsal salt weights in FD-1 and FD-2 were within 5%
of each other after the drain had been completed.
 

135

5.11,8 Loading 233U into FD-2

After the fuel salt had been processed to remove the 235U, the re-
maining carrier salt was transferred from the fuel storage tank to the fuel
drain tanks. The method of addlng 233y enrlchment salt was d1fferent than
that used for adding the flush salt and 233y fuel salt The loadlng equlp—
ment as shown in Fig. 5-17 was attached to the access flange of FD-2, Cans
contalnlng v 7 kg of uranium were lowered 1nto FD-2 whlch contalned half of
the molten carrier salt. When the salt had melted out of the can, it was
withdrawn into the shield. Mlxlng was accompllshed.by transferrlng the
salt between the two drain tanks. .The 233y addition and crltlcality ex-

perlment is described in detail in Ref. Ll.

5.11.9 Heatup and Cooldown Rates.
These are descrlbed under heaters in Sectlon 17.
5.11.10 Discussion and Recommendations

The fuel and flush salt drain tanks have fuhctiqned vefy satisfac-
torily. They were'very sluggishrtfiermally and therefere easy to control
during operation. The afterheat removal syetem'ffinctioned satisfactorily
when needed.

The weigh cells were not completely satisfactory for inventory pur-
poses. Piping stresses should be eliminated or other type instruments
provided. o

As with other offgas lines, the p0931b111ty of plugging should be con-
sidered. Perhaps two lines could be used w1th heaters attached to melt out

any plug which mlght develop. '

 
 

 

S

136

ORNL-DWG 68-967

   
  
   
  
  

Il ——TURF CARRIER

GRAPHITE SAMPLING

SHIELD
PURGE GAS NN
SUPPLY
o ‘.w |“E‘E by
P o TOOL EXTENSION

SEALS

    
   

~ TURNTABLE AND l
STORAGE WELLS Y

CONTAINMENT ENCLOSURE
AND STANDPIPE ASSEMBLY

Fig. 5.17 Arrangement for Adding *°*U Enriching Salt to Fuel Drain Tank’
 

 

137

6. COOLANT SYSTEM

_rC. H. Gabbard J. K. Franzreb
i M. Richardson

6.1 Description

The coolant circulating sYstem consisted of a centrifugal-type coolant
punp and an air-cooled radiator which were connected to the fuel heat ex-
changer by 5-inch INOR-8 piping. The normal operating temperature was 1000
to 1100°F at full power. The coolant salt was circulated at 850 gpm. " Dur-
ing shutdowns the saltdwas—drained-to the coolant drain tank. The piping
had low points on each side of the radiator. Therefore, two drain lines
‘and freeze valves were required. The circulatiug system was filled by

pressurizing the coolant drain tank with helium and venting the coolant

‘pump.
'6.2t Purgihg,Moieture~and Oxygen from the Bystem~ -~

3Before;salt»wa3wcharged;intouthe coolantdrain‘tank,oxygeniandmois;
ture were purged from the coolant circulating system and the drain tank.
This was accompllshed by first pressurlzlng the system to assure that it
was 1eak—t1ght then evacuatlng, ‘and purglng w1th helium during the heatup.
Details of the heatup are glven in Sectlon 17, The hellum was 1ntroduced
at the coolant pump and - coolant draln tank and evacuated or vented through
the offgas system, The coolant pump was operated to prov1de circulation
in the loop.' Purglng contlnued untll the system'was at 1200°F and essenti-
ally no more-m01sture was . belng released Later analy51s of coolant salt

samples 1nd1cated that ‘the purglng had been very effectlve.
6.3 CoolantfiCirculating;SYStem,Calibratibn.anchrainuEime
_'After the coolantosalt:had been added to the COOlant-drain tank (see

Section 6.1) the coolant circulating system was filled and operated for
11 days. A test of the drain time indicated that complete draining could

.:? be accomplished in- approxlmately 12 minutes after the freeze valves thawed.

During the following £111 the c1rculat1ng system was calibrated The drain
tank was pressurized in 1ncrements and the drain tank welght and the dif-
ferential pressure between the drain tank and the coolant pump were re-’

corded. The system calibration curve is shown in Fig. 6-1.

 
DIFFERENTIAL PRESSURE BETWEEN DRAIN TANK AND LOOP (psi)

25

15

10

 

138

ORNL-DWG 73—-601

 

 

 

 

 

 

      

 

 

 

 

 

 

 

 

 

PUMP BOWL ——1g54
£
—jse50 g
o
ol
-
=
S
. O
Q
' Q
RADIATOR >
w
—J8ss o
L
o
-
/ m
L
. w
o
W
-
-l
-
w
: —1840 &
LEAVING MAIN HEAT EXCHANGER oz
ZZENTERING MAIN HEAT EXCHANGER ’<>E
| i
|
SALT TEMPERATURE ~1130°F
DENSITY | ~124.2 Ib/F3 ; | —1 835
800 1600 2400 3200 4000 4800 5500

POUNDS OF SALT IN COOLANT LOOP

Fig. 6.1 Coolant Circulating System Calibration
 

 

139

- 6.h Operation

Other than the difficulties with the radiator described later, the
coolant system 0perated gatisfactorily. No salt leaks were detected at

any time.

6.5 ~Coolant Salt Circulating Pump

6. 5 1 Descrlptlon

The coolant pump used ‘to circulate the L1F-BeF2 coolant salt was simi-

lar to the fuel pump. It did not have‘a spray ring since fission product

gas removel was not required. No overflow tank pr cooling shroud was in-
stalled. At the operating speed of 1750 rpm, it had an»output'of 850 gpm
at a T8-ft salt head. A small helium purge was introduced Just below the

shaft seal in the bearing housing. Most of the gas flowed downward through
the labyrinth between the shaft and the shield block to prevent salt mist

from reachlng the seal. The remainder flowed upward to prevent any oil

which leaked through the seal from getting into the pump bowl. 0il for

_lubrlcatlng the bearings and -cooling the shleld plug was reclrculated by

an externsal pumping system., Heljium bubblers and a float-type level instru-
ment were used to measure the liquid level. (A float-type instrument was
not instelled on.the fuel pump.) The coolant pump was located outside the

reactor cell and could be directly maintained shortly after shutdown.

6.5.2 Operation

The coolent pump ran véry well throughout the eperetien”ef the reactor,

”ffAs W1th the fuel pump, the narrow renge of 0perat1ng 1eve1s caused opere-
f?tlonal dlfflculties. Prior to starting the pump after a f111 the average

loop temperature could not be accurately determlned Slnce the coolant—~

pump ‘level changed sbout: T% per 100°F change in. loop temperature the se-

1ected £i11 p01nt was not always satlsfactory for operatlon. ~In which case

‘the freeze valves had to be thawed and the system level adjusted The nor-

mal operating level was about 55% at full power (the level 1nstrument had
a span of 10 inches for 100% and centerllne of the volute was at 35%).

 
 

140

When the pump was first started after a fill, the level would drop about
9% due to a gas pockét in the loop and ebout 4% due to the pumé fountain
flov. | | |

During load and rod scrams, the system cooled rapidly which decreased
the level below the interlocks required for starting the pump. Without
circulation there was a possibility of freezing the radiator. An emergency
start switch was installed which ensbled the operator to bypass all inter-
locks and start the pump if freezing of the radiator appearéd eminent.

Based on the oil found in the coolant pump offgas lines, there was a
small oil leak into the pump bowl. The offgas line at the pump bowl be-
came'pluggéd occasionally and in November 1968, a clamshell heater,was in-
stalled. This was used once and cléared the line Bufficiéntlyrto permit
operation until shutdofin_Decémber 12, 1969. -

The float level indicator was not used for operation.

6.5.3 Conclusions and Recommendations.

 

The coolant pump was used to circulate salt for 26,026 hours at tem-
peratures between 1000° and 1200°F,

Some plugging of the offgas line was encountered. This should be con-
sidered in future design. Leskage of o0il into the pump bowl ‘could be elimi-
nated by seal-welding a gasketed joint as described for the fuel‘pump
(Section 6.1). | | B

6.6 Radiator

6.6.1 Description

The nuclear power generated by the reactor was finally dissipated by
a coolant salt-to-air radiator. Salt flowed through the 120 radiator tubes
at 850 gpm where it was cooled from about 1075 to 1015°F at full power.
Approximately 200,000 cfm of air from the two stack fans flowed perpenQ
dicular to the tubes and out the coolant stfick. In péssing fhrough-the
radiator, the air was heated about 100°F above the ambient inlet'tempera-
ture. The heat removal was;édjusted to match the nuclear power produced
by raising or lowering the two radiator doois, adjusfing the bypass damper,

or starting and stopping the main blowers.
 

 

 

 

141

The radiator tubes were 0.75 in. OD x 0.072 in. wall x 30 ft long,
zee-shaped to fit into the insulated radiator enclosure. The body of the
enclosure supported the tubes and the electric heaters. Two hundred
twenty-five kW of heat were provided to prevent freezing of salt in the
tubes in case of a sudden lees of nuclear power (see Fig. 6.2).

The upstream and downétream faces of the radiator enclosure were
equipped with doors that could move up and down in a "U" shaped track. As

the doors moved downward 1nto the fully closed position, camming devices

'operated by the weight of the door forced them against the seals located

on the face of the radiator. The doors were raised and lowered by means

of cables driven by a single-motor.' Magnefiic clutches and brakes permitted
independent operation'ef;eitherfldeor. In case of a load scram request or
an eleetrical pdwer.outage, the brakes and clutches were deenergized and”
the doors fell closed at a rate which was limited by the inertia of fly-
wheels attached to the ends of the drive shafts through overrunning
clutches (nT ft/sec). |

6.6.2 Early Tests and Difficuities[

Preliminary heatup and checkout of the radiator began in October 196k,
The main difficulties were due to insufficient insulation and warping of
the doors. As & result, the doors jammed and would not operate properly
and there was too much heat less to reach operating temperatures. The tem-
perature distribution wasep00r?and overheating of thermocouple and electri-
cal leads was encountered. |

After unsuccessful attempts to correct these difflcultles, the doors
were temporarily sealed to allow completlon of the crltlcallty tests while
new doors were being de51gned and built. I

During the last half of 1965, extensive changes, tests, and remodifi-

'cations were made on the radlator assembly. New doors were installed which

| prOV1ded more insulation fac1ng ‘the heat source and addltlonal expan51on

joints to prevent warpage.reThe_ 'soft" seal gaskets were removed from the
doors and relocated on_thefradrator enclosure. They were increased from
3/4 in. to 1-1/2 in. widerfer”better sealing. A modified "hard" seal was

installed on the doors. The door guide channels were repositioned to

 
ORNL~LR-DWG 85844R2

142

 

   

 

o

, \ TR A il Ly B ) e Ay sl
% ] ;

8 .._
Ao NN N
T. <
- - ,.-_b...

 

 

Radiétor Coil and Enclosure

Fig; 6.2
 

 

143

provide more clearance, Four cams were installed on each door to provide
more positive seals when they were closed.

Sheet metal door hoods were installed above the radiator to reduce the
air leakage into the penthouse, Thetsynchro indicating device on the drive
shaft had not provedrrelisble since it was possible for the door to jam and
the shaft to continue to rum in the downward direction. End rollers were
p051tloned to prevent side motion from jamming the doors. Additional upper
limit switches and mechanlcel stops were added to prevent lifting the doors
in the event of a limit switch fallure. Llftlng motor,overload switches
were installed in conjunction with the stops. | |

The new doors weighed about 2700 pounds each which required the ad-
dition of two 20,000-pound capacity, l-in. stroke shock absorbers on each
door. Although it was fecognized that with the additional weight, operation
~of the existing brakes'and clutches was marginal, no changes were made in
the lifting mechanism. Adjustment of the door-lifting cables so that the
doors hung straight did much to ease their operatioh ih the slides,

~ The sbove modificatidns'improved:the heat distribution of the radiator
system. Two major difficulties were.then encountered; one was the exces-
sive temperature rise in the pent house just above the radietor, the other
- was maintaining a negetive pressure in the pent house with respect to the
high—bay pressure, This was needed for beryllium containment.

Efforts to correct thése'prbblems ificluded relocating the heater lead
Junction boxes, pulllng the thermoplastlc wire (194°F) from the junction
f,boxes to the heater leads comlng from the radlator spreadlng out the bead-

ed heater leads to improve the cooling air 01rculatlon_through the W1r1ng,
‘and, modifying the exhaustrdhoting to direct the sweep of cool air across
' the wire troughs (¢500 cfm);':These efforts were successful in reducing
“the temperature to,aboot 100¢F}just above the insulation on top of the
radiator. Heat leakagé;aroofi&.the'dOOr hoods and radietof'eficlosure was
- reduced by weldlng up cracks, addlng 1nsulatlon, and 1nsta111ng boots
- around the door—llftlng cable openings in the door hoods. A 1-7t2 oross-
‘overduct was installed between ‘the door hoods to relleve the pressure on
the inlet hood. The pent house pressure with respect.to_the high bay was
aided by the addition of a duct from the pent house to the inlet side of

 
 

 

 

 

144

the south annulus blower. This reduced the pent house pressure to -0.1h4
to ~0.28 -inches of water without the main blowers. With both blowers in

cperation, the pressure became slightly positive, -
6.6.3 Subsequent Operation and Modifications

The systems were heated and the approach to full power was started
| early in 1966. The radiator functloned sat1sfactor11y, however, due to
heat 1osses through the door seals, ‘it was dlfflcult to heat the empty
radiator to- an acceptable temperature distribution prior to a f111

The heat transfer cepability was found to be sbout 20% less than cal-
~ culated. The cause for this dlscrepancy has not been deflnltely established
but may be due to the large surface-to-alr temperature difference which
existed. This is described in detail in References 20 and 36.

Cooiant salt was frozen in the radiator tubes on two occasions. The
first occurred on May 19, 1966 when a building'pomer failure caused arload
scram and loss of all eiectrical equipment. Two tubes of the radiator were
partially frozen as indicated by the temperatures.reading about T5°F low.
These temperatures recovered a few minutes after'the coolant flow was re-
established 1nd1cating that the plugs had thawed. ,‘

The second incident occurred on June 27, 1966 when again the electri-
cal system failed, this time causing g8 coolant salt drain because of low
radiator outlet temperature, The coolant salt weigh cells indicated that
gbout 200 1bs of salt were held up in the coolant system. OSince it was
probable that the salt was frozen in the radiator which had cooled below
the freezing point, the radiator was reheated and the system refilled When
the pump started, the temperature scanners indlcatefl that 5 tubes were fro-
zen and not circulating. These tubes thawed after several minutes of salt
circulation without damage to the tubes. | B |

To reduce the possibility of this happening again, the control system
was revised_So that 8 rod scram'wouldfalso cause a load scram. The low-
temperature setpoint for a load scram was increased from 900°F, which is
only 60°F sbove the liquidus temperature, to 990°F.2 Other revieions were
made so that the coolant:pump would not be tfirnedleff unless absolutely

aeceesary. A radiator door scram test was eenducted'from full power, and
 

 

 

145

"these revisions were adequate to prevent freezing of the radiator as long
as the coolant pump remained running. The average temperature of the fuel
and coolant leveled out at about 1120°F. | | |
~On July 17, 1966, a sudden loss in power indicated that a main dblower
had been lost. Main blower No. 1 had failed catastrophically and pieces
of the aluminum rotor hub and blading had gone into the radiator duct. Af-
 ter the radiator had cooled, examination revealed that pieces of aluminum |
had passed through the radiator tube bundle, several dented tubes were
found, and several pieées bf,aluminum were stuck_to tubes. Metallographic
éxamination determined there were no punctures in the tubing and the radi-
ator was safe to operate after each tube was thoroughly cleaned. The éxact
cause of the blower failure was never specifically détermined but examina-
tion of numerous "old"-cracks,in.the biades and hub would indicate that
these were the probable point of origin. Flying debris caused some damage
to the heater lead wires, wire trays, ete. S |
It was apparent from inspection of the "hard" sealing surface, which
was mounted on the hot side of the radiator door, that a modification would
bé required. The continuous strip metal seal was badly buckled and frac-
tured due to exposure to the radiator heat. The roughness of this surface
damaged the "soft" seal (two widths of 3/l-in. Johns-Mansville Thermocore
#C=19T7 square asbestos'packing'strips mounted on the face of the radiator)
 so as to render this seal ineffective.
A short evaluation program of six types of "hard" seals resulted in
7 the,selection.of a seal madé*up'of-2-1/h—in. long overlapping steel seg-
.ments spaced l/32—in.}apért'which was installed on the new doors. The test
further indicated that the "I" bar, to which the segments were plug-welded,
bowed as much as 3/16 in. on the 3-ft-long test sectlon.i The "T" bars in
| ~ the new doors were provided with expansion slots plus a ten51on bar mounted
' }'on the cold 51de of the door which was used for stralghtenlng the doors
 after heatlng. b o
 Performance of the radiator enclosure and door was without incident
after the above modificatiéfiéffiérefcbmpleted for the balance of the MSRE
. 0peratifig period. Heat'logsgs_and mechanical'opefatidn were adequate.
Programmed maintenance on the lifting mechanism was perfofmed when the op-

portunity presented itself, Miscellaneous small items, such as replacement

 
 

146

~ of the "soft" insulation along the top of the outlet door, replacement of
broken ceramic insulators on the heater leads within the enclosure, repair
of grounded heaters, etc., was the extent of the maintenance required. The
door on the inlet side of the radiator retained its shape and ‘sealed reason-
ably-fiell,'the seal along the top and sides of the outlet door retained its
seal except for replacement of burned-out tape along the top. The bottom
_ of-thé butlet door continued to distort ahd, even after being straightened
several times, made & poor sesl as & result of the door bowing away from
the soft seal on the radiator face, As long as the side and top seals were
intact, there was no chimney effect and the heat losses were within ussable

6.6.4 Automastic Losd Control

The‘heat removal rate at the MSRE was dependent upon the positions of
the radiator doors and bypass damper and on the operation of the main blow-
ers. These could be manually manifiulated as desired. Instrumentation was
also provided to automatically sequence the operation of these-componefits
so that the operator could use one switch for increasing or decreasing
heat removal rate. An intermediate radiator door setting with one blower
in operation and 1 MW of heat removed was to be used as the starting point.
Since it was found that the heat removal with the doors at théir minimum
setting of 12-in. was about 1.9 MW, the automatic system did not function
properly. The rest of the sequencing functioned properly durihg a'test.
However, in view of the above and since manusally adjfisting the heat load
was not a disadvantage, the sutomatic load control was never used in the

operation of the reactor.

6.6.5 Comments and Recommendations

Assessment of the radiator containment, doors, and door-lifting mech-
enism is complicated by the many changes, large and small, which were re- |
quired before acceptable performance was achieved. o

The radiator enclosure difficulties can be attributed prlncipally to
details of installation such as gross air leakage into and around the en-
closurg as a rgsult of cracks, unsealed openings, insufficient insulation,
lgck of a sufficient number of expansion joints, ete. The prificip&l_radi-

ator enclosure support "H" beam members were not vertical and therefore
 

 

147

the adjustable "U" shapped'dooriguides located within these members had
,to be remachlned to prevent the doors from binding., ,

The lifting mechanism was unduly complex. A substantial increase in
both rellabillty and control would have been achieved by prov1d1ng each
~ door with its own motor-gear reducer unit with a built-in brake on the mo-
tor rotor shaft. A small fast-acting brake in the high-speed end of the
drive would have reduced the nuflber of electromechanical devices and re-
-duced or eliminated the problem of coordinating time constants between
' clutches breakes, and motor;-'The ability to operate the doors indepen-
dently was probably unnecessary

The unrealistic close tolerances between the doors and the radiator
enclosure end distortions produced by thermal gradients, primarily in the
doors, required a massive effort to correct in order to,provide the re-
: quired seal. These factorsjshould be given careful consideration in any

u'future design.
6.7 Perfbrmance of Main Blowers, MB-1 and MB-3

6 7.1 Descrlgtion

- The two main blowers are a part of the heat reJection equipment where
the heat generated by the reactor is dlssipated to the atmosphere. The
blowers were manufactured by the Joy Menufacturing Co. and are the "Axivane"
type Model AR-600-360D-1225, The blowers are direct-connected to 250-hp,
1750~rpm motors by & short drlve shaft with a flexible shim - coupllng on
- each end. _ : L o -

. Although the blowers had been origlnally purchased for. another program,
~_ithey were essentlally unused at. the beglnnlng of the ‘MSRE program. The
blowers were each orlginally rated at . 82,500 efm at 15 inches of water or
'“illh 000 cfm of free air delivery.. The MSRE radiator design requirement of
"100,00070£m-from each blowerrab 9 inches of water was'witbin ‘the normal
L}_operating range'df'the bléwérs._ Had new blowers been procured for the

iMSRE this same type of equlpment would have been considered & satis—
"ffactory choice. L | :

 
 

 

 

 

 

148

6.7.2 Preoperational Testing -

The pre0perationa1‘testing of the blowers consisted of sbout 25 hours
of total operation., When the blowers were first deliveredénd installed,
‘they were operated & short time (less than 10 hours) to provide the elec-
tricaliload for testing the diesel generators. The blowers were also oper-
ated for a short period of time prior to power opération of the MSRE to
check thé'radiatof tubes for self-excited vibrations, to.permit a calibra-
tion of the stack air flow instrumentation, and td determine if control
circuitry was necessary to avoid certain abnormal opérating conditions.

During the abnormal operation testing, the blade-pitch was found to be
less than the specified 20 degrees. The pifch was set corréctly and the
test program was repeated. The tests indicated that the bloWerwaould
operate satisfactorily at all the normal conditions, but that a "surge"
condition would occur if the byrass damper were partialiy closed when the
radiator doors were also closed. The manufacturer had stated during dis-
cussions of the test program that the "surge" condition could be expected
but that no damage would occur other than overloading the drive motors.

The blowers were shut down immediately after the "surge" condition was en-
countered and administrative control was used to prevent further operation
under surge conditions. A second pitch change was made to 22-1/2 degrees
immediately after the first full-power operation. This 22-1/2 degree pitch
gave the maximum blower performance within thé power'rating 6f the drive

motors.

6.7.3 Operating History

The-aerodynamic performance of the blowers was satisfactory through-
out the operating life of the reactor. However, there were several me-
chanical failures which are listed in Teble 6.1 and which are discussed .
below. This table also shows the effect of tne failure oh‘the~e§eration
of the reactor. | - | |

MB-1 Coupling Failure — After sbout 550 hduré of operation, the flexi-
ble coupling on the motor end of the floating drive shaft failed during '
operation of MB-1l. The motor end of the shaft was then supported'bnly by
the coupling guard, andrthe shaft whip that occurred as the blower rotor
 

‘?; ”TaB1e 6.1 'Summary of Main:Blower Failures

 

 

Date ~Failure - Effect on Reactor Operation
6/14/66  Coupling Failure MB-1 Zero~Power Operation for 14 hours
7/17/66 ~ Hub and Blade Failure MB-1 Premature Shutdown from Run No. 7

3/6/67_'
1/16/68

~ 8/31/68

9/17/69

9/19/69

 ; Bearing:Fai1uréfMB~3 .

. Bearing FfiilfiféfipMB-l'M3;3_

*"Be&ting Replééépent MB-3.

Beafing and Mount Replacement MB-1

Faulty Bearing Replacement MB-1

~ Partial Power‘Operation During Run No. 8
' Zéro-Power.Operation'f@r 3 days

" Partial Power Operation for 6 days.

Maintenance Performed During Scheduled
Zero-Power Operation

~ Completed During Scheduled Maintensnce

~ Period

"Zero-Power Opération for 2 days

Partial Power Operation 3 days
Zero~Power Operation for 3 days

 

YT

 

 
 

 

150

coasted down destroyed the guard and also the coupling on the end of the
shaft near the blower. The shaft came to rest embedded in the sheet-metal
nose with the blower end of the coupling near, but completely severed from,
its mating flange on the rotor. The motor end of the shaft-ektended radi-
ally across the blower inlet at about the L o'clock position. 

Scratches and dents on the blades indicated that some of the debris
from the failed couplings had gone through the blowe} into the radiator
duct. However, the damsage, except'to the couplings, appeared to be super-
ficial and of little consequence. No additional inspecfionrof the blower
was made. | » ' h

The primary cause of the failure, wé beliéve, was a fatigue failure of
the flexible shims. The-exact cause of the shim failurerisrunkndwn.‘ Some
home-made washers were found in the drbrisrthat wére not roundedias they
should have been.  High stresses at the corners of these washers could have
started the failure, or the coupling could have been running with excessive
misalignment,- The blower rotor and drive shaft assembly had been rotated
by hand when the final blade pitch change was made about 216 bperating-
hours before the failure. There may have been some cracked shims in the
coupling at that time, but the shims were still supporting the shaft.

The couplings were rebuilt with new bolts, washers, and shims. The
shaft was realigned, and the blower was returned to service. The couplings
on MB-3 were also disassembled and new shims were installed.- The original
shims in these couplings were not in bad condition. - | |

- MB-1 Hub and Blade Failure.— After about 650 hours of additional op-
eration, there was a second failure of MB-1l. This was a-éatastrophic fail-
ure of the blading and the hub of MB-1. The outer periphery of the rotor
hub disintegrated and all of the blading was deStroyed.'rfibst of the frag-
ments wére contained in the blower casing, but numerous pieces of the cast
aluminum-alloy hub gnd blades entered the radiator duct and some of these
actually passed thrdugh but did not seriously damage the radiator'tube
bundle., Figure 6.3 shows the failed hub and typical pieces of the blading.

An inspection of the broken pieces of MB-1l revealed numerous "old"
cracks in the blades and in the hub as evidenced by darkened or dirty areas

on the fractured surfaces. One blade in particular had failed along a
 

151

 

   
   

 

 

 

' "% PHOTO 84351
rFig. 6.3 Outer Periphery ofrFailed MB-1 Hub and Typical Blade Fractures

 

 

 

 
 

152

large "0ld" crack. The hub had contained short, 1-1/2 to 2-inch, circum-
ferential cracks at the base of 8 of the 16 blade sockets and the failure
generally followed these cracks. Figure 6.4 is a piece of the hub showing
the darkened areas that indicated previous cracking at the basé of the
blade sockets. Similar cracking was found in the hub of the spare blower
“that had been in storage and a continuous crack extending dbout 35%vof the
circumference as well as some of the shorter cracks were found in the MB-3
“hub. No cracking was found in the bladihg of MB-3 or the spare blower.
The exact cafise of the MB-1 failure isruncertain, but a blade failure at

- one of the existing cracks'éeemsrto be the‘most probable fifst,evént.

All three of the blowers were rebuilt by the manufacturer with new re-
designed hubs and lighter magnesium-alloy blading. The hubs were rein-
forced with radial ribs to reduce the behding moments in the areas where
the cracking had been observed. There have been no further cracking prob-
lems in either of the rebuilt units in about 11,000 - 12,000 hdurs of op-
eration. A more complete discussion of the failures and of the corrective
action that was taken is given in References 42 and 43.

Thrust Bearing Failures — There were a total of six failures assoc-

 

iated with the main thrust bearing or its mounting. Main blower MB-3 was
taken out of service on March 6, 1967 for a bearing replacement after the
vibrations had increased from a normal value of about .8 to 2.5 mils. The
bearing contained an excessive quantity of very old appearing grease, and
the bearing surfaces were severely worn and pitted. This bearing was ap-
parently from the replacement blower that had been in storage, and the
bearing apparently hfid not been cleaned and relubricated when the rebuilt
rotor was assembled. Improper lubrication was judged to be the cause of
failure., The bearing had operated about 1800 hours.

Main blower MB-3 was shut down a second time for a bearing replsacement
on January 16, 1968, because of ;n increase in bearing vibration. Although
the vibration and temperature had been normal, the MB-1l bearing was also
found to be defective and was replaced at the same time. The MB-1l bearing
had one pitted ball, but the MB-3 bearing was severely damaged and one ball
had actually fractured. Figure 6.5 is a photograph of these two balls.

Since improper lubrication did not seem to be the cause of these two

failures and since three bearings had failed in relatively short operating
 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 6.4

 

e

£ST

 

 
 

 

 

i

 

 

 

 

 

 

154

 

 

 

 

P;l;t 1 ™ |§ Y ¥ ‘.tij T
BORATORY

         

 

”
£

   

 

 

 

| OAK RIDGE NATIONAL LA

 

 

Fig., 6.5 Ball Bearing Failures Taken from MSRE Main Blowers MB-1 and
' MB-3 in January 1968 '

   

 
 

155

 times, the expected'life of these bearings was-calculated from an estimated
set of load conditions. The predicted life was 5000 hours as compared to
operating times of 6630 and 4800 hours respectively forLMB—l and MB-3. As
& result of this evaluatlon the original radial-type ball bearlngs were
lreplaced with an angular—contact type having identical mountlng dimensions.
The angular ‘contact bearing had a greater thrust load capacity.

" The angular contect bearinQS'were'apparently capable of satisfactory
1life, but additional difficulty vas experienced with the self-aligning
mount. The MB-3 bearing was replaced dur1ng a scheduled malntenance period
lbecause of a thumping noise that wvas heard when the blower was turned by
hand. The bearing was found to be in good condition, but the wear pattern
indicated that the outer race'had been misaligned with the shaft. The mis-
‘alignment plus the high radial clearance in this type bearing csused the
noise. No further dlfflculty was experienced with MB-3.
| Blower MB-1 was shut down on September 17, 1969 because of an increase
'_ in vibratlon. The vibration increase had been caused by excessive wear in
the spherical surfaces of the self-aligning mount. The replacement of the
mount also required replacement of the bearing. The new bearing failed by
overheating during its_testlrun beceuserof insufficient clearnaces in the
phenolic ball retainer. This bearing was then replaced with a bearing of
the original radial type that was on hand. This bearing was satisfactory
for the remaining operation of the reactor. The defective bearing was re-

turned to the vendor.
~ 6.7.4 Surveillance of Blower Operation

 Following the'hub'andflblade failure of MB-1, a surveillance program

'f[for monltoring the operatlon of the blowers was 1n1t1ated Vibretion pick--‘

 

ups were mounted horlzontally on each bearlng of the blowers and drlve

amotors and thermocouples were installed 1n each of the blower bearlngs.

- A v1brat10n meter w1th fbur 1nput channels was prov1ded for each blower,

'and the v1brat1on of each bearlng was recorded in peak-to—peak dlsplacement
once per shift by the operating personnel.. The thermocouples were moni-
tored by the computer and 1n1tially an slarm would be given 1f the tempera-

ture exceeded & preselected llmit. Thls was revised later to alarm if the

 
 

156

temperature rise above ambient exceeded a limit. This permltted much
tighter operatlng llmlts because the ambient fluctuated with the outside
air temperature. _

In addltion to the dlsplacement readings ‘taken from the V1brat10n
;meters, an output -was also available to display thevylbretlon trece on an
oscilloscope or other high-speed reccrder, Oscillosccpe photographs were
taken for reference on new bearings and-when the vibration meters indicated
an increase. Figure 6,6 shows. a comparlson of vibration traces from a nor-
mal bearlng and from the bearlng containing the fractured ball.

A complete inspection of,the blowers including a dye-penetraht‘examina-

ticu'cf the hubs and blading wes completed annually duringrshutdown periods.
6.7.5 Discussion and Conclusions

Following the major rebuilding of the main blowers.in 1966, thefcper-
ation has been generelly satisfactory even though there have been several
power reductions caused by bearing problems. The total lost operating time
caused by the variousbeering‘failures_was 8 days of zero-power operation
and 9 days of partial-power operation. A better thrust bearing arrangement
could have been required. | - -

The surveillance program to monitor the mechanical'terformance'6f the
blowers has been very effective. All the bearing problems except the MB-1
pitted ball in January 1968 have been found by an increase in vibration.
When the ball fractured in the MB-3 bearing, a temperature‘ihcrease had
preceded the vibration increase by several days. However, this-tempera-
ture increase had gone unnoticed beceuse the bearing temperature was well
below the limit that had been set. The computer program was therefbre re-
vised to alarm on the temperature rise above ambient, and much tighter.
limits were selected based on the previous operating experience. This
alarm system was effective in detecting the defective new bearing on Sep-
tenber 19, 1969. :

The 0perat1ng experlence W1th the v1brat10n monitoring system has
| clearly shown that the displacement meter alome is not adequate to mcn;tor
the condition of the bearings. Some system of maintaining a history of
the frequency spectrum is also requlred We used photographs of the oscil-

loscope trace to document the normsal operatlng vibrations, and additional
 

157

; .
.-
" S s iy i sl ik I

 

Trace with Fractured Ball — Taken 1/16/68

Fig. 6.6 ' Main Blower MB-3 'Vib-rati:on Traces

 
 

158

photographs were taken for comparison when an increased vibration was indi-
cated by the dispiacement meters.- A mbdification tortheieofiputer to in-
crease its samplihg rate would have{permittednthe f:equeney spectra to be
generated by the computer. This would have provided'a more convenient and
more_preciée spectrum, but some degreerof:backgrofind experience would still
be required for_properAinterprefation of the spectra in either case because
there can be several normal frequencies. There #ere'several instances of
abnormal vibrations when the bearlngs were not at fault. Tfie final de-
cision to replace a bearing was made in each case by the sound and feel of

the blower durlng a coastdown from operatlng speed.
6.7.6 Recommendatlons

The main blowers were conventional commercial equipment which would
normally be expected to operate with a minimumrof attention or maintenance.
The instrumentation of all rotating equipment is'probably impractical in
regard to both cost and manpower. However, certain classes of equipment
should be instrumented and monitored on a routine basis. The equipment
should be selected on the basis of:

l. Size and cost

2. The'accessibility durihg operation

3. The effect of shutdown on overall plant operation

i, The potential for self-destruction |
Ideally the monitoring should be computerized but this is not actually
required. _ , |

Each type and piece of equipment will have its own operating characte-
ristics in regard to vibration or temperature, and_these characteristics
may vary with the insfalletion or operating conditibns.' Preliminary limits
may be obtained from the manufacturer or from Reference hh but the final
operating limits should be selected from a background of normal operating
experience. )

The experiences W1th ‘the cracked hubs and with the orlglnal thrust
bearing design are indications that the Quality Assurance Program should
extend into the design phase as well as the fabrication phase of major or

critical items of equipment.
 

 

 

159

6.8 Coolant Drain Tank

6. 8 1 Description

The coolant drain tank vas - similar to but smaller ‘than the fuel flush

_ tank. It was 40 in. in diameter by 78-in. high end hed & volume of 50 ft3.
As there was essentially nofafterheet in the coolant salt, cooling was not

- provided;_ Two selt lines (204 and 206) and their associated freeze valves
were used for filling and draining the coolant circulating loop. !

- The weight of selt was indicated by forced balance weigh cells. The
weights were recorded on,strip oharts but ‘could be read more accuretely
from installed mercury manometers., Two conductivity type%prObes (one near
the bottom and the other_near;the top elevation of the salt) were provided

for use as reference points.
6.8.2 Calibration of the Weigh Cells Using Lead Weights . .

The weigh cells were calibrated by loading the drain tank with lead
billets. The drain.tank was at ambient temperature during the calibration.
There was good agreement between the manometer read1ngs and the actual
weight added.. Difficulty was encountered with air and mercury leaks in
the readout system. The weigh cells were,further calibrated during the
addition of coolant salt. - d |

 6.8.3 Addition of Coolant Salt:

Twenty-two batches of coolant salt (5,756 1bs) were added from cans in
& portable furnace directly to the coolant drain tank through a heated line
attached to the top of the tank. After the addition was complete, this

B line:vas cut off gnd,welde&fgfifit;"fAVgood:linear plot of the manometer

f readings vs aCtual fieight,was”obtained_except for a 150-1b step in the
'5eurve'fihich‘occurred wheh-a”brokeh fitting in the weigh cell tubing was

;repaired. The relationship between the weigh cell readings and the loca-

:rtions of the probe 1ights was estdblished.

The weighing system on. the coolant drain tank proved to be con31derably

| V;fmore steble than on the fuel drain tanks. With 5756 1b of salt in the tank

 

 

at ebout 1200°F the extreme spread of 40 1nd1cated welghts, taken over a

period of a week, was t 22 1b.

 

 
e o A e " S

 

160

6.8.4 Heatup and Cooldown Rates
These are described under heaters in Section 17.
6.8.5 Discussion snd Recommendations “
| : N6 unfisual difficulties were encountered with the coolant drain tank.
The weigh cells fUnctiohed satisfactorily, however, the trouble with the

fuel drain tank weigh system emphasizes the need for cereful consideration

of piping stresses or the use of other type instruments.
 

161

7. COVER-GAS SYSTEM

P, H. Harley

| 7 The MSRE cover-gas system consisted of a helium trailer for normal sup-
ply, two banks of three helium'cylipders.forremergency_use, two parallel
,treating_stations_for_rempvingrmoisture and oxygen to a nominal 1 ppm by
7_roiume, a treated helium surge tank, and two hesders,_35 psig,end-250-psig,
~which supplied cover gas to the various. systems. Each of the treating sta-
tions consisted of a dryer (molecular sieve) preheater, and an oxygen-
removal unit - (heated tltanium sponge) Mbisture and oxygen anaiyzers could

be valved in to monitorrtheetreated or untreated helium as desired.
7.1 Initiel Testing

The components of the treating station were developed and designed
by the Reactor Division Development Section. The'performance of the sys-
tem was tested by measuring oxygen and moisture content of the cover gas.
upstream and downstream;of:theetreating station to insure proper opera-

. tion. Other testing at the MSRE consisted of 1eak-testing the system and

, checking flow control and temperature controls, Thermal expansion of the

1 0, removal ‘beds during the initial heatup caused the 3-in., ring joint
flanged heads to leak, After retightening the flanges, the units were kept
hot to prevent thermal cycling the flanges.f , |

i Late tests indlcated that the Fisher Mbdel 67—H pressure regulator
}.which was installed to reduce helium from traller pressure to 250 psig was
_}susceptible to moisture diffu51on through the diaphram., These tests indi-

) ';_cated Victor.Model VPS-201 - and Grove Model RBX-20h-15 were better from .

 

 
 

 

 

162

this standpoint. The Victor model was less complicated and therefore was
installed 1n the MSRE.

The low-pressure cover-gas header was 1n1t1ally a hO-p31 header which
was protected by a U48-psig rupture disc. During initial testlng, this rup-
ture disc failed on several occasions withoutVapparent'cause; The header
pressure was subsequently lowered to 35 psig'which was sufficient for all
normal operating requirements and did alleviate the problem, The safety
 1imits2% (Lo He supply pressure) had to be lowered from 30 to 28 p51g to
provide s margin from the normal operating pressure.

Four helium control valves failed in the first month of operatlon.
Some galling of the close flttlng trim (17-4 PH plug and Stelllte seat)
was apparently caused by the extremely clean, non—lubrlcatlng_dry helium.
Other valves in the system, inclu&ing the four replacements have  operated

satisfactorily.
7.2 Normal Operation

Since putting the cover-gas system in operation in the fall of 196k,
sixteen trailers of helium have been used. About 21,800 ft3 of the
29,000 ft3 in each trailer was used before returning the trailer to be re-
- filled. A trailer lasted about 4 months on the average or an average usage
of approximately 180 cubic feet per day. Normal usage when the reactor was
- operating was 260 cubic feet per day. Larger amounts were used during pur-
ging and filling procedures and very little was used during maintenance

periods.

Although helium was the normal éover'gas, during maintenance periods:"

‘fhe reactor system was purged-with argon. Argon was also used during a
special study of circulating gas in the fuel. Bottles of argon containing
<b-ppm moisture were placed on one of the emergency cover-gas headers and
then fed through the heatihg station into the regular cover-gasrdistribua
tion system. -

The cover gas was changed from oée treating station to the other only
once in the five years of operation. This was done in February 1967 when
trouble was encountered in the power supply control circuit to the opera-

ting'unit. ‘Although the dryer which had been in operation prior to the

s
163

: changerwas-regenerated,ithererhas been no indication that the capacity of
any dryer or Oz-remoValsunit'has been exceeded. - k |
Spectrographic analysis of the helium received averaged 2 ppm oxygen
.and L, 6-ppm m01sture. After. passing through the treating station, the |
3helium contained v 0, 2-ppm 02 and from 0.5 to 10-ppm m01sture. The vari-
| ation in. moisture analysis appeared to follow the ambient temperature of
- -the helium trailer and-treating station. During hot weather the analyses
~ were high and.during cold weather the results'wereflow}_ Fluctuations could
be seen from day_to_nightLtemperaturerchanges'of several:ppmfmoisture.
~ This fluctuation of~moisture!analYSisrwas.decreased:by better heating of
. the sample;station-locationtin the winter ‘and better ventilationgofrthe
area during the summer months. ) o
. While: investigating the moisture analysis. problem, another dryer was
t~installed in the main helium header between the normal treating station and
- the line to the analyzers. . Whether this second. drying step»helped is doubt-

 

ful but it did give assurance that the helium'was as dry as possible,

The electrolytic cell in the moisture analyzer caused'cons1derable-
difficulty during early operation. The cell was-replaced four times be-
‘tween June and-Octoberf1965;‘once.in June 1966, and in January 1967. No
cell replecements have been'necessary;since‘then. Apparently the quality
of the cells was improved. Only"one'electrolytic cell has failed in the
oxygen anelyzer and thet was in June 1965

To -check the final performance of the treating station in October 1969,
a cylinder of He containing moisture was connected to the analyzers. The
7analyzers read 20-ppm moisture ‘and 0. 6-ppm O,. After the gas was run
--through the treating statlon it analyzed 0. 9-ppm.moisture ‘and 0.2-ppnm O3.

' There has only been one faeilure of the 48-psig rupture disc since the
‘initial testing. This was: caused.by 1mproper installation. ‘Each time the

60-p51g fuel system pressure ‘test was run, pressure was put on the back

'”iVS1de of this rupture disc to prevent it from'breaking.r "In August 1969,

 

when the baokpressure was increased the rupture disc started 1eaking.—
'During replacement the vacuum support was found to be on the wrong side
of the disc. Repeated flexing, when backpressure had been applied, was
the apparent cause of the failure. More 1mportantly,,the faulty 1nstalla~
tion had nullified the safety function of the disc.
 

 

164

T.3 Conclusions and Recommendations

The cover-gas system performed quite satisfactorily with a minimum
amount of maintenance or operastor time. Since the moisture and oxygen in'
the heiium as received was very low, it is doubtful that the treating sta-
tions were réally hecessary. By obtaining analysis. of the cover gas before
usage and having an on-line analyzer, the treating station could probably
be eliminated. In larger reactors, the treating-statiofi might be needed
to dry the early recycle gas. The MSRE charcoal beds remained "wet" for
a long period during early operstion. |

We have had considerable fluctuation in the indicatéd mbiéture. It
the treating station and analyzers were located in an area which had sa cofl—
stant temperature, the analyzers would be more consistent. If no treating
station is used, the helium supply should be kept at a constant tempera-
ture. Rising temperature drives moisture out of the container walls and

decreasing temperature causés the walls to absorb moisture.
 

BB o ARk AR AR 4 L R BTSSR e pans e

165

8. VOFFGAS SYSTEM
A. I. Krakoviak

8.1 rDescription

-

The offgas systems were comprised of those 1ines,'filters, charcoal
beds, and valves through which cover—gas flowed from the salt systems to

the environment. The off—gas systems for the coolant salt and for the

- fuel salt were similar in many,respects, but the_coolant off—gas system

was simpler because it -did not have to deal withvthe intenseiy radioactive

fission products found in the fuel off-gas during pdwer operation.

8. 1.1 Coolant Off-gas System -
Figure 8.1 is the flowsheet for the coolant off-gas system (simplified

. by the omission of elements which are not pertinent to the presentation

of-operating,experience—inrthis_report).f'Gas entered the coolant pump bowl

through bubbler level-indiCatdrsfandias a purge to the pump shaft annulus.

‘The shaft purge flow was split, with some flowing downward into the pump

bowl andrthe,rest_upwards'and out through line 526. The upflow was to
preyent.oil fumes'from)coming;downitHE'annuffis into the pump bowl. How-
ever, some oil (about_O.SgFfil.b cc/day) did get~into the pump bowl by a
leekage'path:around the shield plug. This oil was thermally decomposed
into yapors“and'sootrthatileft'fiith,the off-gas through line 528.

Gas flows into‘and“out of'the pump'bowl were continuous. Gas was

"admitted at.a fixed rate and the pump bowl pressure was controlled by :
rthrottling PCV-528 in the off—gas system. The driving force for the |
- upper gas flow (1ine 526) was the pressure differential across PCV-528.
‘gg'Gas ‘was normally vented from the coolant drain tank only’ after a fill of
m'the circulating loop, which was done by pressurizing the tank.

‘All piping in the coolant off*gas system was 1/2—inch Sch-40 stain-

iless steel pipe. The valves, filters, and traps are described in the sec-
'Lftion on experience._ The system'was located in the coolant drain cell,

- which could be safely entered;whenever the reactor power was shut down.

 
 

 

 

 

 

 

166

ORNL DWG 73-k2Th

 

 

 

 

 

   
  
    

| S |
O, 09_5&”\ N
| V=526
olL CATCH '
14 SLm &z TANK 526 FILTER
:CF =SF |
—3_|_[“ A 2 CAPILLARY
C%%L}fg?‘ vsigh | L2 OFLow l'z\ksmnc-roa

  

  

  

 

 

coT

 

 

(L

 

 

% g
SE
Q
i o%
7PL N Vi L
2 98
Cle o
0ao.
n
|
|
| Ta
L Je 35k
HV-527 HCV-536 1z
Zlrd
T Lo i
HCV- 547
| V-560A
STNE 524 FROM FUEL PUMP
COOLANT
DRAN
TANK

P4—NORMAL CLOSED

DJ—NORMAL OPEN
B<—THROTTLE D |
-==—LATE ADDITIONS TO THE SYSTEM

Fig. 8.1 Simplified Flowsheet of MSRE Coolant Off-Gas Systemr-

LINES57

- FROM

AL'ro STACK

- HCV-557C

 

-

CHARCOAL BEDS
 

 

167

8.1.2 Fuel Off-gas System**~"~

As seen in Fig. 8.2, the flowsheet for the fuel salt off-gas was
basically similar to that for the coolant off—gas but was more complex.
Three drain tanks were required in the fuel system (two for fuel, omne for
flush salt) and an overflow tank was required for the fuel pump bowl.r A
'major difference was that all. fuel system off-gas was sent through charcoal
beds for fission product absorption before release to the stack. The gas
hold—up and cooler consisted of a 68-ft length of 4-inch Sch-40 stainless
steel pipe. The remainder of the off-gas piping with the exception of - the
flexible jumper line near the pump bowl and the line in the vicinity of
~ the particle trap and line 557 was 1/2—in. Sch~40 stainless steel pipe.
Therefore, at the normal off—gas flow rate of 4.1 liters per minute, the
fission gases were at least 45 minutes old on-arrival at the particle trap
and approximately 100 minutes old on arrival at the ‘main charcoal bed.

An auxiliary charcoal . bed was provided to accommodate large venting
rates (for short intervals) .such -as were required during filling and
draining the reactor or during salt transfers.- A sampling system was also
provided to study the concentrations of fission products and oil decompo-
sition products in the fuel offwgas. nf

The fuel off—gas system was contained and shielded and therefore re-
quired remote maintenance techniques for work on any portion upstream of

the charcoal beds. o

8.2 Experience -With' Coolant Salt Off—gas system

JA summary of the difficulties encountered with the coolant off-gas
system is given in “Table 8.1. A description of - these is given in the
following sections.  ':'1;;{757

"_s 2.1 Filter and Valve Difficulties o
During early checkout of the coolant system, the original pressure
,-control valve (PCV—528) which throttled the off-gas from the coolant pump
~'was too small (C =0, 003) for adequate pressure relief and was replaced

by a valve having larger flow coefficient (Cv = 0 02).

 
 

168

J

wo3sAs seH-330 TN MISH 3JO 3I29ysmord parITrdmrs gz°g °*S14

TrS-AS3

. , o W3HDUND -
T30 HOLDVIY WOMNY d=—"—=1 .HIIJNVS WOUJ PS-A
T R -2

 

   
           

 

 

, ~3ASVYHd
NOISNIAYSY HIMOG

. ONENA QIAOWAY

SININOIWOD ==-=

g311105nL P

'a3507D WWYON pd

@ N3dO IYWUON []

 

 AGN393T

 

 

 

 

 

 

 

         

 

 

L dai B
m.;....wl_.m..t#w

 

 

 

 

 

 

 

 

 

 

 

ANVL
HoLY D O

 

 

SWaLsAS Lwi 100
Y WO woyd

 

SLZN=EL DMT INNO

 

 
169

 

 

June

71 Sept

1969

Table 8.1. Summary of Difficulties with the Coolant Off-gas System
+ Date Describtion
Jan. 1965 __Installed larger control valve (PCV;SZB) (C 0. 032
‘ ' ' to 0.02) ' |
Jan. 1965 Added a 25u'filter ahead of the control valve |
 Jan. 1965 ' Replaced the 25 filter (1.4 sq in.) |
Feb., 1965 :’Installed:a 1argér 25u fiiter'(SO'sq in.)
- June 1965 Control valve plugged —-cleaned this and installed
S ‘a 1p filter (50 sq in.) ; :
Jan. 1966 Replaced 1lup filter
' ‘Feb. 1966 wo.oow oo
Feb. 1967 m
~ Feb. 1967 W
“June 1967 " woeon
 Sept 1967 noooowoow
Jan. 1968 oo n
Sept 1968 on mooon
Sept 1968 'Restriction at coolant pump — cleared by backblowing
‘Nov. 1968 Restriction at coolant pump — cleared by heating
- with a torch. Installed heaters on.the line.
Cleaned off gas lines and replaced 1lu filter.
Feb. 1969 Restriction at coolant pump — cleared by heating
o with electrical heaters.
1969 Control valve ly filter and check valve cleaned

and back—diffusion preventer. removed..

' Replaced lu filter._

 

 
 

170

A sintered metal filter (tubular cartridge one-inch long by 1/2-inch
diameter) capable of stopping particles greater than 25y was installed in
the off-gas line to protect the pressure control valve. It was necessary
to replace this filter (because of excessive pressure drop) after only 24
days of service with salt in the coolant loop. Inspection showed that
the plugged filter was covered with amorphous carbon containing traces of
the coolant salt and INOR-8. A second identical cartridge plugged after
20 daye of service and at the completion of Run 1, (see Table 3-2 for dates
of rune) it was replaced'with one having 35 times the surface area. Cool~-
ant salt -was not circulated again for 3-1/2 months and then for only 118
hr. During circulation, pressure control again became erratic which indi-
cated an obstruction of either the,filter or the valve. Both were removed
for inspection, and although there was no deposit on the filter, the valve
was partially obstructed by a black, granular material. Rinsing with ace-
tone restored the original flow characteristics of the valve, and it was
reinstalled for the next run. The filter was replaced with a similar one
but capable of removing particles greater than 1lu in diameter. The smaller
pore size was chosen because the black granular material was determined to
be a mixture of glassy spheroids of coolant salt approximately lu in di-
smeter and carbon or carbonaceous material.® It is not known whether
these materials were carried into the line continuously or were swept out
of the pump bowl by sudden venting. The valve was relatively trouble-free
after the installation of the 1p filter. The filter, however, experienced
gradual plugging and was change& or cleaned twice in 1966, three times in
1967, three times in 1968, and twice in 1969. The filters, when removed;
from the system, showed no evidence of foreign matter other than & light
film of 0il in the filter housing and a light film of oil was transferred
from the porous metal filter onto the plastic bag used in remo#ing the
filter. | _

Flow tests on the plugged filters showed pressure drops of less than
0.1 psi at flow rates two and three times greater than the normal MSRE rate
(essentially the same results were obtained on a test of the fuel off-gas
filter)., Plugging in these filters was attributed to the liquefaction of
oil in the pores of the filter. The fact that the pressure drop (at nor- |
mal MSRE flow rates) changed from V10 psi before removal to essentially
 

 

 

171

zero when tested after removal was attributed to the removal of some'of
the oil from the pores onto the plastic bag and rubber gloves during
removal ‘and insertion of the filter in the flow test rig.' The filters
were rinsed in acetone before reuse. '
“ On several occasions when the filter became restricted and could not
"be'replaced during power operations, ‘the output from the pressure controller
was switched from PCV-528 to HCV-536. The coolant pumppressurewas then
" controlled with this 1arge by~-pass valve (C = 3,5). Although the valve
was operated at essentially its fully open or fully closed position, ade-
' quate pressure control was~achieved because the check valve and back-
diffusion preventer offered an appreciable resistance to large flows.
During the May 1967 shutdown, ‘the ‘off-gas piping downstream of HCV-536 was
_changed.to join line 528_downstream of PCV-528 rather than line 560. This
;change routed the gas by the ‘coolant system radiation’monitor"so that the
7coolant off-gas could be. monitored when it became necessary to use HCV#536
" to control the system pressure. ' ’ |
* On one occasion late in reactor operations (September 1969) when the

filter becaue restricted the coolant system pressure was allowed to reach
10 psig before remedial-action-was taken. During this test the pressure
1n the system meandered,backfandlforth'between 10 and <5 psig for approxi-
" mately one month before exceeding IOppsi and thus requiring pressure con-
trol with the vent-valve;(HCV—536). This pressure drop is in agreement
‘with the gas pressure required to enlarge or burst a film of oil from a
| one-micron pore if one-assumes;that"the bubble radius is the same_as'that
"vfbtfthé'pbre;- The equilibriumlequation for a spheriCal'bubble,is
_-AP ';:-EEL -

A=

bd where g = surface tension = 16 dynes/cm for oil. in the coolant system; and
- T =bubble radius = pore radius = 0. 5 x 10°* cm. ST
- " This calculation indicates that any pore in the sintered metal filter
smaller than lu, once plugged with ‘condensed oil vapors, would remain:
| plugged at pressure differentials of less than 9. 3 psi. ,The slow pressure
| oscillation experienced could be attributed to a near equilbrium condition

of condensation and-evaporation'of oil at the pores of the filter.

 
 

 

 

 

172

- 8.2.2 Restriction at the Pump Bowl _

In September of 1968 approximately one month after the start of Run
15, in addition to a restriction at the filter, a restriction developed
in the off-gas line at the exit of thé coolant pump bowl. This restriction
was blown clear once by closing the pump offfgas énd equalizer valves-and
allowing the pump pressure to increase to 15-psig before suddenly venting
the pump to the drain tank which was at atmospheric pressure. The reactor
was shut down in November of 1968 when this restriction reappeared and a
similar restriction developed in the fuel off—gas system.,

During the shutdown, line 528 was cut approximately_B feet dbwnstream
of the_pump bowl and a clean-out tool which was inserted toward the pump
bowlkcame,put_covered with black, heavy grease. A brown vapor (hydro-
carbons) was collected wfien this section of line was heated with a torch.
The restriction was not cleared, however, until the junction of the off- |
gas line and the pump bowl was heated to red heat. There was practically
no evolution of vapor from the line at that time, suggesting that the re-
striction was primarily salt or a combination of salt and hydrocarbons.

A heater was installed on the line at the exit from the pump bowl;
also a trap which was packed with stainless-steel mesh was installed in
the line approximately six feet from the pump bow1 in an effort to con-
dense oil vapors at this location and thué prevent or minimize oil trans-
port downstream to the filters and valves.

Also during this shutdown all coolant off-gas valves in the coolant
drain cell were dismantled and cleaned. The valves and interconnecting
lines and filter housing were found to have a light layer of oil on the
inner surface, and approximately 10 to 20 cc of o0il was cleaned from low
pockets in the interconnecting lines.

Approximately six weeks after resumption of operations, the line at
the pump bowl again showed evidence of a rgstriCtion. The heater was
energized and the restriction cleared when the temperature of the pipe
under the heater (normally at 485°C) increased to 62060. No further off-
gas problems were-encountered at this location. However, about six months
later résfirictions appeared in the pressure control valve, check valve,
~and the back-diffusion preventer. During the June 1969 shutdown, the fil-

ter, control valve, and nearby check valve were cleaned and reinstalled;
 

 

 

173

"the back—diffusibn preventer was removed completely. Although all four
components were coated with oil which impeded off-gas flow, it is believed
that the back—diffusion preventer was the major restriction.

With the exception of the restriction in the sintered metal filter in
September 1969, no further coolant off-gas problems were encountered before
final shutdown. | 7 7 o S

In summary, the restrictions in the coolant off-gas lines were pri-
-marily related to oil leahage into the pump bowl.h In'retrospect a filter
of an intermediate pore size (say 2-4u may have been a better choice. The
restriction at the pump exit was probably‘more of a salt mist carryover
phenomenon rather than hydrocarbon related. - The plugging at the coolant
pump bowl exit, although occurring less frequently, seems to be similar

to that experienced at themfuel pump which will be discussed later.
8.3 Egperience withfithe:Fuel Off-gas System

‘A summary of the fuel off—gas experience is given in Fig. 8.3. The
salt circulating times and power operations are shown and the type diffi-
culties encountered are tabulated These are discussed in the following

sections.

8. 3 1 Experience with Fuel Sgstem Off—gas System Prior to Power Operation
The problems encountered with the fuel off—gas system during the pre-

~critical and low-power operations (<25 kW) were somewhat similar to those

in the coolant system.," _"' o _ _'i j : ,

" 8. 3 1. 1 Filter and valve difficulties. After about two months of

flush salt circulation during the precritical testing,_the fuel system

:rpressure control became erratie.u During the shutdown at the end of Run 1

- the filter and va1Ve were removed - The filter (pore diameter of v25u)

*i"appeared clean but the valve"was partially plugged The obstruction was

'-blown out with gas and the valve was washed with acetone, reassembled and

;'both filter and valve reinstalled.r Approximately a week after the start

'f'of ‘the next run, the valve again became partially plugged. This time it

flgwas replaced with one having a larger Cv (0. 077 instead of 0 022) The
smaller valve was then cut open_and a black deposit of amorphous carbon

and glassy beads was found'partially'covering the tapered stem. The. acetone

 

 
- 17L

Portion of Off-Gas System lrvolved

   

 

 

 

 

 

 

 

 

Fig. 8.3 Summary

 

 

 

 

 

 

 

of 0ff-Gas Experience

 

 

 

 

 

 

 

 

 

Saltin Salt in Portion of Off—-Gas System Involved ORNL—-DWG 73-2031
Loop 522 Loop 522 [Particle
RunNo. Date L% | e ket | oo Other Flush O3 : Neor | Trap | MCB | ACB
" Fiush id res nlet RunNo. Date  Fuel Power (Mw} FP | Area | Intet | Inlet| Other
‘ 26 26. MCH's plugging — unplugged using electrical heaters.
\ 12
1. PCV-522 plugged (C, = 6.022)
2. Cleaned PCV—522 — 25u filter appears OK. 12
' 27. Line 522 st FP starting 1o piug. Cleared by forward blowing.
! 3. PCV-522 plugged — replaced with larger vaive . 27 28. Starting November 25, 1967, operated for 130 days on MCB—1A
(¢, = 0.077). 28 28 to test hotdup capacity. 522 restriction at FP appeared and then
cleared itself.
' 29 29. Line 522 a1 FP starting to plug.
4, 522 filter repiaced with a 1 x 0.34 sq. ft. surface
! area fitter.
i 30 30. Line plug in 522 at FP disappeared.
i‘ 14 .
! 3 31, Line 522 at FP starting to plug.
5. Plugs at PCY—522, FE—521, CV-£33, and inlets to
b MCB's. Unplugged PCV—522 by operating it several
times,
' 8. Replaced PCV--522 with a large hand valve, replaced ) )
! £272 filter with a 50y x 0.34 sq. in. surface srea. 32 32. Mechanically resmed out Line 522 at FP.
! Removed FE~521 and CV—533 and! did not replace.
l MCB inlet plugs dissppeared.
| 7. Plugging at HV-522 B and at MCB and ACS injets.
8. Blew outoff—gaslines, removed 522 fifter and
] installed the first (Mark 1) particle trap and a
charcoal trap, Instatied a larger PCV-522 (G, = 0.885)
and replaced MCB inlet valves with larger valves,
{ 9. Poppet trom one of the drain tank vent CV's in MCB—1
' iniet plugged the line. Checks indicated that norw of
the CV's will closs. .

10. MC3 inlet plugged — cleared by forward- and back:

blowing. .

11, MCB and ACB inlet plugging — unplugging about

. weekly by back—blowing. The particle trap piugged. 3 . Line 5! Juagi
FP pressure increased from 3 to 12 psig and seemed 15 3 33. Line 522 ot FF plugging.
to be power dependent. Rerouted Line 524 during

: this period. V-557B now used to control FP pressure.

12, Particle rap AP decreased to ~1 psi at Zero power
and remained low during later power Operation. .

: 13. Particle trap plugging. FP pressure increased to 8.7 34 34. Mechanically reamed out Line 522 at FOD.
: psig.

14, FP pressure decreased to normal when power was 16
reduced. :

15. Flush salt overtilt.’ 35. Line 522 at FP starting to plug. MCB’s plugging — unplugged

) ) using heaters and back-blowing. Inlet handle on MCB—2A

16, Cleaned up oft—gas lines from overfill. Repiaced 35 35 inl v ke with in ch Tt
particle trap with another identical one (i.e. Mark 1}. } inlet vatve broks with vafva in closed position.
instatied heater on ACB infet.

17. For 26 days the off-gas system had not plugged. Lo , .

Then 2 days after raising power, 522 plugged at the 7 k. 36. Plugin Line 522 st FP blew out while emptying OFT.
fuel pump. Back-blowing was unsuccessful. MCB

started plugging and al! 4 beds wera put on--stream, . .

The particle trap plugged periodically. 37. ACS inlet partially plugged. Back--biew to clear.

18. Unplugged 522 at FP by heating and back-blowing. a8 38, Line 522 at 533 plugged § times. Back-blowing clesred the

19. Line 522 at FP plugged and particle trap AP increased. plug. Off--gas was vented through the drain tanks ~1 day.

20. Reamed out 522 at FP and replaced jumper. tnstalled
heaters on MCB 1A and 1B inlets. Heating for B hours 39 39. Line 522 at FP plugged.
removed the plugs. 40 40, Particle trap plugging — valved in spare particle trap. ACS

inl iald 8 | .

21. Partiche trap plugging — rerouted otf—gas through 4 inlet partially plugged. Forward—blew 10 clear
DT's which allowed enough coeling to eliminate 18 42 41, OFT vent (523) plugged.
plugging. 42, FD-2 vent plugged.

22, Replaced particie trap with two Mark I particle 43 4] 43, Installed heater on Line 522 at FP. Heated line to unplug.
traps (one in service and one valved out). Removed Replaced Line 523, Heated and back—blew FD-2 vent to
PCV-522 and charcoal trap and did not reinstall ramove piug.
either.

23. MCB 1A and 1B — AP had increased to 7 psi from
~2 psi at the start of Run 11. Therefore changed
to MCB 2A and 28,

24, MCB-2A AP increased from 2.6 to 9 psi in 10 days.

Heating removed plugs from MCB 1A and 18 and
back—blowing removed the plugs from MCB 2A 19
and 28. 44 44, FD-1 vent partially plugged. Back-biowing did not remove
45 plug.
46 45, MCB inlets plugged twice and were cleared by back-blowing
and heating,
. . Restriction i v f . 2. Cleared b
25. {nstalled heaters on MCB 2A and 28 inlets. 4 o;::?:::c.;:_' ® upstream of FT No. 2. Clea v
20
47 47. Line 522 at FP starting to plug.

 

 
 

 

175

rinselfrom the first cleaning (Run 1) and the deposit removed from the
valve during Run 2 contained 1- to 5-p glassy beads which proved to have
the composition of flush salt. The salt beads in the off—gas lines were
probably frozen droplets of mist produced by the stripper spray in the
fuel pump. Just as in the coolant system, it is not known whether the
mist was transported continuously or was swept out by sudden venting.
The performance of the fuel off-gas system,was satisfactory for the
remainder of Run 2 and the . zero~power experiments (Run 3). During the
'shutdown following Run 3 larger sintered metal filters (filter area =
0.34 ft ) capable of removing particles greater than 1lu in diameter were

installed ahead of the pressure control valves® in both the fuel and coolant

-53‘. o
T

off-gas lines.

8.3.1. 2 Check " valve malfunctions. The check valves in the off-gas

 

system are springfloaded poppetfitype valves and were designed for low
pressure drops. During'precritical checkout of the system when a bit of
debris'caused one such:valve.torremain open, all the other similar valves
were removed for'inspection,;cleaning, and pressure drop checks before
reinstallation. 'Those which;did not open with a forward pressure of at
least 5—in.)H30-and recloseiwith a forward pressure differential of n2-in.
H20 were replaced with new valves meeting these criteria.

The next indication of a check valve malfunction was in January 1966
when a restriction developed in the check valve downstream of HCV-533. The
blockage occurred after the_fuel system pressure was vented for a short
period of time through'this valve to the auxiliary charcoal bed. Although
some foreign material, shown in Fig. 8.4, was found in the poppet of the

.'check valve, it was judged not sufficient to stop the gas flow (the plug
- may have been dislodged‘during disassembly) The soft O-ring (which makes
‘the seal)'and the cone'ofrthe_poppet were covered with an amber varnish-
 like material which was identified as a hydrocarbon.‘r The check valve was
. not replaced and administrative control of HCV-533 was used to prevent a

 backflow of gas into the fuelépumppgas space during venting of the drain
tanks. | o | | |

In April 1966 difficulty in venting through the auxiliary charcoal
bed was found to be due to a check valve poppet which had become lodged in
the inlet line to the bed. Apparently the poppet had vibrated loose from

 
 

 

176

PHOTOAR27593'

 

Fig. 8.4 Check Valve CV-533

 
 

 

177

a check valve at one of the drain tank lines and had been carried down-
stream during venting: operations._ The inability of the check valves to

.prevent reverse flow in the exit lines of all three drain tanks implies

that any and all of these.valve poppets may have vibrated_loose.

 8.4. Difficulties with the Fuel Off-pas System

During Power Ascension

The fuel system off-gas problems became somewhat more complex during
power ascension than thoSe,encountered earlier because of fission fragment
production in the reactor and subsequent transport of the fission gases and

"some solid decay daughters in'the off-gas lines. Some of the lubricating
7_oil which leaked into the pump bowl (0.5 to 1.5 g/day), was polymerized to
‘a varnish-like substance by the high radiation field in the off-gas lines
‘and components. This material, amorphous carbon from the cracking process,
 and solid decay daUghters"of;the:fission'gases were trapped in the various
components of the off—gas system such as filters, check valves, capillaries,
valves, traps, and presumably in ‘the charcoal beds. The difficulties ex—

. perienced during this period are reported by run numbers or major shutdowns

8 4. 1 ‘Run Run 4

 The first indication of this type of restriction occurred when the
'jpower level was raised_to 0. S'MW on January 23, 1966 after only 3 MWhrs of
total integrated power. 'At that time, system pressure began to rise slowly
 because of a restriction in the vicinity of PCV—522 ‘Shortly after the

r,'start of this pressure transient the reSponse of the drain—tank pressures

-f-indicated an abnormal restriction at a capillary flow restrictor in line

' 521 the fuel-loop—to—drain—tank equalizing line. The excess pressure in

" “the fuel system was relieved by venting gas through HCV-533 to the auxili-

l'aary charcoal bed until the restriction at PCV-522 was cleared by Operating

":Jihthe valve several times through its full stroke.- These operations also

_ revealed at least a partial restriction in the line to the auxiliary char-
71coal bed, apparently at - HCV—533. The reactor was made subcritical after

'}ff4_1/2-hr operation at: 500 kW-;::r

 

Satisfactory pressure control was maintained for about 24 hr, but the
'Aplugging at PCV-522 recurred shortly after the reactor power was ralsed to

 
 

 

 

 

178

1 MW. Again, the operation of this valve through its full stroke was ef-
 fective in relieving at least: part of ‘the restriction. Also on this oc-
casion, evidence of restrictions at the charcoal-bed inlets began to appear.
These restrictions were bypassed by putting the two spare charcoai-bed sec-
tions. (2A and 2B) in service and closing the inlet valves to the two that
were restricted (1A and 1B). Within about 6 hr the inlets of beds 2A and
2B also plugged. When the preViously plugged beds were checked, at least
one was found to be clear and off-gas flow was reestablished through bed

1B. | |
The combination of difficulties in the off-gas system resulted in a
reactor shutdown. to correct these conditions. The restrictions in the
equaliéing line (521) and the auxiliary vent line (533) were relieved by
removing the capillary (FE-521) and a check valve (CV-533, reported earlier).
The throttle valve and filter assembly, and a second valve that had been
tried briefly, were removed from line 522. A large, relatively coarse (50u)
filter and a large, open hand valve were installed at the PCV-522 location.-
This arrangement eliminated the small, easily plugged passages associated
with PCV-522, and it appeared that the filter would still remoVe any parti-
cles large enough to plug valve 522B, which was to be used for system pres-
Sure control. The original valve; filter (pore diam. ¢1u), and the flow
element were subseguently examined at the High Radiation Level Examination
Laboratory (HRLEL). |

Although nothing was done to clear the charcoal beds other than re-—
alignment'of the operating mechanisms to the valves, successive pressure
drop measurements showed that the restrictions in the beds had decreased
significantly and before the start of Run 5, flows through the beds were
hack to normal. Flow measurementsthrough‘the auxiliary bed showed that
the bed had also returned to normal.

Examination of FE-521 — The flow restrictor consisted of a short length
of 0.08-in. ID tubing welded into line 521 at one end and coiled to fit
inside a 3-in. ID container. The other end of the capillary was left free
in the container. When the container was cut open, the entrance region was
found to be clean, and only a small deposit was found on the container wall
near the end of the discharge region as shown in Fig. 8.5. These isotropic

oarticles.were characterized as partly coalesced amber globules with a
: 179

 

 

 

 

 

 

   

   

 

   

 

 

 

7
1

   
 

180

refractive index of 1.520. Spectrochemical analyses indicated microgram
quantities of lithium, beryllium, and zirconium. The restrictor was not
completely plugged when examined in the hot cell.

-,/ Examination of PCV-522 — The stem of the control valve shown in Fig.
8.6A, was covered with an amber oil-like coating and there was an accumu~
lation of oil in the recess formed by the bellows-to-stem adapter. The
tapered flati(flow area) on the stem had small accumulations of solids;
the body was coated with a similar material and had a semisolid mass of
material on the surface near the seat port as shown in Fig. 8.6B. A gamma
scan® ‘of this material indicated the presence of ”Sr, 140p,,. and l"°La..

Examination of the Filter Preceding PCV—522 —-Filter-522 was a l-in.
diameter by 15-1n. long cylindrical type—316 stainless-steel sintered metal
element enclosed by a 1—1/2—in. schedule—40—pipe. The pore diameter of this
particular filter was %lu and the element thickness was 1/16 in. The filter
area'was 0.34 ft* and the'flow was from the outside in. The filter assem-
bly was removed: on.February 8,s1966.and examined at HRLELlon the following
day. The upper_third of_therelenent;was steel gray inrappearance;_the
lower two-thirds hadja frosty white appearance. At no place was there any
evidence of buildnp'or cake,, A day later the frosty white areas had be-
come darker and tended toward thersteel graj appearance. The index of re-~
fraction of some of the material scraped from the sides of the element was,

£ lhoBa’ lkOL

1.524 and a gamma scan‘ of the sample‘indicated the presence o
19%Ru, and **7Cs. |

The filter was then reassembled for flow tests. The data indicate

2,

that for a given pressure drop the filter was passing‘only 5% of the corre-
sponding flow of a clean filter.® However, extrapolation of the data to

the normal MSRE off-gas flow indicates that even though 957 plugged, the
contribution of the filter (0.075 psi measured at the hot cell) to the total
pressure drop (>5 psi during Run 4) was negligible. Exposure to atmosphere
could have opened some of the pores in the filter; however, a more likely

- explanation is that handling and sampling opened enough pores in the ele-

ment to produce these results.
 

 

 

 

 

 

 

 

 

 

181

Fig. 8.6B PCV-522 Valve

 

PHOTO R27812

 

 

 
 

 

 

182

8.4.2 Run 5 o

Satisfactory manuél‘presshre—control by throttling V-522B was demon-
strated during theipreceding shufidown; however, at the beginning of Run 5
after only 2-1/2 hburs of power bperation (1 MW), this valve showed evi-
dence of rapid plugging and required comstant adjfistment until it was in
the fully-open position fifie hours later. During one such adjustment,'
rapid plugging of the charcoal beds occurred. At the conclusion of salt
circulation (terminated by a space-cooler motor failure), the restriction
at the inlet to charcoal bed 1B could not be dlearéd; therefore, the inlet
valve (HV-621) was removed for examination at the HRLEL. With the valve
out, the pressure drop aérossrthe bed was much lower but it was still higher
than for a normal unrestricted bed. An excess of heiium (forward blow) was
forced through the bed and the pressure drop suddenly decreased to normal.

The cone of the inlet valve to this bed (shown in Fig.78.7) was shiny
as though wet with an oil-like material. The small metallic chips near the
large end of the cone were a result of the sawing operation. There was
some white amorphous powder on the tapered section o0f the valve trim and
appeared to be adhered fairly strongly to the metal surface. A similar
material was found in the valve body. The material removed from the stem
was described as an isotrbpic,.faintly colored material, varnishlike in
‘appearance with a pebbly surface. The predominant isotope in the material
was ***Te. The refractive index of the varnishlike material was 1.526 com-
pared with 1.50 for a distilled fraction of the lubricating oil used in the
fuel pump.* ' |

8.4.3 Experiments and Alterations During the March-~1966 Shutdown
Off-gas samples taken while the reactor was shut down showed an in—

creasing hydrocarbon content ifi]the gas as the reactor cell temperature was
- increased, lending support to the'hypothesis that there was a reservoir of
hydrocarbons in the holdup—voluhe portion of the off-gas line. It was not
practicable to clean the‘68-ft-1ong, 4-in.-diam pipe, so the off-gas line
was disconnected at the fuel pump and in the vent house; large quantities
of helium were blown through the line in the forward and reverse directions

at velocities up to 20 times normal.®»” Very little visible material was
 

 

 

 

 

 

 

 

 

 

 

 

 

183

 

 

  

   

 

 

  

 

~ PHOTO R28500

 

 
 

184

collected on filters at the ends of the line, but there were fission prod- &fi#
ucts, and the amount doubled when the cell was heated from 50 to 80°C.
Visual observation showed that the head end of the holdup volume was clean
except for a barely perceptible dustlike film. A thermocouple was attached
to the holdup pipe near the head end for monitoring temperatures during
power operation. (When the power was subsequently raised, the temperature
'indeed rose from cell air temperature of about 55°C at zero power to about
113°C at 7.5_MW; The temperature increase occurred with a time constant of
about 30 min which was not inconsistent with buildup of gaéeous fission
products in the line.) | | |
- As a result of extensive laboratory tests on orgénic vapor traps,’
the filter-valve assembly was replaced with a particle trap and an organic
vapor trap in series and upstream of the pressure control valve (PCV-522)
| whose flow coefficient was increased to 0.865. The Mark I pérticle trap,’
shown in Fig. 8.8 was designed to remove particulates and mist. .Gas from
the fuel-pump bowl entered at the bottom of the unit through a central pipe,
reversed flow in the Yorkmesh, and passed in succession through two con- o
centric cylinders of porous metal (felt metal) and a bed of inorganic fibers. kfig
The first felt metal filter was capable of stopping 96.7Z of particles |
greater than 0.8y and the second — 99.47% particles greater than 0.3p.
The organic vapor trap’ consisted of713 feet of 1-in. sch.-40 stainless-
steel pipe, arranged in three hairpin sections of approximately equal lengths.
The vapor trap was loaded with 1092 gm of PittsBurgh PCB charcoal and instru—
mented with thermocouples at 5, 12, 51, 59, 105, and 113 inches, respectively,
from the bed inlet. |
As a result of the experience with and examination of HV-621 and other
off-gas components, the inlet valves to the main and auxiliary charcoal beds
and V-522A were replaced with ones having larger flow coefficients.

8.4.4 Runs 6 and 7

At the start of Run 6, the pressure drops (at normal flow rate) across
the particle trap, organic vapor trap, and charcoal beds (sections 1A and
1B in parallei service) was <0,05, 0.7, and 1.6 psi respectively. The off-
| gas system performance was satisfactory during.the first ten days of Run 6

with the reactor operating at 1 M{ or less. However, when the power was

Y
   

   
  
  

- FINE METAI‘.L'IC_ FILTER‘

FIBERFRAX

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ORNL-DWG 66-11444R

 

  

 

 

 

 

 

   

 

 

 

 

 

 

COARSE METALLIC FILTER'; E (LONG FIBER)N © -+ . —THERMOCOUPLE (TYP)
’ s A = """"5 3 7z '“'/ flw,s /d’/.@'/fl/////fi'/mflfl //l/l/;/
¥ ||5|'<|a?3 N - 9
i i 2 A
m | ‘ ‘ l ' il i fi E T ey m;: wum e, mr::,;:::_::a
: QIIIIIIhLI I b ! rgwg ,gfi ' 4
b L e ’*r*I?’ifilm
7 g Ao ¥ % rxrat /‘9" {
7 b 1 z Zz2 u”u””u/uu//fl' //fl”.wf// ////// ////4’///#/ /A”//////”./ll//
- ‘ : S oM : o it
~~SAMPLES A LSAMPLE 3A L8
5B AND 6 |

STAINLESS STEEL MESH NICKEL BAFFLE (8)

Fig. 8.8 Line 522 Particle Trap (Mark I) and Location of Sampling
o Points During Subsequent Examination

68T

 

 
 

186

increased to 2.5 MW, the pressure drop across the charcoal beds also in- Qfij
creased. Pressurization and equalization experiments established that the
restrictions were at the inlets to'the’beds,.probably at the packing of »
steel wool above the charcozl.- It,was‘found that the pressfire drop could
be reduced, usually to near the normal 1.5 psi, by blowing helium at 35
psig either forward or backward through the bed; back-blowing seemed to be
more effective. Back-blowing of the beds was done whenever the pressure
drop of two sections in parallel'approéched.B psi. Later, higher pressure
drops were tolerated before baCk-blowing was initiated. Plugging occurred
when thé power was raised during the épproach'to full power with section 1B
plugging more often than the others. Plugging became less frequent later,
but at the end of Run 6, it was still necéssary to back-blow the beds about
once'a week.
Particle Trap Performance. After approximately 10 days of operation
at the 5-MW level, the pressufe'drop across the particle trap started to
increase and reached 6 psi abprbximately a week later. When the powet was
increased to 7.5 MW, the pump pressure increased to 9 psi in less than 12
hours at full power. Since the pressure drop across the organic vapor trap (fiJ
was less than 0.5 psi and that across the charcoal bed was kept below 3 psi
by periodic back-blowing, the variation in pump-bowl pressure was inter-
preted as a cdrresponding restriction in the particle trap. About three
hours after an unscheduled power interruption, the fuel-pump pressure de-
creased to 6.5 psig and the following day it decreased to 4.3 psig. How-
ever, a few hours after the reactor power was again raised to 7.5 MW, the
pump pressure increased to 8 psig indicating a restriction in the particle
trap. At this time the helium purge to the punp shaft was decreased from
2.4 to 1.9 liters per minute in an effort to reduce the fuel-system pres—
 sure. The reactor was then taken subcritical for two days. When the power
was again increased to 5 MW, the pump pressure increased from 3'psig at
zero power to 8 psig and eventually to 12 psig at full power. The pump
pressure gradually returned to the 6 to 8 psig range after approximately
a week at full power. After Run 6 was terminated by a componenf cooling
pump failure on May 28, 1966, the particle trap pressure drop decreased to -
less than 1 psi and it remained relatively unrestricted for a month after
full-power operation was restored in Run 7. The pump pressure then gradu- Qfiyf

ally increased to 8.7 psig before another power reduction (MB-1 failure)
 

187

occurred'ending Run 7. A few hours after the power reduction, the pump

pressure returned to a normal value of 3 psig and remained at this pressure

- for approximately a month of zero-power operation before the restriction

~ gradually returned. '

. In summary, pump pressure increases (restrictions in the particle

traps) were usually associated with periods of power increases. Pressure

decreases were sometimes unexplained but usually were associated with power

.- .decreases or'with'deliberatelattempts-to clear the trap such as reverse-
blowing with helium. - ConEny ) | |

Rerouting of Line 524 Helium, at a rate of 2.4 %/min, was introduced

 

through line 516 into the'fuel—pump shaft annulus just below the lower shaft
oil seal;a:Thejlargerjpart-of.this flow'goes'down the shaft into the pump
bowl thuszpreventing fission gases and salt mist from entering this radi-
‘ation sensitive region of,thé"fiuap. 'The smaller portion of this flow
(<0.1 2/min) goes up the shaft to prevent oil vapors from migrating to the
'fueiusait;'it also aids in”transpOrting any oil seal leakage to the oil
r_catch tank. The driving force;for the smaller flow (line 524) is the pres—
 sure difference between the fuel pump and line 522 downstream of the fuel
- .system pressure control valve (PCV-522). However, when PCV-522 was opened
fully (because of restrictionS'in'the charcoal beds), line 524 had no pres-
sure differential and consequently no gas flow; therefore, it was rerouted
- to enter the off-gas line at V-560A downstreau of the main charcoal bed and
V-557B. V-557B was thenrthrottled to control system pressure. On two oc-
| casions in Run 6, sudden increases in pump pressure (at the completion of
salt recovery from the overflow tank) caused a gas flow reversal at the
pump- shaft and gaseous. activity was carried into line 524. Since this line
-bypassed the charcoal bed the gas flow had very little decay time. As a

*fj;consequence, gaseous- activity triggered -the . closure of the radiation block

-~ valve on the main off—gas line. Therefore in June of 1966 'a small charcoal

 

1bbed’ was added to 1line 524 to: hold “up krypton and xenon for 2-1/2 and 30
days, respectively.r The bed - consisted of 9 ft of 3—in. sched -10 stainless—: -
steel pipe loaded with 15.8 lb of Pittsburgh PCB charcoal On at. least four
: occasions in the subsequent run, fission product activity was blown or dif-
-'fused up the»pump shaftrannuius.i Although the activity level in the oil
'catch'tank-and line 524 increased, there was essentially no activity re-

iease, thus proving the effectiveness of the added charcoal bed. However,

 
 

188

the last two releases from the pump bowl caused the flow element in line o
524 to plug partially and then completely. The plug was fbund to be in
‘the sintered stainless-steel disc at the inlet to the matrix—type flow ele-

ment. The element was replaced with a capillary tube.

8.4.5 Expgriments and Alteration During the June-September Shutdown (1966)

At the conclusion of Run 7, it was decided to flush out ;he'residual
fuel in the overflow tank and to check the indicated level at which over-
fldw occurred. Because of an insufficiency of salt in the drain tank, pres-
surizing gas from the drain tank entered the reactor and floodéd'the pump
bowl with salt, thus binding the pump shaft with frozen salt and pushing
salt in the gas lines at the top of the pump bowl. |

- The frozen salt was cleared from the reference bubbler line with the
use of femotely—applied external heaters; salt in the sampler line was
melted by the same technique but it ran down and refroze at the junction
with the pump bowl. This obstruction and the frozen salt in the énnulfis
around the pump shaft were cleared when the pump bowl was refilled and
heated to 650°C. Although some salt entered the main off-gas line as indi-
cated by a temporary rise in temperature at TE-522-2 (Fig. 8.9), pressure ‘E,)
drop measurements showed no significant difference from the clean condition.
This was attributed to the blast of compressed helium from the drain tank
that was released backward through the off-gas line, before the salt had
time to freeze completely, fihen the overfill triggered an automatic drain.®
Therefore the only work done on the main off-gas line (522) during this
shutdown was to replace the short, flexible "jumper" section of the line,
where the convolutions would be expected to hold some salt.

The particle trap was also removed from line 522 and sent to HRLEL for
detailed examination. An identical replacement (Mark I) was installed to
permit operation while the original was examined and a new one designed and
constructed.

Also during this shutdown, a remotely replaceable heater was designed

and installed on the inlet section of the auxiliary charcoal bed (ACB).

8.4.6 Runs 8, 9, and 10
With the resumption of operations the off-gas flowed freely, with no

unusual pressure drop for 26 days of circulating helium,‘flush_éalt; and
 

 

 

 

 

 

189

- ORNL-DWG 67 - 4764

 

 
  
 

OFFGAS HOLDUP
VOLUME , 4in. DIAM

 

 

 

 

 

 

 

TO PARTICLE TRAP - J
AND CHARCOAL BEDS ==—}h | /

Fig. 8.9 Off-Gas Piping Near Fuel Pump and Overflow Tank

 
 

190

fuel salt at low power. Then, tfio days after power operation was resumed
at 5.8 MW, a plug developed in line 522 somewhere between the pump bowl

and the junction of the overflow tank vent with the 4-in. holdup line. The
first indication was a decrease over a few hours from 107°C to 71°C at TE-
522-2, as the plug caused the off-gas to bypass through the overflow tank
(OFT). The presence of the plug was confirfied when HCV-523 was closed to
build up pressure in the OFT to return salt from the tank to the pump bowl;
pressure in the pump bowl also built up. Efforts to remove the restriction
by applying a 10-psi differential either forward or backward were unsuc-
cessful. ) |

The bypassiné of'off;gasyihfofigh the OFT did not hinder operations
except for one specific job: recovery of salt from the overflow tank.
 Salt sidwly but continuously accumulated in the tank during the entire
operating life of the MSRE, and it was therefore essentiél to return salt
to the pump bowl two or three times a wéek in order to maintain proper
levels. With a plug in the off-gas line of the pump bowl, it was necessary
to greatly reduce helium flows into the pump so that the overflow tank
pressure could be increased faster than that in the pump bowl to make the
salt transfer. Through the remainder of Run 8, salt was returned (burped)
from the OFT six times, and on at least four of these occasions some fission
product activity was blown or diffused up the pump shaft annulus into the
oil collection sp;qe. This was a consequence of the reduced helium purge
down the shaft annulus and the uhavoidable,sudden preésurization of the
pump bowl that occurred at times in the procedure.

After Run 8 was terminated, steps were-taken to clear the plug from
the off~gas line so- that the normal salt recovery proceduré could be used.
Frozen salt was suspected.as the cause of the plug, so the fuel loop was
flushed to reduce radiation levels, the reactor cell was opened, and speci-
ally built electrical heaters were applied to the line between the pump
bowl and the first flange. Heating alone did not clear the plug, but when,
with the line hot, 10 psi was applied backward across the plug, it blew
through. The pfessure drop came down as more helium was blown through un-
til it became indistinguishable from the normal drop in a clean pipe.

In Run 9, the power operation was begun 8 hr after fuel circulation

had commenced, but TE~522-~2 came up to only 66°C, indicating that the line
 

 

 

191

. was again plugged. Whileftools"and procedures were being devised, the re~-
~actor was kept in operation, but great care was taken to avoid getting fis-
“sion products or salt-spray7up the pump shaft annulus again., This entailed
',lowering ‘the power to 10 kW, 24 hr before the overflow tank was to be emp~-
tied, then stopping the pump 4 hr beforehand to let the salt mist settle.
During the salt transfer, the fuel pump was vented through the sampler and
=auxiliary charcoal bed. After three cycles of this, the reactor was drained
and flushed again:in preparation ‘for working on the off—gas line.

This time heat was applied to the short section of line between the
second flange and the top of the 4—1n. decay pipe. 'When heating to about
600°C did not open:the line, the flexible jumper was disconnected to permit
“clearing the obstruction mechanically. In the flange above the 4-in. line,
the 1/2~-in. bore was completel& blocked, but the weight of'a:chisel tool
broke through what appeared to be only a thin crust of ‘salt. 'Borescope
inspection showed that the rest of the vertical line was practically clean,r
and there was only a thin layeriof salt in the bottom of the horizontal rum
of the 4-in. pipe. Helium-was:blown through the line at five times the
normal flow, and the'pressureidrop indicated no restriction. 'The flange
‘near the pump bowl contained;arsimilar plug which was easily broken. A
1/4~in. flexible tool was . then inserted all the way into the pump bowl to
prove that a good-sized passage existed A new jumper line was installed
and operation was resumed., R o |

‘Because obtaining samples remotely without spreading contamination
- would have been most difficult, no analyses were made of the material in
. the flanges.- But it appeared that salt: had frozen in the line almost com-
pletely blocking it during the overfill of July 1966. Material in the |
';foff—gas stream during subsequent operation ‘then plugged the small passages.

‘During the July cleanout the heaters apparently melted the salt out of the
'frpipe, but left behind a thin bridge of salt in the cooler flanges. After
re'the flanges were mechanically cleaned and reassembled “this’ part of the

-doffngas system operated satisfactorily during Run 10.

8. 4 7 MK-I Particle Trap 1~?l**€i‘f' ‘ )
: The second particle trap -served through Runs - 8 9,_and 10., This'unit
" behaved in Runs 8 and 9 much as had’ the first trap; the pressure drop oc-

- casionally built up to 5 to 10 psi, beginning two ‘days after power operations

 
 

 

192

started in Run 8. Back-blowing with helium-was effective in reducing the
pressure drop to 2 to 4 psi; however, in Run 10, after the first week of
power operation, back-blowing became ineffective. Various tacticsa were
used to get fission gas to the particle trap with as little delay as possi-
ble, to see if increasing the fission product;héatifig in the trap would
drive off the material at the restriction. After this proved to be inef-
fiegtive and recognizing that heating caused the central inlet tube to ex-
pand farther into the Yorkmesh filter in the trap, the opposite- approach
was used. Eight hours of delay was obtained by routing the gas through an
equaLizing line to the empty drain tank, through the tank and salt fill
1lines to the other tank where it bubbled through several inches of salt
heel, then out through the drain-tank vent line to the particle trap. The
pressure drop across the trap was 16 psi when the gas was first rerouted,
but within a fewfihours it was below 2 psi. When the original'route was -
again ;ried,.the restriction began to build up almost immediately; there-
fore,.;he delayed route was used until the end of Run 10 which was termi-
nated for in-cell repairs and also because a newly constructed particle
trap (Mark II) based on the Hot Cell examination of the first trap was -
ready for installation. | ) . S

Examination of MR-I Particle Trap.® Flow tests on the first trap (in
ser?icg from April through July of 1966) indicated that (1) the pressure.
drop after service was about 20 times the "clean" pressure drop. .Assuming'
orifice type flow, this would represent a reduction in flow area of about
80%. (2) The bulk of the pressure -drop occurred in the Yorkmesh and Felt-
me;al_sectioqs,of the filter. The Fiberfrax section was essentially clean.

The area of the Yorkmesh which had been immediately below the inlet
pipe was cqvered with‘a blue~gray to black mat which hadrcompletely filled
tfie.spéce between the wires of the mesh (Fig. 8.10). The shapeaof the mat
éorresponded to the bottom of the inlet pipe, and it is likely that this
matérial was the major restriction to gas flow while in service.

Since the inlet pipe temperature probably increased several hundred -
degrees during power operation of the reactor, it is believed that this
restriction behaved similar to a thermal valve. This could account for the
unexpected increase in pressure drop while at power and the decrease in .
pressure drop when the power was reduced or when the off—gas‘was_deléyed |

via the drain tank.
 

 

 

 

 

 

 

 

 Fig. 8.10

193

 

 

 

 

 

 

 

: Ttap Mark I

 
 

 

194

A radiation survey made around the outside surface of the lower section
of the trap gave readings ranging from 6900 to 8000 R/hr. The probe indi-
cated 11;400 R/hr when inserted into the position formerly océupied by the
inlet pipe. The radiation level dropped off to 4100 R/hr at the bottom of
the trap. The bottom of the inlet pipe had a deposit of blackish material
which corresponded to that in the Yorkmesh. The inside of the inlet pipe
at the upper end of the bellows appeared clean and free of deposit. The
eXtétior of the bellows had some of the light-yellow powdery material on
a background of dark brown. = '

When the Yorkmesh was remofied, it was found that the surface of the
outer wires of the mesh bundle was covered with a thin layer of amber-
colored organic.materiél.' Much of fihis material evéporated from the heat
of the floodlamp used to makéfthe photographs. As the bundle was unrolled,
the color of the film on the fiire changed from amber to brown to black near
the center. The black material waé thicker than the wire by a factor of 2
‘or 3. This material was brittle, as was the wire, and much of it came
-léose as the wire was flexed. Samples of this material,’designated Nos.

SB and 6 (Fig. 8.8) were taken for examination and chemical analysis. Met-
éllographic examination of a piece of the wire covered with the black ma-
terial showed that the wire was heavily carburizgd with a continuous net-
work of carbide in the.grain boundéries; There was no evidence of melting
of the wire; however, the grain growth and other changes indicated oper-
ating temperatures of at least 650°C. The nonmetallic deposit observed on
the wire mesh was apparently of a carbonaceous nature and appeared to have
been deposited in layers. These "growth rings'" were probably the result
of off-gas temperature, reactor power, and gas flow-rate changes.

The perforated plate of the coarse filter section and the lower flange
of the filter assembly were covered with a stratified scale (view A-A,

Fig. 8.8). The colors varied from a very light yellow to orange. One
stratum in the lower flange area appeared gray, almost black. A sample

(No. 9) was taken of the light-colored material, including some of the
black material. The perforated plate of the fine filter section was covered
with a8 thin, dark-brown coating, which seemed to be evenly distributed over
the surface of the plate. The inner surface of the outer wall of the trap

was covered with allight—amber coating, which was also evenly distributed. -

s
 

 

| 195

It is believed that_these coatings were deposited by condensation and sput-
:tering.of the oil from the adjacent filter and that the dark brown color
indicates that the porous'metal screen had operated at a'much higher tem—
.perature than had the- outertwall (during operation the trap was immersed
'jin a tank of water for cooling by natural convection) ‘The lower surface
'of the upper flange (view BwB Fig. 8.8) contained a depoeit Which'had the
-tappearance of organic residue.s,The deposit was amber colored, and the frac-
'ftured edge (Fig. 8. 11) gave the impression that the material was brittle.
‘There is as yet no explanation for formation of this deposit or how it came
to be formed in this particular location, The radiation level on the out-
_'side of this section of the filter was 200 to 360 R/hr.
'_ The material at the entrance of the Fiberfrax section: showed an oil-
'- likefidiscoloration, but there was no. evidence of any significant accumula-
Vtion of ‘material. Comparison of the weights of the different layers with
‘the weights of the material originally loaded indicated changes of less
than 0.2 8 An interesting observation relates to the very ‘low radioac-
tivity level of the Fiberfrax at the entrance section which is separated
"only‘by the Feltmetal filter from an area containing material with activity
- levels: of thousands of R/hr._ The only detectable activity (above the exami-
ination cell background of 4.2 R/hr) was, at the discharge end of the Fiber-
frax section. It is probable that this activity resulted from back—blowing
the trap or from pressure transients which could have carried gaseous decay

'products from the vapor trap back upstream and into the particle trap.

| .;Even so, the activity level was only 1.8 R/hr above background.__

 

7 Analytical Results. | A total of four samples from three different lo-
' cations (Fig. 8.8) were subjected to a variety of analytical tests. The

samples were identified as follows. -

_S ple No.-i'j_ ’jEf' ], o " Taken from .
| 3A ..;_lf-r'-';Coarse section of porous metal filter
9 . Scale on lower flange of porous metal section
5B % Mat at inlet to Yorkmesh section
6 ';Li'Mat at inlet to Yorkmesh section -

. For sections of sample 3A it was found that about the same weight loss
(0.2% of sample weight) resulted from heating to 600°C in helium as from
_ dipping in a trichlorethylene bath. The material removed by heating was

 
 

 

196
11 Upper End of Porous Metal Section — MSRE Particle Trap Mark I

8

Fig.

 

 
 
 

 

. 197 Rl s

cold trapped and found to be effectively decontaminated; however, the tri-
chlorethylene wash was contaminated with fission products.
| Samples 5B and 9 were compared for low-temperature volatiles; at 150°C,
No. 9 lost 52 and No. 5B lost none. When raised to 600°C, the weight losses
were 35% for sample 5B and 32% for sample 9. "Analysis of the carbon content
gave none for sample 9 and 97 for sample 58. This indicates that sample 9
'had not reached as high a temperature as had sample 5B. '

The mass spectrographic;analysis»of sample 6 indicated that there was
a very high fraction'of'fission products. These are estimated to be 20
- wt % Ba, 15 wt % Sr, and 0.2 wt Z Y. In the same analysis the salt con-
stituents Be and Zr were estimated to be 0.0l and 0.05 wt % respectively.

In addition~the'materiel-in'samples 5B and 6 contained small quanti-
ties of Cr, Fe, and Ni, while sample 9 did not. The reliability of these
values was compromised by difficulties caused by the presence of organics

and small sections of wire in the sample.

 

‘The gamma-ray spectrographic work indicated therpreSence of the fol-
lowing isotopes: '®7Cs, ®°Sr, *°°Ru or *°°Ru, '*°MAg, °°Nb, and *“°La.

All rhree samples were chemically analyzed for Be, and the level was
below the detectable limit of 0.1l%. Attempts to analyze for Zr were com-
plicated by the presence of large quantities of Sr. '

The fraction of soluble hydrocarbons was determined using CS:, and
the values were 5B, 60%; 6, 73%Z; and 9, 80%. The extract solutions from
ssmples 5B and 6 were allowed to evaporate, and a few milligrams of the

residue was mounted between salt crystals for infrared analysis. The sam-—

' ples were identical and were characteristic 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,cause breakdown of the organic into elemental

carbon and volatiles.
 

198

8.4.8 MK-ITI Particle Trap

As a result of the operating experience with and detailed examination
of the first particle trap, the MK-II trap,® (shown in Fig. 8.12) was de-
 signed with the following features:

- 1. The trap housing was increased from 4 to 6-in. ID, resulting in
an increase in cross-sectional area of 2254 in both the Yorkmesh and Fiber-
~ frax sections, . ,

2. The unit was in effect turned upside down so that the Yorkmesh
section is at the top of the unit and the Fiberfrax section is at the bot-
tom. This change permits heating of the Yorkmesh section (using beta decay
heat -and lowering  the water level) while still maintaining cooling on the
. other two sections. '

3. The disposition of the Yorkmesh was modified to provide increased
frontal area in.the directien normal- to the flow.

4. Since the first trap had shown veryrlittlerloading-infthe,fine
Feltmetal section, only the coarse Feltmetal was used. ,

5. The total filter area was increased from 22 in? for the original
to 288 in® for the new trap.

6. The depth of the Fiberfrzx was reduced by 50%.

7. The pressure drop at 15 2/m was less than 1 in. Hz0 and the trap

.
<

efficiency was 99.9Z for particles greater than 0.8u.

8.4.9 Organic Vapor Trap

It was expected that accumulation of orgenic material in the charcoal
trap immediately downstream of the particle trap would result in progressive
poisoning along the length of the trap. Such poisoning would shift the lo-
‘cation of maximum fission product adsorption and produce a shift in the
temperature profile of the trap. Except for the upward shift due to in-
creased power level, the basic shape of the temperature profile did not
change, indicating that no significant poisoning by organics occurred.

Since the pressure control valve (PCV-522) was operated in thelfully—

)opened position and it appeared that nothing would be lost by eliminating
the charcoal trap, both were removed to meke room for two new MK-II parti-

cle traps which were installed in parallel in January 1967.
 

 

ORNL~DWG 67-4766

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 0 i 2 3

 

INCHES

_ YORKMESH A | L
. /_-T.E. _ /FELTMETAL . N /T.E. . \FIBERFRAX . 'FE.
L Sl 2 AL L L [L T PP P T f
e L — [~
re—— ! Vo T A
e = = et - — - ’ ———— et
B L \\ N Y s/ -
' ’%’
7 RS o ~ ™~ A
2 t rF VY
! —: —— —i e
A. & y 4 ) T
=
/ 7T 77 777 77 777 7
L 1] : -
NICKEL BAFFLES
s,

Fig. 8.12 Line 522 Particle Trap Mark II, Particle Trap Subéssembly

66T

 

 
 

e g e g g et <t e 31 W e 1 3

 

 

 

 

200

8.4.10 Charcoal Bed Performance
After the removal of the check valve poppet in April 1966, venting

through the auxiliary charcoal bed was uneventful through the approach to
power operation. However, an intermittent restrictionfwaS‘nbted early in
July. During the fill and drain after the July 17 shutdown, the restric-
tion became continuous and.more-éevere and fias fidt amengble to back-blowing.
A remotely placeable_assembly of electric heaters.was designed and placed
around the junction of the inlet pipe and steei,wooi-sectionrof the auxili-

ary charcoal -bed. In the brief shutdown between Runs 8 and 9 the level in

the pit was lowered below the bed inlets and the heater was energized.

Some- improvement was observed when the heater temperature reached 385°C.
After the temperature was raised to 670°C and then cooled, the.pressure
drop was downby a factor of 5, to a satisfactory level. The auxiliary bed
operated relatively trouble-free until it became partially restricted in
March 1969 and again about two months later. Back-blowing cleared the bed
in March; a forward-brdw in May was adequate but not as effective. No fur-
ther operational probléms with the auxiliary bed were reported.

Sections 1A and 1B of the main charcoal bed had been used almost ex-
clfisively during earlier runs, and they were on-line when Run 8 started.
The pressure drop at the inlet of these-sectidns began to increase a few

hours after the power was raised<afid when the pressure drop reached 7 psi,

sections 2A and 2B were also brought on line. Since back-blowing had be-

come progressively less effective and in light of the’success with heaters
on the inlet of the auxiliary bed, an attempt was madé to clear the re-
striction in section 1B by heating the inlet with a torgh (electric heaters
to fit this bed were not then available). Althofigh the bed temperature
reached 465°C, there was no improvement and all four sections of the main
bed were operated in pérallel for Run 9. After this run, electric heaters
were installed on the inlets of sections'iA and 1B, Heating to 400°C for

8 hours brought the pressure drop back to the normal range for clean beds.
201

8.5 Subsequent Qperating'Experience with Fuel Off-gas System

_ With the major part of the MSRE operating time yet to come, the off-gas
Vproblems during this period while still present were somewhat ameliorated by
- more efficient particle traps and methods of clearing these restrictions.
Some new restrictions appeared but basically they were similar to those
already described. Remote maintenance methods were used to cope with some

of these.

8. 5.1 Particle Trap
- The MKrII trap (Unit No.nl) was in service from January 1967 to May

of 1969 with no noticeable restriction' however, during the first week in
May, the pressure drop across the trap increased to 0. 4 psi. A week later
when this restriction_increased to 0.75_psi, the reserve_nnit, which had
been valved out -up to thisltime; was put in parallel service. This lowered
the preSsnre drop toVO 1 pSi;or 1ess. “Also temperatnres in the trap re-
flecting fission—gas heating indicated that nearly all of the gas was flow-
'ing through the fresh bed. In October 1969, pressure drop measurements on
~ the off-gas system indicated a restriction of 0.3 psi across trap No. 2;
‘however, cycling the inlet valve; V-522C, between its closed and open po-
sitions cleared the restriction' this indicated a restriction in the valve
T.rather than the trap. Three days later the restriction reappeared and was
dcleared by the same method -No further problems in_this_area were en-

countered

8.5.2 Introduction of Oxygen into Fuel Drain Tank No. 1

| -In_the_event of a_salt spill into the reactor-ceil;;provisions had

d“*heen.made to vent thejreactor:cell ‘to the auxiliarj'charcoallbed through
a line equipped with two block valves. - In January 1967, theservalves,
EIV#571 A and B, were inadvertently opened while the reactor cell was being
. pressure-tested prior to Runlll.,;Reactor-cell air-entered the auxiliary

charcoalgbed which,at,thertinveasibeing used to vent the empty but hot

'.g,reactOr'via'FD;l. The"slowhleakagetof cell air into the ACB resulted in

:”fan increase in fuel system pressure and was incorrectly diagnosed as plug-
'iging in the auxiliary charcoal bed. To clear the bed, it was pressurized
S to 50 psig with helium and released into FD-1 (V-571A was discovered to be

about 1/2 turn open during the back~blowing operation).

 
 

 

 

 

 

 

202

It was estimated that a maximum of 1.25 g~moles of 0; could have been &fij
back-blown into FD-1 which contained approximately 2 cu ft of molten fuel
salt.®® It is believed that the oxygen was later purged out of the drain
tank without appreciable mixing with the fuel system piping. |

Subsequently, line 571 which joined the two systems was cut and both

ends were cap-welded.

8.5.3 Charcoal Beds |

The restrictions at the inlet sections of the main charcoal bed con-
tinued to exhibit chronic plugging throughout the remainder of MSRE power
operations. Sections 1A and 1B began to plug slowly after the two weeks of
power operation in Run 11, After 4 fiéeks of operation the pressure drop
increased from 2.5 to 5 psi and reached 7 psi two months afEér_the start - of
the run. When sections ZA'and-ZB_were;put in service, the pressure drop in-
creased from 2;6 to 9 psi in only ten days. Sections 2A and 2B were cleared
by forward-blowiflg with helium and sections 1A and 1B were cleared with the
previously installed heaters. These operations did not requiré a reduction
in reactor power but only a temporary icwering of the water :level ‘in the _ (ij
charcoal ‘bed pit to allow the heaters to function. Approximately one month
later this operation had to be repeated.

Owing to the success of the heaters on the auxiliary bed and sections
1A and 1B of the main bed, simiiar heaters were installed at the inlets of
sections 2A and 2B during the shutdown between Runs 11 and-12.

Approximately one month after the start of Run 12, it was again neces-
sary to clear all four sectiomns. | |

During MSRE.startup, injections of °®°Kr indicated that the charcoal
bed hold~up time was 5-1/2 days for krypton with two sections in parallel
servicej the equivalent xenon. hold-up time was caiculated to be 88 days.
Thermocouples are_located_at the entrance regions of the 1~1/2-, 3-, and
6-inch diameter sections of the charcoal. beds. Thesé sections contain 4.75,
27.4 and 67.85 percent, respectively, of the total charcoal in the beds.

On the basis of a2.5.5 day hold-up time for krypton, the lag in temperature

response to fission gas adsorption between the 1-1/2— and 3~inch diameter

sections should be 6.3 hours (4.75% of 5.5 days) when two beds are:in ser-

vice and 3.1:houré'when,®ne;bed is in service. During full-power operation &;fi
 

 

203

on September 15, 1967, the temperature response interval for beds 1A and
1B was 11-1/2 hours. During a similar startup on September 21, 1967, the
response interval was 9—1/2 hours for bed 1A and 11-1/2 hours for bed 1B
indicating that bed 1B had developed more of a restriction than bed 1A.
During the startup of November 25, 1967 ‘with beds 2A and 2B in service, the
response interval. was 8nl/2 hours for bed 2A and no response in bed 2B in-
dicating an appreciable restriction in bed 2B. Beds 2A and 2B were then
valved off and bed 1A was put in service; the temperature response interval
- was 6*1/2 hours. Bed lA remained in service for the remainder of Run 14
(130 days) with the exception of a 10—day period in the middle of February
* when two sections were put in service to achieve the low fuel—pump pressure
- required for a ,reactivity test. The maximum pressure‘drop across the single
section'during Run 14rwas 4. 9 psi. No discernible‘increase'in effluent
activity was detected by the radiation monitors downstream of the charcoal
'trap during the single bed test.
~ No plugging in the bed'was experienced during the precritical and low
_m_power (<25 kW) operation using 233y fuel (Runs 15 and 16). However, during
"the next runs, gradual plugging recurred after the first 12 days of power
'operation (5 MW). The electric heaters, previously installed on the inlet |

section, were used to clear the beds, although two of the three heater ele-

.b-ments installed on sections lA and lB and one element on 2B failed This

procedure along with a back—blow at the end of the 8-hour heating cycle was
N required to clear the beds approximately three months later. Unfortunately,
during the valving operations involved the stem or stem extension on the

b inlet valve to section 2A broke with the valve in the closed position. Re-
”__pairs were not attempted because_of the high radiation level at the valve

o and the sufficiency of the other three sections.

During portions of the next run (19) argon was substituted as a cover

- gas to study its effect on core void fraction and reactivity changes. No

mdfiunusual off—gas problems associated with argon were manifest other than
.'interpretation of flow and pressure drops through the system during the
.'several switchovers from one gas ‘to the other. A barely perceptible in-
crease : in the effluent stack iodine indicator (attributed to mneutron acti- -

:vation of argon in the . reactor) was noticed Duringrthis run, it was neces-

, sary.to ‘clear the charcoal beds twice, using the inlet heater and a

 
 

 

 

204

forward-blow with helium at 25 psig. No further problems were encountered
with the bed during the last run.

8.5.4 Restriction at the Pump'Bowl
The restriction in the off-gas line at the pump bowl reappeared in

November 1967, one year since the-line was last cleared with a flexible
reamer. The restriction became noticeable approximately 2-1/2 days after
the reactor power was reduced to 10 kW to repair the sampler-enricher.
During this period the operational and maintenance valves weie bpen, and

a 0.6-1iter/min helium purge was maihtained down the sampler tube to the
pump bowl to prevent contamination of the sampler by fission gases. The
restriction was evidenced by an increase (0.5 psig) in the fuel-pump bowl
'pressure when the overflow tank vent valve was closed during a rbutine
return of salt from the overflow tank to the pump bowl. After the sampler
was repaired, the restriction was relieved by pressurizing the pump bowl to
6.0 psig and suddenly venting the gas into a drain tank which was at 3 psig.
The pressure drop then appeared notmal (<0.l1l psi), but after three more
days of low-power opération, an abnormal pressure rise (0.2 psi) was again
observed when salt was being retfirned from the overflow tank. Repetition
of the mild blow-through to the drain tank had little or no effect this
time. The line was not completely blogked, however, and full;power-opéra—
tion was resumed. At first, temperatures on the overflow tank and the of f-
gas line (responding to fission product heating) clearly indicated that

" there was enough pressure drop at the pump-bowl outlet to cause much of the
off-gas to bubble through the overflow tank and out its vent line. Then,
~ after 15 hr at full power, the'bypéss flow stopped, indicating that the re-
striction had decreased significantly; it remained below the limits of de-
tection throughout the next nine weeks while the reactor was opérating'at
7.2 MW or 5 MW. After two days ét 10 kW, the pressure drdp again became
detectable and continued to increase over the next six days at very low |
power. When power operation was resumed at 5 MW, temperatures indicafed
that thexe'was again bypass flow through the overflow tank. The restric-
tion increased during two weeks of operation at 5 MW, but the line never
became completely plugged. Four days after the resumption of full-power

~operation in February 1968 while the overflow tank was being emptied, the

U
 

~ stream was suspended in the entrance of the 4-in. off-gas hold-up pipe.

205 -

restriction partially blew out, bringing the'pressure drop again below the
- 1limit of detection. The. ‘pressure drop remained just at or below the limit
| until the March-1968 shutdown for fuel processing. N

~ The behavior was quite”unlike the plugging that occurred earlier in
the same section of line as a result of the overfill.with flush salt, main-
ly in that it never completely plugged and there was a suggestion of some
effect of power level. |
| After the fuel was drained the flexible section of off—gas line was
removed. and_a felxible_tool_was-run through the -28-in. of line back to the
,pump bowl. The'tools,jal/4-in.lcable_with a diamond drill tip, encountered
' some,resistance at first and came out with a considerable amount of solids .
on it. Although.the flexible Jumper and the line downstream were not re-
- stricted, the'jumper line along.with the flexible reaming tools were sent
to the HighflRadiation—Level—Examination Laboratory for cut-up and exami-
,nation., , o l, , , |
| The replacement Jumper. line was equipped with four thermocouples to
provide more information during power operation. A slender basket contain-

ing devices of metal and graphite to collect material from the off-gas

The restriction at thelpump'bowl reappeared-afterflapproximately one
- month of operation with the new fuel in September 1968. The restriction
- was blown clear once during pressure-drop measurements, but reappeared a
week later and could not be dislodged by back—blowing-helium from the drain
tank to the pump bowl. The pressure drop across the restriction had in-
'gcreased to 4 psi by the last week in November 1968, when the fuel and" cool-

ant systems were drained to clean out the off-gas lines of both systems and

 

”:also to ‘mix the. fuel before ‘the. start of full-power operation.

. The flange at the fuel—pump bowl was again disconnected, and a flexi-

:;ble cable assembly, connected to a filter and a vacuum pump, was used to

f-ream out the off-gas 1ine to the fuel pump -and . also to collect ‘a sample of
- the restricting material Only a very small amount of blackish dust was
3frecovered on the filter paper. and ‘from the flexible tubing. The flange
was reconnected and the reactor was refilled however, ‘after about three
'_.weeks of operatien, the restriction reappeared in mid—January and gradually

increased to a AP of 5.5 psi. Because the rapid recurrence of the restriction
 

 

 

206

at the pump-bowl exit threatened to become a serious nuisance, a compact
2-kW heater unit was designed for installation on this line at the first
opportunity; however, in the last week in February during a routine trans-
fer of fuel from the overflow tank to the fuel pump, the restriction was
blown clear, resulting in a complete transfer of fuel from the overflow
tank to the pump bowl. The pressure drop in this section of line was then
below the limit of detection (about 0.1 psi).

On five other occasions in the next 9 weeks the pressure drop became
noticeable, but each time it blew out when the overflow tank was burped.

On May 1 it reappeared and remained detectable through the June-1969 shut-
down. The restriction caused most of the pump-bowl purge gas flow to bub-
ble through the overflow tank except when the overflbw tank was being pres-
surized to push accumulated salt back to the pump bowl. As a consequence,
on May 25, the off-gas line from the overflow tank became almost completely
plugged and it became necessary to turn off the overflow tank bubbler flow
most of the time. Thereafter all the pump purge gas was forced out through
the off-gas line from the pump bowl. This situation caused some inconveni-
ence during the burping operation, but by reducing purge flow to the pump
bowl at these times, salt could be transferred back to the fuel pump with-
- out exceeding 15 psig in the pump bowl.

During the June shutdown, the 2-kW heater that had been built earlier
was installed remotely on the pipe as near to the pump as possible and con-
nected to spare power and thermocouple leads in the reactor cell. The pump-
tank furnace heaters and the new heater were turned up to bring the section
of off-gas line to near 650°C, then gas pressure was applied to blow the
‘restricting material back toward the pump bowl. The heater was subsequent-
ly turned off but left connected so that it could be used again without re-
opening the reactor cell. ' -

The set of specimens exposed in the off-gas hold-up volume since July
1968 was also removed during the June shutdown. ‘

Although the restriction did not reappear during Run 19, the pressure
modifier and recorder which was installed on the fuel=-pump preséure trans-

mitter to detect high frequency pressure noise in the reactor core indi-

cated that a restriction at the pump exit was again in its formative stage.'®
 

 

 

- 207

The restriction did not become evident (by previous standards) until ap-

proximately two weeks after .the fuel fill of Run 20. In the three days
before final reactor shutdown, the restriction increased from 1.3 to 2.2

psi as measured with the procedure for salt recovery from the overflow tank.

- Because of the proximity of the reactor shutdown date and the desirability

of not interrupting theiexperiment in progress, the heater which had previ-

ously been installed and successfully used to clear a similar restriction,

was not used. | h . |
Examination of the 522 Jumper Line Removed March 1968.'' An 8-mg sam-

ple of what appeared to be soot was collected from the upstream flange of

- the flexible jumper line; a'15-mg sample of a similar‘material was collected

from the downstream flange. The 8—mg sample indicated about 80 R/hr and the
15-mg sample indicated about 200 R/hr at “contact." Each of the flanges and

'the convolutions of the flexible hose appeared to have remaining on them a

thin dull-black film. A one—inch 'section cut from the upstream end of the

| jumper line indicated 150 R/hr; a similar section from the downstream end
indicated 350 R/hr.

The reaming tools used inlclearing the restriction at‘the pump  bowl

- were covered with blackish, pasty, granular material which was identified

by x—ray as fuel salt particles. The relatively high values of lithium
and beryllium found in this sample may indicate pickup of scme residual

| Lin-Bng flush salt which had entered the off—gas line as a result of the

pump overfill in 1966. _ ,
Electron photomicrographs of the upstream dust showed relatively solid

particles of 1 surrounded by a material of lighter and different structure

'which appeared to be amorphous carbon., Chemical analyses of the dust sam-

ples showed 12 to 164 carbon, 28 to 54% fuel salt and 4% structural metals.

~_Based. on activity data, fission products could have amounted to 2 to 3% of

the sample weight. The unaccountability of the remainder is attributable

“to the small amount of ‘sample’ available.

From the dust collected from 1ine 522 onto a filter during the December-

12

1968 shutdown, the following conclusions were reached: (l) the isotopes

333y, '“%Ba, '*“Ce, d;?’Zr:were transported only as salt constituents;

(2) the isotopes °°Sr, °'Y, and '*®7Cs, which have noble gas precursors, were

 
e s e e

 

208

present in significantly greater proportions consistent'with’a mode of
transport other than salt particles; (3) the isotopes °*Nb, °°Mo, *°°Ru,

and '?*MTe were present in even greater proportions, indicating that they
were transported more vigorously than fuel salt, ' ' o

A more complete discussion of these findings and the results of exami-

nation of the sample specimens removed from the éwint hold-up volume will
be covered in a report to be issued by Compere and Bohlmann entitled "Fis-
sion Product Behavior in the MSRE During 2®°Uranium Operation" (ORNL—TM-
2753).

8.5.5 Other Restrictions in Line 522 7

On five occasions during Runs 17 and 18, complete flow blockage oc-
curred in the main off-gas line somewhere between thezé-in.édiam hold-up
pipe and the'junction-of Line 522 and the auxiliary charooal.beddline, 533.
Although ‘the plug could have occurred at the inlet port of V—522A it is
unlikely that the plug formed in the valve itself because manipulating the
valve stem to its full travel in both directions did not break the plug
loose. 1In each case the fuel system pressure was vented for short. periods .‘ii}
of time through the equalizer and drain tank vent lines to the auxiliary
charcoal bed until 1line 522 was unplugged by back—flowing helium from line
571;:through lines 561, 533, and 522 to the hold-up volume. It Seems that
each plug became successively more difficult to dislodge. To dislodge‘the
last plug, it was necessary to vent the fuel»Sjstem to 2 psig, pressurize
line 533 to 60 psig, and quickly open V-522A. The last effort apparently

‘restored the line to near its original condition because plugging at this

location had not recurred at the end of reactor operations. Because of the

remoteness and inaccessiblity of this line, no further effort was made to

locate the exact position or characterize the nature of the plug.

8. 5.6 Restriction in the Overflow Tank Exit Line

The fuel pump was provided with an overflow pipe. (shown in Fig 8 9)

~and a catch tank (5.4 ft®) to prevent_salt-from becoming high enough to al-

low it to enter the gas and lubricating passages of the pump as a result

of unusually high gas-entrainment.in{the circulating sglt or possibly as a

result of a high-temperature excursion while operating. The overflow pipe B
from the pump dips to within less than 1/2—in. of the bottom and into a fifij
 

 

209

1/2-in.-deep dimple pressed in the lower tank head to permit more complete
retrieval of its contents. During operations, salt spray and/or bubbles
slowly spilled into the overflow tank (OFT); the salt was periodically re-
turned to the pump bowl by closing the OFT gas discharge valve and allowing
helium from the bubble-type level indicators to build up the required pres-
sure (3. 5 psi > pump bowl pressure) to force the salt back to the pump bowl.
“When the OFT discharge is opened, the salt remaining in the tank and that
remaining in the pipe is Sufficient to provide a seal.at.the bottom of the
tank and thus prevent-fission_product;gases from'entering the overflow tank.
However, when the main_off-gas line, 522, became restricted near the pump
'bowl; practically all of the pump purge gas and gaseous fission products
were bubbled through the salt heel in the OFT and out through line 523 to
join the main off-gas line downstream of TE—522 2. From TE-522-2 data and
pressure measurements during salt recovery operations, it was estimated

that fission.product gases were first diverted through the OFT on.October 15,
1966. - Since the fuel—pump overfill on July 24, 1966, it is estimated that '
‘most - of the fuel—pump gases were forced through the overflow tank for ap-
proximately 28 days in 1966 6 days in 1967, 76 days in 1968 and 68 days
in 1969, before the overflow tank gas exit line (523) became plugged on
-rMay 25. Thereafter, all of the,pump purge -gas was forced.out through the
. restriction inlthe off-gasilinerat-the exit of the.punp bowl. This situ-
ation caused some inconvenience duringrthe salt recovery operation,'but by
oreducing-purge flow to the:pump atothese times, the operation could be done
-without exceeding 15 psig in- the pump bowl. | |

| During the shutdown in June 1969 when the overflow tank vent line was

 scanned with the remote: gammaflspectrometer,_an.unusually strong source was
Tobserved at the air;operatedrvalve:(HCV;523), about 33'ftudonnstream from
;-the'overflow tank. ?langes'fiere'opened~and pressure observations showed
that the restriction was in the flanged section containing both the air-
'operated valve and a hand valve. This section was removed to a hot cell
where polymerized hydrocarbons and fission products were found to be block-
h'ing the hand valve inlet port. A replacement section containing,an,air—

‘operated valve but no hand:valve;was:installed;_ Although the restriction

: atrthe exit of the pump bowl reappeared during the last week of MSRE opera-
tion, the OFT discharge line functioned normally.

 
 

210

8.5.7 Restrictions at Exits of Drain Tanks

During the reactor drain in June of 1969, pressure measurements indi—
cated that a restriction had developed in the gas line somewhere between
Fuel Drain Tank No. 2 (FD-2) and the junction of the inlet line (574) and
the exit line (575). The first half of the drain appeared normal at a drain
rate of v2 cu ft of salt per minute after which it tapered off to m0,25 cu
ft per minute when the trapped gas in the drain tank balanced the salt in
the fuel system. Two and a'half hours were'required for the drain as op-
posed to a normal drain time of less than 40 minutes when drained into one
tank. The restriction was later cleared by heating the tank to 680°C and
applying a pressure‘differential of 60 psig across the plug via line 561.
A similar but lesser restriction in'the’gaszline at the.other fuel drain
‘tank was partially cleared during the fuel system pressure'test in'Auéust
1969 by heating the tank to 680°C and flowing helium from FD-2 (at 50 psig)
to FD-1 (at 2 psig). ,

Restrictions in these lines were most probably deposited (but not to
the poiut of detection) durihg December of 1966 and January of 1967 when
the reactor off-gas was routed to FD-1, through FV-106 and FVAIOS,Vthrough
the salt heel in FD-2, and out through line 575 (Sect. 5.6.1). Fission gas
decay heating, during this time, was appreciable in the drain tanks and re-
quired downward adjustments to the drain tank heaters. The offegas £low
through the drain tanks was reversed after the first week in an effort to
minimize plugging. - . | | ,

The off-gas flow was again routed through drain tank No. 2 gas lines
for a day when line 522 plugged in May.of 1969. |

In October of 1969 during a pressure release test When the fuel—pump

3

gas was released into FD-1, it was discovered that the gas line was again
restricted near the drain tank. Two attempts were made to clear this line
by back-blowing helium, first at 30 psig and then at 50 psig, from line 561
through HCV-573 and into FD-1. Although only marginallimprovement was ob-
~ tained, the gas flow through this line was judged adequate for any emer-
gency drain situation.

‘During the subsequent drain in November 1969, approximately 4300 1b of
salt drained into FD-1 and 5100 lbs drained into FD-2 iudicating that a
slight restriction existed in the exit gasvline of FD-1. During the .
211

December -drain, the‘exit,gas valve from FD-2 was closed during part of the.
drain to compensate forrthe partial restriction in FD-1 gas line; however,
overcompensation resulted and 5400 1bs drained into FD-1 and 4100 1bs

drained into FD-2. The drain times for both drains were less than 20 min-

utes which is normal for a drain when both tanks are used.

8.5.8 Restrictions at,therbffégas Sampler 7

A system to permit the analysis of the reactor offegas stream was in-
stalled downstream of the particle trap as shown in Fig. 8.2. The sampler
contained two thermaleconductivitycells, a copper oxide converter, and two
molecular Sieves,'one operating atfiliquid nitrogen temperature and the other
at room temperature.® Since the sampler wasran integral part of primary
containment during samplingboperations, and since some components of the
| sampler did not meet the requirements of primary containment solenoid
- block valves were installed in the inlet and outlet lines which connect -
the sampler to the reactor .system. ‘Two fail—closed valves in series were
-installed in each line and were instrumented to close on high sampler ac—~
| tivity, high.reactor'cell-pressure, and high fuel-pump pressure.
| , Although located downstream'of the particle trap, the inlet line to
the off-gas sampler periodically developed a restrictionrin the vicinity
of'the'safety block. valves. The block valves have 3/32—in.-diam ports and
the inlet piping is 0. 083 in ID autoclave tubing. The restriction was suc-
cessfully cleared each time by back—blowing with helium, Aslencountered
in other parts of the off—gas system, successively higher pressures were
_required to clear the restriction each time. During the latter part of
Run 18, the inlet block valves would not shut of f tightly and were replaced
in July of 1969. Visual. inspection of the faulty inlet valves did not re-
'veal the .reason for the. leaking valves nor- the nature of ‘the restriction
"(the restriction had been blown: clear before the valves were -replaced).
_During the August startup, 1t was again necessary to blow out the restric—

tion at the inlet to the sampler.-

 
 

212
8.6 Discussion and Conclusions

The periodic plugging in the coolant system off-gas filter and also
in the fuel system filter during the precritical and low-power (<25 kW)
operation can be attributed primarily to the accumulation of oil in the
0.7 to 1lu diameter pores of the sintered metal filters (2 to 4p pores might
have been a better choice). However, subsequent filters inrthe fuel system
were constructed of felt metal (Huyck Nos. FM-225 and FM-204) which was not
only efficient at stopping solid fission products but also was apparently
immune to plugging by hydrocarbons. Virtually all of the solid decay daugh-
‘ters were stopped by the Yorkmesh and filter before reaching the Fiberfrax
section of the particle trap.

0il vapor in the off-gas stream condensed on various parts of the off-
gas system components and apparently enhanced the adsorption or trapping of
particulate matter, particularly on the Yorkmesh portion of the particle
trap. | | 7

The fact that such large amounts of solid decay daughters were trapped
in the particle trap indicates that most of these solids tend to remain in
the gas phase until trapped onto a surface wetted by hydrocarbons.f It is
not clear whether the solid decay daughters agglomerate in the hold-up vol-
ume and if so whether they agglomerate with others of their own species,
with other species, with hydrocarbons, or‘possibly combine chemically to
form both volatile and non-volatile compounds.r A large fraction of the
noble gas fission products decay in the 4-inch-diameter hold-up volume and
their decay daughters apparently plate out or are adsorbed onto the first
flow channel restriction where the gas flow changes'direction"andlor is
close to a relatively cold surface such as valves and floufrestrictors;
| The high hydrocarbon content (>60%Z) of the material collected on the
Yorkmeshris inconsistent*with.the high temperature (¢650°C) in this region
| during power operation; hydrocarbons are vaporized and cracked at this tem-
.perature. One must assume, then, that the hydrocarbons were collected
sometime after power operation had ceased and after the fission product de-
cay heat became negligible. The growth rings on the Yorkmesh were probably

caused by alternate periods of power and zero-power operation.
 

 

213

Since spectrographic analysis did not show any of the alkali metals,
cesium or rubidium, on the Yorkmesh, they were apparently boiled off also
by fission product decay heating in this region. If this be the case, ap-
preciable quantities of these materials would be expected to be collected
in the charcoal beds, On the other hand,; large quantities of barium and
strontium were reported'on'the Yorkmesh; this would argue for effective
trapping of their precursors, cesium and ribidium. However, to be trapped
on Yorkmesfi.at 650°C, thelalkali:metals would necessarily have to be chemi-
cally combined (as halides for instance) rather than in the elemental form,
otherwise they would boil off and be carried downstream.

_ Although none of the solid decay daughters in the 4-in.-diam pipe got
past the particle trap, another batch (though somewhat smaller number) of
solid decay daughters are born (and presumably remain in the gas phase) in
the larger hold-up volume ahead of. the main charcoal beds. These fission
products along with hydrocetbdné-are;probably the source of the plugging
experienced at the entrance'region‘of—the beds.

After examination of the MK-I particle trap, it is understandable why
the entrance region of the charcoal beds plugged periodically; the entrance
pipe (1/4-in..schedf40)'enters the bed normal to the section where the
stainless steel wool is packed, thus the cross-sectional flow area (trap-
ping area) at the entrance to the stainless steel wool is only 0.1 sq. in.
| Although a particle trap was not installed between the second hold-up
volume and the charcoal bed, it was considered and was definitely needed.
Also the entrance section of future beds should be redesigned such that gas
enters the empty chamber above the steel wool and thus offer a larger cross
section of stainless steelrwool to gas flow as in the MK-II trap..

i 'Anothef argfiment.that'fiSSionlpfodficts, as well as hjdrocarbbhs, are
- involved in the plugging mechanism at- the charcoal bed entrance, is the fact
'that the beds remained free of plugging until power operations were begun.
- For a better understanding of the plugging mechanism, the MR-TI particle
trap and the entrance seetiOn_tO'one of the charcoal?beds would need to be
ekemieed. Also at least one bed could be examined along its length to de~
~termine the fission‘produet'adsorption characteristics of the bed.

The probable reason that the'ovetflow tank exit line remained clear
for such a long time was the fact that pressure differentials (&3.5 psi)

 
 

214

were periodically released through the line each:time salt was returned to
- the pump bowl. The pressure pulses probably cleared any incipient plug
-which may have been in its formative stage. -

The restriction at the fuel pump bowl does not seem to be related to
fission products since the plug seems to form as readily if not more:so
while subcritical and at low power as it does at full power. The gas exit
line from the pump bowl should be modified to eliminate mist or liquid car-
‘ryover into the off-gas stream. A heated cyclone~type separator’? or a
-large well baffled and cooled region which could be later heated to melt
down any salt formation are two possibilities. o

I1f hydrocarbons had not.beenrpresent°in_the pump bowl, fission gas be-
‘havior in the MSRE might have been somewhat different from that experienced.
;SometspeculationS'on-probable results are (1) less serious overall plugging
because there would be no semisolid varnish-like buildup due to hydrocar-
bons, (2) the York mesh would probably be less efficient at trapping the
solid decay daughters and more of this material would then be trapped on
the felt metal filter, (3) the fission product "cake' would probably be
1oose1y-packed on the filter and more easily disrupted by backflow since

it would not contain a "binder" or paste material (hydrocarbons).
 

 

 

 

215

9. FUEL AND COOLANT PUMP LUBE OIL SYSTEMS
| J. K. Franzreb

9.1 Description

Two identical oil systems served tolubricate and-cool-the fuel and
coolant pump bearings and to cool the shield plugs located between the
bearings and the bowls of the pumps. Each was a closed loop designed to
meet containment requirements, -The oil used was Gulf-Spin 35.

Each system consisted of two 5-hp, 60-gpm at 160-ft head, 3500-rpm
Allis Chalmers Electri-Cand pumps, (one normally in operation with the other
in standby), an in-line Cuno. EFS oil filter, and an oil tank of 22-gal op-
erating capacity- hav1ng "brazed on" water coils cap&ble of removing 41, QC0
Btu/hr. | | | .

The oil flow to the bearings of each of the salt pumps was sabout 4 gpm.
- with about 8 gpm to each shield plug. The remaining 48igpm was recycled
through the oil tank to aid in cooling. A scavenging jefi_was installed in
the o0il lines near the salt pump so that the shield plug oil flow aided the
return of the bearing oil to the storage tank,

Lubricating oil seeping past the lower shaft seal of the salt pump was
piped to the oil eatch tank  for measuring the rate of leakage.

The pumps, Storage'tank,'filter and much of the instrumentation and
valving for each system were mounted in an angle iron frame to facilitate
moving to the site, This was'COmmohly referred to as a lube-oil package.
The two packages were interconnected 80 that either could.be used in an

~emergency to supply both salt,pumps.

-

9.2. TInstallation and Barly Problems

- Before installation at the MSRE, both lube-o0il packages were operated
."in'a'test stand., Heafi‘loadiafid'preSSure drop data were obtained and the

- performance of the sysfiems;was;eheeked.' A problem of gas entfainment_wae

- found which affected the'prifiingzof the standby pumps. Approximately
30 seconds were requiredxto,fiiime'a gpare pump if it had'notrbeen.operated

for several hours. Priming time was reduced to 5 sec by installing gas

 
 

 

216

vents from the pump volute casing and the pump discharge line to the gas
space in the oil tank. This gxperiencé led to é modus operendi of starting
the spare pump of each packagé once per shift and running it for about

15 min. . S :

The packages were installed at the MSRE in 1964, Initial operation
-was hampered by failure of the stator insulation in one of the pump motors
and by low resistance to ground in stators of others. Since moisture was
suspected, a potting compound (Dow Corning Silastic RTV-T31) was used to
seal the motor hofising Joints and & moisture resistant coating of paint
(Sherwin Williems epoxy white B69W6) was applied to the exterior surfaces
of all four operating motors, plus two spares. , ,

‘The loss of prime of the standby pumps continued to be a problem until:
’_the scavenging jets used to return oil to the reservoirs from the salt pumps
- were replaced with jets of lower capacity to reduce the entrainment of gas
- in the return oil streams. These replacements were done in September 1965,
After modifications, it was possible to reduce the frequency of priming of
- the standby pumps to once a week.

Becsuse of the lowered capacity of the new return Jets, it was found
necessary to limit the flow of oil to the fuel-pump motor to 4 gpm, as a
flow of 5 gpm would result in a buildup of o0il in the salt pump motor cavity.

'

9.3 Addition of Syphon Tanks to the Qil Catch Tanks

The oil catch tanks were febricated of & L6-1/2-in.-long section of

" 2=-in. pipe topped by a 20-in;—longrsectibn of 8-in. pipe; The lower sec-
tion allowed accurate measurement of the oil accfimnlationjrate whereas the
upper portion provided sufficient volume to handle possibie gross:le&kage.
The catch tanks were drained periodically to keep the.level in the lower.
section. Radiation would not permit doing this during power operation and
since the volume of the lower section was around 2500 cc and the allowable.
seal lesk rate was 100 cc/day, it was possible that the reactor would have
_to be shut down to drain the oil catch tanks during extended runs. There-
fore, in October-1965, equipment was installed to automatically siphon .the
_0il from the catch tanks when they became full.
 

 

217

After instellation, tests were run on these and they performed well.

- However, they failed to function properly at the low oil leaskage rates that

actually occurred during operation. The oil flowed over the high point of
the syphon tubes, as over a weir, without bridging the tube to form a sy- -

phon. We therefore reverted to manfialrdraining during shutdowns. Fortu-

nately the leak rates did not get hlgh enough to interfere with operation

of the reactor,
9.4 0il Leakage
Continuous records of‘oilntankrandwoil catch tank levels were main- °

tained during all of the MSRE'operoting life to determine what losses were

takifig place. Samples removed from the sysfems were'carefully measured,

as were additions. ‘The oil storage tanks were large and therefore small
. changes in level 1ndicatlon caused large errors in 1nventory. Successive

‘ log-readings could vary by 500 to 600 cc. Fairly accurate indication of '

the leskage through the rotating oil seals of the salt pumps to the oil

. seals of the salt.pumps to the oil catch tanks were possible over one month

or longer periods. The lesk rates during operation varied from a few cc

. -per day for either salt pump to around 25 cc/day. The seals seemed to get

"worse or improve for no apparent reason.

For the 24-month period through August 1969, inventorles showed unac-

counted for losses from the fuel pump oil system of 5.4 F 1.5

—3.0
the coolant o:l.l system of. D 6 4% (5) liters or a.vera.ge da.lly losses of 7.5 cc

liters; from

- (5.25 gms), and T. 8 cc (5.4 gms) respectivelyu Anelyses indicated that

. there were approximately 1 to 2 gms of oil products per day in the fuel
_;pump off-gas stream. ‘Some oil- nmywhave been held up in the lines, 6--f“t3
':Fholdup volume, and. partlcle traps that were upstream of this sampllng point.

In conclu51on, the best estimates showed an oil loss of 5+ gms/day,

,'of which only l-2 gms could be found as hydrocarbons in samples of off-

gas from the primary system,_;e&v1ng approxlmately:3 gms/day unaccounted

'?efor.

 
 

 

218
9.5 Change-Out of Oil Pumps .

‘After the initiel trouble with the electrical insulation due to mois-
ture, the pumps gave very satisfactory service. Two pumpS'fiere removed

from service, one because of excessive vibration and the other because of

‘an electrical short in the motor winding. This was done during the six-

month period ending August 31, 1967. 7

The pump with excessive vibration had one of the two balancing disks
loose on the shaft. This loose disk was reattached, and the shaft assembly
was dynamically balanced. The pump was reassembled, tested under operating
conditions and made ready for service. The pump with the shorted winding

was rewound and put back into service.

9.6 Test Check of One 0il System B
- Supplying Both the Fuel and Coolant Salt Pumps

~ On March 20, 1966, the two oil systems were valved so that the fuel
oil package (FOP-2 running) supplied oil to both.saltrpumps; This was done
to check calculations; and to be sure that this could indeed be done in
some future emergency. The coolant pump was circulating salt at 1200°F;
the fuel pump was idle and at a‘temperaturé of 125°F, Adequate flows to
both salt pumps were maintained (3.35 gpm to the bearings and 6.5 and 7.3
gpm to the coolant and fuel pump shield plugs). Under these conditions the
oil supply temperature equilibrated at 127°F. When the fuel pump oil flow
was stppped,_thé oil supply temperature increased ~ T°F, indicafing that

 the fuel pump was removing part of the heat load of the oil system.

Although this test didAnét,prove conclusively that one o0il cooling
system was adequate for an emergency wherein one package would be used to
supply both_operating salt pumps, subsequent heat exchanger tests and cal-
culations indicated that one oil cooling system would be adequate,’

9.7 0il Temperature Problems

The total oil flow from one lube oil pump was normally 60 gpm. Of
thig,.B gpm_was directed to the shield plug of the salt pump in the par-
ticular system, 4 gpm to the salt pump bearings, and the balance was by-

passed back into the oil tanks and caused to flow down khe outer tank walls.
 

219

This amount plus the warm oil from the shield and bearings was cooled by
means of water flow through coils_brazed_to the outside of_theloarbon steel
tanks. | . _ o o o | \

As the system was eubjected to extended service, scale built up on the
inside surfaces of these coils, causing a drop in_the cooling capacity.
This in turn caused the temperature of the supply oil-to climb from its
normal 134°F to 140°F. This higher temperature was reached during March
of 1966. R )

Both of the lube 011 tanks coils were flushed with a 15-wgt % solution
of acetic ‘acid. Considerable material was removed and the 0il tempera-

tures, upon: restarting, were reduced T°F in the case of the coolant pump

oil system, and 3°F in the case of the fuel-pump o0il system.

During subsequent operation the temperature of the fuel pump oil grad-
ually rose, and in August. 1966 this acetic acid treatment was again given
both coolers.

This problem of gradual foullng of the water side of the cooling coils
recurred throughout subsequent operations, and by December 1967, the tem-
perature of the oil supplied to the pumps was up to 150°F. The water sup-

ply was changed from tower water to cooler untreated ' 'process water" which

in the case of the MSRE was taken directly from the potable water system

via a backflow preventer. The system'was left on process water for two
months and the heat'transfer_improved enough, probably by phyéical flushing
out'offithe18cale,'so that tower water could again be used and afford satis-
factory cooling of the 011. This suitching had to’he.done at least twice

more before the reactor was finally shut down in December 1969 A semi~-
'-permanent liose and piping system.was,prov1ded S0 thls:could be done expe-

‘ditiously.

- _9,8 Increase in Radiation.Level at the Lube 0il Packages

At thedbeginningrof]powerioperation iniAuguSte1969, a radiation moni-

‘tor at the reservoir of the fuel pump oil system indicated radiation from

the upper part of the tank waerincreasing and'decreaSing'with reactor power.

:U31ng the on-site gamma spectrometer, it was determlned that the activ1ty

was qlArgon. This was. apparently produced by activation of the blanket gas

 
220

'in the upper part of the fuel pump motor cavity. Prior to this time only
‘helium had been used during power operation, and no gas'actiyation had oc-
curred. Argon was being used at this time to investigate bubbles in the
fuel loop. Since radietion levels did not ‘exceed 2 rR/hr and decreased
 when theé cover gas was changed back to helium, this was not a serious

- problem.
9.9 ”Analysis of 0il

Samples were taken of each new supply drum and from.the two lube o0il
systems after circulating for a short time. Analyses of these were used as
.controls, to compare with periodic samples taken from the systems during
operation. The main concern was that the oil, especially in the fuel pump
system, might undergo some changes due to long-time circulation through the
.high flux field in the pump or that there might be some thermal decompo-
sition. , ‘ o

Table 9.1 is a compilation of some typical oil analyses performed
during the life of the MSRE. The only significent or even minor changes
that were found were that the OH radical 1ncreased somewhat and some C=0
was formed, probebly indicating some small degree of oxidation.__Tritium
buildup was insignificant. | |

The- oil which leaked through the seals to the oil catch tanks al-
though dark_in-color, showed no significant chemical,or physical changes.

An addition fractional distillation test analysis was run on & sample
of new Gulfspin-35 oil on September 2, 196h This wes distilled under_re-
duced pressure (3—4 mm Hg) in an 18-in. reflux colum. As indicated below, |
most of the distillation occurred between 148 and 178°C | |

9.10 Replacement of 0il

The first charge of Gulfspin-35 oil was added in September l96h as the
7011 tanks were calibrated Each tank took about 35 8allons.= Periodlc ad—
-ditions were made to compenseate for losses. _ |
As 1ndicated previously, no signiflcant physical oL chemical changes
occurred which would indicate that the oil should.be replaced., Various oil

t.companies were contacted and they did not have any suggestions as to other
221

 

 

 

 

 

 

Table S
o .
Viscosity -~ Viscosity Viscosity
MSRE MSRE Date | Centistokes SsU SSU Carbon
Sample No. Run No.  Sampled Description = At 25°C At 100°F - At 210°F z
L0-1 thru LO-5 | \ L 15.44 ‘ - 85.52
and o _ 5 ea 55-4al drums to . ’ s : to
L0-9 thru LO-13: 8/25/65 new Gulfspin-35 - 15.99 - - 86.43
10-6 o 5 2/18/66  FOP-CONTROL  NOT ANALYZED
Lo-7 s 2/18/66  COP~CONTROL  NOT ANALYZED
L0-47 | | - 8/8/66  New 0il for FOP o 9.94  2.64  86.88
10-49 - 8/13/66  New Oil for COP | _ '
L0-69 11 2/13/67  From COP N 9.94 2.6 86.21
L0-70 11 2/13/67  From FOP , , 10.26 2.67 86.6
L0-140 5/16/68  ** . NOT ANALYZED
- L0-141 5/16/68  %kx NOT ANALYZED
' ' _ Solids - ppm o
LO-154 : 5/19/69  From FOP 180 9.94 - 2,58 78.0
LO-155 5/19/69  From COP 180 10.52 2.67 76.7

 

References: MSRE Sample Log for Water, Helium, Miscellaneous; and MSRE Data File — Section 6E-1 {Lube O:
ek : :
01l in service from 5/20/67 to 5/16/68 — Collected in the Waste Oil Receiver from the fuel pump oil cat

Ak
0i1 in service from 5/20/67 to 5/16/68 — Collected in the Waste 01l Receiver from the coolant pump oil ¢

 
-- TYPICAL ANALYSIS*

 

 

 

 

 

 

 

 

.1 MSRE GULFSPIN 35 LUBE OIL |
i o :
H,0
Sulphur Bromine Flash Sediment SPECTROGRAPHIC ANALYSES -- Micrograms/ML ‘ N .
% Number Point 2 viv Al Be Ca Ce Cr Cu Fe La Mg Mo " Ni P Pb Sn Sr T4 V Zn Zr Comments and Miscellaneous
9
1;26 3%3 F . <0,04 (all) ANALYZED AS ESSENTIALLY THE SAME AS LO-6 and LO-7
1.95 325°F
<} <«0,02 124 <10 0,3 <0.4 0.2 <2 0.1 <0.4 42 74 <20 <20 0.5 <0.6 <2 96 <0,6
<1 <0.02 352 <10 0.5 <0.4 1.9 <2 0.2 <0.4 fiz 250 <20 <20 2.5 <0.6 <2 136 <0.6
’ Mn
Ha0-ppm , ,
0.012 1.26 111°F 160 <l <0.01 <200 <20 <0,5 <0,2 Al - <0.1 ;grs %4 %50 --- <20 20 <05 <1 ~100 <0.5. yged o1l showed an 1ncsea§e in the free
313°p 1 <0.01 <200 <20 <0.5 <0.2 Al == <0.1 <0.5 {4 ~S0 == 520 20  <0.5 <1 A100  <0.5  gu oagd € V3630 &, and an increase
0.10 1.9 322°F H=8-§6 3.9 -~ 648 ~== -= <2 3 -— <6 = 4= 108 === - 4.6 - =-- 2.0 ==~ Interfacial tension 15.4 (normal = 18)
0.07 1.7 326°F Hag_gg <1 - 520 ww~ == <2 <2 -- <6 -— == 76 === - 3.4 - == 130 -- Interfacial tension 16.0 (normal = 18)
- . : Ag LL K - Mo Na 31
1 <1.0 ~200 1.0 10 5.0 10 1.0 0.5 20 Tr <5 1,0 — - == 200 == Dark appearance
Ag Li K Mn  Ka s1
0.2 <1.0 200 1.0 10 5,0 10 1.0 0.5 16 Tr 50 1.0 -~ -— - 200 -- Dark appearance
- K a 51 _
0.012 1.29 P s 50 0.5 A2 A5 <0.5 | ?5 1 100 Had C=0 band indication some oxidation
’ ’ - K Na 51 , N
0.05 1.58 ey o 50 Al A2 A5 L5 3 10 ~100 Had no C=0 band increase over mew oil
. - ) .
1 Analyses).
h tank.
atch tank.

 

 

 
 

 

222

means of checking for damage. However, as protection against undetected

changes, the oil was replaced in August 1966 and again in May 1967.
9.11 Discussion and Recommendations

Except for the priming difficulties with the pumps and the moisture
troubles with the motors, the systems operated very satisfactorily.

No noticesble change occurred in the lubricating oil. However, more
information is needed as to what chemical analysis or physical tests should
be made to detect undesirable changes. A more realistic sampling schedule
needs to be established. T

Improvements in future heat exchangers for this service should be made.
Excess capacity should be provided to compensate for fouling of the heat

transfer surfaces or better cooling water inhibitors should be used,

 

 
 

 

 

 

 

223

10.  COMPONENT COOLING SYSTEMS
‘ P. H. Harley

The primary and secondary component coolant systems consisted of_comr
pressors and associated equipment required to gas-cool various components
in the fuel and coolant systems. Although they performed similar functions,

- the two were comiletely independent and are therefore described separately.
10.1  Primary System

10.1.1 Descrlption
Cell atmosphere gas (95% No, 5% 05), commonly called "cell air" was

rec1rcu1ated by one of two T5-hp belt-driven p031t1ve dlsplacement pumps
(component coolant pumps). - The other pump was a spare which could be manu-
ally started if needed. Each pump was located in & COntainment enclosure -
or dome which could be opened for maintenance. After compression by the
component coolant pump, the cell air passed through a water-cooled shell-
and-tube heat exchanger and then to the reactor and drain tank cells for
distribution to the freeze valves, control rods; reactor neck, and fuel-
~-pump cooling shroud. ‘A-eide'stream was ciréulated past a radiation moni-
tqr.to indicate cell-gir activity and could be .exhausted to the containment

stack for evacuating the cells.
10 1.2 Inltlal Test;;g

The initial testing was. prlmarily a check of the system design. Flows
to equlpmentzwere.measuredTand:interdependence of flows was checked. The
- eystem wasuleak-tested and-tfiemheat'transfer coefficient of the gas cooler

-;was measured. el | |

Permanent flowmeters were not 1nstalled since most of the flows were
set 1n1t1a11y and then not- changed Durlng installation of piping, various
flows vere measured with temporarlly 1nstalled rotameters. Since air flow
“flto the fuel-pump shroud needed to be varled a permanent orifice was in-
stalled when tests 1ndicated that air. loadlng on the control valve was not
an adequate indicetion of air flow,

Initial flows were adequate except for cooling the control rods. Suf-

ficient flow was obtained by removing a discharge block valve. This block

 
 

 

 

224

valve was intended to olose on high cell-air activity to prevent gross con-
temination of the cell in case & control rod thimble ruptured. In lieu of
the block ‘valve, the discharge air was routed through-aVSS—gailon stain-
less steel drum located in the reactor cell to retain -any salt which leaked
due to the rupture.

. The measured flows are given in Table 10.1 Tests indicated that chang-
ing one flow had no effect on the other flows even when the cell was being
evacuated at sbout 200 scfm. . The pressure control valve (PCV-960) which
discharges excess air back to the cell adequately compensated for the vary-=

ing flows and maintained a constant discharge pressure,

-10 1. 3 Blower quacigy

Design of in-cell MSRE components initially required a gas-coollng
capacity of 885 scfm., However, development tests indicated that cooling
air would not be required on the freeze flanges. Therefore, the blower

capacity was lowered to 61T scfm by installing a smaller sheave on the:

drive motor.
After sbout two-years operation, more circulaetion was deemed necessary
between drain tank cell and reactor cell. This was to give a more rapid

response on the cell air activity monitors in case of a leask into the drain

- tank cell. A larger diameter drive sheave was installed to give a blower

capacity of TUO scfm and a 2-in. line was installed from the freeze valve
supply header to the drain tank cell. This gas (v200 scfm) flowed from the
drain tank cell into the reactor cell and back to the component coolant
pump suction. This T4O scfm was used for all Subsequent operation. During
rapid evacuation of the cell, it was necessary to close a hand valve in

this line in order to maintain an adequate flow to other components.

10 1. h Gas Cooler

The heat‘removal by the gas cooler was adequate throughout- the opers-
tion. The initial heat transfer coefficient calculated to be 36.5 Btu/hr
ft2 °F with a flow rate of 885 scfm. After two years of operation and at

& flow rate of ThO scfm, it was 28 Btu/hr £t2 °F and hes remained at ap-

proximately this same value., No inspection has been made to determine how
much oil and/or belt dust is deposited on the gas side (shell side). of the

tubes.,
225

- Table 10,1 Flow Distribution of
~ Primary Component Cooling System

 

 

N ¥ %%
L Pch-960

Measured
- , — o Flow :
EquipmenteSupplied scfm Design
No. 1 Control Rod Moter - 1.33 5
~ No., 1-Control Rod 3.8
~ No. 2 Control Rod Motor 1.45 - 5
No, 2 Control Rod 3.9 '
No. 3 Control Rod Motor 1.k 5
‘No. 3 Control Rod 3.9
Reactor Neck Outside. 18 15
Reactor Neck Inside 18 15
Graphite Sampler 16.2 15
FV-10h o 5
FV-10b 26.6 15
FV-105 27.7 15
FV-106 2k.8 15
FV-10T 15
FV-108 15
FV-109 = | , 15
Cell Alr Radlation Mbnitor (RE-565) . 100
L—952 260
: Fuel-Pump**' 30 100

 

| L-952 was 1nstalled in September 1966 to increase circu-
_latlon between reactor and drain tank cells.

*®%
Initial design of 200 scfm‘was lowered to. 100 scfm
Inltlal normal use was 30 scfm ‘but later no flow was requlred

PCV—960 dlscharged excess capaclty to malntain a con-

stant discharge pressure. -

 
 

 

226

10.1.5 CCP Drive Units

Originally the component doolant'pumps were driven by five matched
belts using & 15.9-in, drive sheave and a 21-in. drive sheave. When the
blowef capaéify was feduced by changing to & 10.9-in. drive sheave, belt
difficulties developed. The first set of belts failed after 1450 hours of
operation and the second set only 11 houfsfilatef., The smaller diameter
drive sheave increased the bending stresses to values in excess of the belt
ratings. To correct this, in February 1966, the drive and drive sheave dia-
meters were increased to 12.5 end 24 in. respectively. To obtain a better
belt loading, the initial matched belt sefs were replaced by a poly-V-belt.
Then in August 1966, the drive sheave was increased to 15-in. diameter to
increase blower capacity. This further reduced bending'stresses on the
belt, L

The original belts were rated at 200°F whereas the pdly—VAbelts were
only rated at 130°F, This caused difficulty since the ambient temperature
in the domes was sbout 1T0°F due to motor heat and leakage of 326°F gas
from the discharge relief valves. Rupture discs were installed to prevent
leakage but they would not withstand the shock received when the blowers
were started., Baffles were then installed which deflected the cooler in-
coming gas across the belts and thus lowered the ambient temperature to
around 150°F. .

To magnify the drive belt difficulties, the motor support was not sat-
isfactory. The motor could slip on its support plate and flexing of the
motor support legs caused a pulsating load on the drive belts. Small blocks
were welded to the support bracket to keep the motor from slipping and heavy
bracing of the stand minimized the vibration. Since the largest:strain oc-
curred on the drive belts when the blower was started, one blower was oper=-
ated continuously during a run instead of alternating the blowers twice a
month as was done initially. This increased the belt life to more than
8000 hours and fininterruptediruns of over,3000.hours wefe mfide without belt

adjustments.

10.1.6 '0il System

One of the component coolant pumps (CCP-1) had to be stopped during
Run 10 in January 1967 because of low oil pressure, and CCP-2 was used for
227

- the remainder of the run. After Run 10, a cracked copper fitting in the
0il system was repaired, and 2 gallon of oil was added to bring the oil .
level back.to normal. Drains were also installed to return any oil seal
1eakage back to the oil reservoir, |

More trouble was encountered in the CCP-1 oil,ci?culating $§stem during
Run 12, First, a loose tubing connection caused the loss of "2 gal., of oil.
This was repaired -in August.1967. When the blower was restarted, intermit-
rtent low oil-pressure alarms again occurred, An investigation indicated
no significant loss of oil, but a slow oil-pressure response was observed
whénrthe blower was.started. The suspected oil pump and pressure-relief
valve were replaced with spare units to correct the trouble. Later inspec-
tion indicated fhat'the oil pump was not damaged but the pressure-relief
valve was relieving before the normal oil pressure developed., In spite of
this difficulty, adequate lubrication had been maintained and there was no
damage to the blower. The oil pressure relief valve on CCP-2 was inspected
end also found to relieve at too low a pressure and was readjusted.

In September 1967, after six days of flush-salt éiréulation and two
dsys of nuclear operation, CCP-2 was shut down by low oil pressure. Since
the discharge valve would not close leak-tight, the run was terminated.
Repairs consisted of repairing the leaking discharge'valvé, capping a leak-
ing drain line and tightening seversl packing nuts and fittings. The other
blower, CCP-1, was used for 873 hours into November 1967. At this time a
serious o0il leak was indicated by iow 0il pressure and accumulation of about
2 gal., of 0il:in the condensate collection tank connected to the blower
containment. The standby?unit, CCP-2, was immediately started up and oper-

ated without further difficulties throughout the remainder of the run,
| 35036 hours. S B S | |
- In-July 1968, the b?azedgtubing 0il system on .both blowers was re-
placed by welded pipe and flexible hose. This eliminated most of the pro-
blems with the oil system. A leaky shaft oil seal had to be replaced in
‘December 1969. | | |

10.1.7 ©Strainer:

 In August 1966, a temporary strainer with a'l/l6-in;'scféen was in-

stalled in the component-cooling-pump discharge line to collect rubber dust

 
 

 

 

228

from the drive belts, and prevent blowing this material onto the hot metal
components being cooled. Although the temporary strainer worked satisfac-
torily a new strainer made of 100-mesh screen, which had been on order for
a year, was installed in the line in May 1967. Over an eight-month period,
the temporary strainer had accumulated 30 to 50 g ofkblack, dry pdwdery ma-
terial that appeared to be dust from abrasion of the drive belts., After
being decontaminated, the strainer was examined and was found to be in very
good condition. The surface was slightly etchéd, but no more than would
be caused by the decontemination process. -

The new strainer developed excessive pressure drop in September 1967

and the screen was replaced., In 1,800 hours of operation the 100-mesh screen

had become partially plugged with rubber dust from the drive belts, It was
damaged in removal and a replacement basket was made using 16-mesh), 0.023-
in. wire screen. This showed no pressure buildup over the subsequent 2

years of 6peration.
10.1.8 Condensate Collection

After a reactor cell space cooler began to leak water into the cell in
March 1966, condensate accumilated in the 10-in. suction line to the com-
ponent coolant pumps at a rate of 1 to 2 gpd. The water was vaporizing in
the reactor cell and condensing on cold surfaces at the component coolant
pumps. Simple drains were installed, but these could be used only when the
reactor was subcritical due to radiation in the coolant drain tank cell.
Handling of the drained water was complicated by the tritium (up to 915
uc/me produced from the 6Ii in the treated-water corrosion inhibitor or
from diffusion through primary piping. During the shutdown in August 1966,
piping was installed to permit draining the domes during power operation
to a tank in the sump room. It was periodically pumped from this tank to
the liquid waste tenk. ' s |

© 10.1.9 Electrical

In June 1966, the wiring to CCP-1 drive motor shorted out in the pene-
tration into the containment enclosure. The short tripped the main bresker
to generator bus No. 3 as well as the component coolant pump breaker. The

.pump and other equipment on bus No. 3 stopped. Emergency diesel power was
 

229

started immediately, but because of a misinterpretation of the problem, the

‘reactor drained before'cooling was restored to the freeze valves by switc-

hing to CCP-2. The short occurred'in an epoxy seal on the end of the
copper-sheathed mineral insulated ceble and was caused either by moisture
from 1n—cell leskage or & breakdown of the epoxy potting compound.

" Wiring changes were made to both component cooling drive motors to
prevent a reoccurrence. Each phase of the three phase circuits was brought
through a separate penétratibn and the epoxy was omitted from the end seals.
No further difficultiesrwere encountered.

10.1,10 - Valve Problems

Isolation valves were provided on each component coolant pump to per-

mit repairing one while the other was in operation. "These large valves

- (10 in. on the inlet and 6 in. on the outlét) had back-seating stems on tef-

lon seats -to prevent leakage when the valves were open. These proved to

be very satisfactory. After sbout 3-years service, one of the 6-in. iso-

lation valves leaked in excess of 100 cfd. An inspection revealed several

- large pits in the seating surface of the valve which had been filled with

epoxy during manufacture.' These were repalred'by filling with weld metal
and then reflnishlng the seating surface. When reassembled, the leakage
through the wvalve was-NO 5.cfd which was considered to be satisfactory.

Check valves were 1nstalled on the discharge of the blowers to prevent

_backflow through the stand-by unit. These were 6-in. butterfly-type valves
, W1th silicone rubber hinges holdlng two flappers in place. In January 1966,

after 1640 hours of serV1ce, “the hinge on the CCP-2 valve broke allowing

-one,flapperfito fall off,e;Nomspares-were available so flappers and hingers

were made at ORNL. This hed not.failed after 1800 hours but was replaced
by factory-built parts. tIn'Navember*1966- the butterfly'check'valve from
CCP-1 was found to be inoperative and was repaired. - The fallure also oc-

: ;curred at the silicone rubber hinge which supports the two wings of the

check valve.-

On one occasion- during the perlod when belt prdblems ‘were causing loss

'“-of'blower-capaCLty, the stem_of_the in-cell pressure control valve (PdCV-

960) froze in a partially openrposition and failed to close when the valve

operator air loading increased. This was the only failure of an in-cell

 
 

230

~component in this system. The valve was removed remotely and répacked.

The stem was polished and lightlyglubricated'beforé reinstallation. To
prevent the valve from remaining in one throttled position for an extended
time, the pressure controller was siubsequently cycled periodically to be
sure the valve was functioning. Moisture in the circulating air probably
contributed to the valve failure, and the periodic draining of the collected

moisture has helped prevent a reoccurrence.
- 10.1.11 Conclusions and Recommendations

Although many difficulties have been encountered with equipment, bnly
two reactor drains have been caused directly by component cooling system
failures. One was loss of one blower while the other was isolated for main-
tenance and the second was & failure of electrical cable to the motor of
one compressor (CCP-1). One run was restarted after repairing -an oil leak
which developed early in the run.
| Better design specifications would have assured more serviceasble items,
'For,instancé,“the pressure relief valves had a specified capacity at 15
psig, however, they discharged a small amount of gas continuously at nor-
mal operating pressure. This caused excess heating of the equipment in the
containment domes. Combining overpressure protection with an unloading
falve end controls would have permitted blower startup without load and
caused less wear on the drive belts. A more dureble hinge would have pre-
vented the check valve failures. _

_Mbre instrumentation on gas flows and on the o0il system would have
given earlier indication of pending difficulties. O0il sump level and dis-
- charge pressure with indicators outside the containment would have helped
here, » _ ‘ o 7

Although sufficient capacity was available, only minimum capacities
~were available for cooling at some locations because of limiting pressure
drqps of piping and control valves. More design consideration should have
been given to the possibility of needing additional coolihg. 1After cor-
,_recting blower drive problems during the first two years of ofieration, sev-
eral reactor runs were made in excess of 3000 hours with only preventative

meintenance between runs.
231

10.2 Secondary System

10.2.1 Description

Ambient air for cooling out-of-cell components was supplied by either
& 60-scfm positive displacement pump (CCP-3) or a piston-type air com-
pressor (AC-3). The air was distributed to the coolant system drain freeze

valves and to thé chemical pr0cessing plant freeze valves,
10.2.2 Operating Experience

. Barly testing revealed that the flow to FV-20k and FV-206 was.insuf—
ficient to establish good plugs in the freeze valves. The air-operated
control valve trim was ehlarged'to reduce the pressure drop and the blower
discharge pressure was increased from 8 to 9.25 psig.' With these changes, .
‘adequate air capacity, 25 scfm, was obtained to each freeze valve.

In April 1967, the drive ‘shaft of the rotary blower, CCP-3, seized in
the brass sleeve bearlngrof_thc blower housing and did extensive damage to
the drive shaft, Operation was continued using the service air compressor,
AC-3, which was the standby supply to keep FV-20l and FVf206 frozen., A
second besring failure occurred in CCP-3 after only 200 hours of operation
following the first failure. Since that time, AC-3 has been used as the
rfiormal cooiing air supply with;CCP—B as & standby,unit. CCP=-3 was used
during programmed maintenance on AC-3 and while freezing FV-204 and FV-206
when more air was-rcquired than'the air regulator from AC-3 would supply.

A brass swinging gate check valve which prevented backflow through
CCP—3 -when the service air compressor was used, failed because of wear
caused—by.pu;sating flow vhep_CCP—3"was71n service. Thls_check.valve was
"~ replaced by a hand valve. Thén; vhen the air compressor was changédvto the

rnormally operated supply, a“név check valve was installed so the blower
_could be restarted remotely if ‘a low header pressure alarm occurred The

hand valve was 1eft installed S0 the blower could be 1solated if the check

valve falled again. .
10 2.3 Conclus1ons

 The positive dlsplacement pump (CCP—3) whlch‘was 1ntended as the prime
air mover functioned poorly. However, the service alr compressor (AC-1)
was very reliasble and thus llttle or no Operatlng time was lost due to this

system,

 
 

232

11. VENTILATION SYSTEM
" P. H. Harley

11.1 Description

~ The purpose'éf the ventilationsystem;was to exhaust air from areas
where there was a potential danger_of release of radioactivity. The air
flowed through a network of ducts to the filter pit which contained 3 paral-
lel banks of roughing filters,and absolute filters. It was pulled through
the filter pit by one of the tfiorstack fans and was then discharged up a
100-ft stqck to the atmosphere. The stack was cbntinuously monitored to
assure that the radioactivity released was within acceptasble limits. OSen-
sitive differential pressure gages were installed to indicate the direction

of.air flow between the various areas.
11.2 Operating Expériencé

After installation of the ventilation éystem.was'complete, a series of
tests was run to insure that the‘criteria would be met. Several minor.
changes were made in the ducting, cell blocks were caulked, and leaks re-
paired. The biggest problem encountered was_seaiing the hi-bay. Gaskets
were installed on the doors, all joints in the lining were taped, and a
number  of large holes were plugged. The leakage into the hi-bay was meas-—
ured as follows. The dampers in the ventilation ducts from all other'areas
were closed. All inlet vents and doors to fihe hi:bay'were closed. The
dampers between the hi-bay and the stack fans were throttled until the hi-
bay pressure was —=0.3 inches of water. The in-leakage (stack flow) was”
found to be 4250 ftalmin. Although this was above the design-valué of
1000 £t3/min, it was accepted. ' ) S

The pitot tube stack gas flow meter was calibrated using & hot wire
anemometer. - Several traverses were made under various flow conditions.

. The flow under normal operating conditions was found to be around 23,000
cfm, The stack calibration curve and details of the calibration are given
in MSRE Test Memo and Test Report 2,3.15. |

Tests indicated that adequate ventilation could be maintained with 2
of fihé 3 perallel filter banks in service. .Therefore, it wes possible to
replace one banfi of filtérs without interrupting opefations.
 

233

After installation, the sbsolute filters did not pass the standard
ORNL-DOP smoke test,®5 (99.95%'efficiency is required.) Visual inspection
.indicated leakage around the fremes which required considerable caulking
and some minor revisions. On 11/1h/64 all three banks passed the DOP test
(East = 99.970%; Center = 99.972%; and West 99,967% efficient). The next
test in May of 1965 indicated that all three banks had dropped to less than
99. 90%-effic1ent Therefore, they were removed and exten81ve recaulking
and revisions were made to improve their efficiency and assure that they
could be replaced remotely. The results of subsequent DOP tests are shown
in Table 11,1. | | |

It has notlbeen necessary,to replace the &bsoiute,filters since 1965,
The pressure drop through them had changed very little during the i years
of operation which indicates that the'roughing filters performed well and
removed most of theeparticulates.

7 The roughing filters_#ere_changed when the pressure drop across them
- reached 5 to 6 inches of water at which time the stack flow would be down
to~&18;000 cfm. During earlier operation fihile considerable maintenance
~and construction was still being done, these plugged rather repidly. Small
easily changed fiberglass filters installed on several of the inlet ducts
reduced the plugging_rate.HWSUbsequent filter changes are shown in Tsable
11.2. The last set‘of'rdughing filters read “500 mR/hr at 3 inches after
they were removed. o |

Some difficulty;was encouhtered inrearly Operafion-with stack fan No.
1 (SF-1) bearings. The bearings were initially grease-lubricated and there
.was difficulty in maintalnlng proper lubrlcatlon. The normal bearing life
was asbout four months, In 1966 an oil-lubricating system'was installed

and the sfibsequent bearing llfe has been sbout 18 months. SF-1 was run

- continuously except when necessary to do maintenance on it SF-2 served

 

 

as a standby uit, = -?e“

o Sensztive dlfferentlal pressure gages (full scale = 2 inches of water)
.were provided to 1nd1cate dlrection_of air flow. These were recorded each
shift on the building log and dampers were adjusted‘es'neceSSary; Venti-
lation flow was from the less hazardous to the more hazardous ares except
in the coolant cell during power operation. When the main radiator blowers

were operating, considerable air leasked from the radiator and pressurized

 
 

234

Table 11.1 Results of DOP Tests of Absolute Filters

 

 

 

Efficiency _
Bank 10/65 3/66 6/67 9/68 8/69
East 99.999%  99.998%  99.979%  99.994%  99.993%
Center 99.99k 99.995  99.998  99.996  99.995
West 99.994 99.997 99.99h 199.994 99.985

 

Table 11.2 Roughing Filter Changes

 

East | Center West
October 1964 x x X
October 1966 X X x
August 1967 x x

September 1968
August 1969 oo x x

 
 

 

_' 235

the coolant cell. In an effort to correct this, the inlet to one of the
radiator suxiliary blowers (MB-2) was connected to the coolant cell. This
reduced the pressure below atmospheric but it was still above the hi-bay
pressure. Since the only hazard was from beryllium and since the coolant
stack was-continoofisly:monitored for.ber&liiom, this was considered ac-
ceptable. | - '

During malntenance, containment is prov1ded by the flow of air into’
the cells. The system has adequately supplied thls need for all mainte-
nance operatlons.- |

There has been very llttle activity released to the atmosPhere except
during maintenance perlods. Even then the releases have been below the
ORNL permiSsible limits. - The permissible release rate from the 5 ORNL
stacks is A1 curie per week, Table 11.3 is a tebulatlon of the amount of
activity released from the MSRE stack. The maximum released in any week
~at the MSRE was 568 mCi durdhg the week endifig December 21, 1969. This
weé mainly'ffofi'the-cofefepeeimen-removal plus about 50 mCi ffom the pri-
mary system lesk which occfirred_during the final shutdown.25
| Due to a galled bolt, the,core specifien hold-down flange could not be
reinstelled and was left hanging-in the standpipe. Another flange was in-
stalled to close the system. ©Salt on the hold-down assembly probably con-
tributed to the larger than usual release. The standpipe vacuum was also

.inedvertently left off on tvo occasions when it should have been on.
11.3 Conclusions and Recommendations

The ventilation.systei has.fUnctioned-satisfactorily_during.operation

‘and maintenance. The installtion of easily removable filters at inlets to

. the ventiletion ducts prolonged the life of the roughlng filters and showld

~ be considered in future de51gns.

 

, .The stack monitorlng system could be improved The.recorders used
'were very difficult to make notes on. This, together with the.renge'chang-
"ing feature of the.system,"made'later interpretation of the charts very
difficult., In addition to this, the monitors were never calibrated. This
should have been done and limits on their readinge should have been es-

tablished before Starting ofieration.

 
 

 

236

Tdblé 11.3 Semiannuel Stack Release

 

" Dates

 

 

9/69  1/70

Particulate  Gaseous
From To mCi mCi . Main Source of Activity
3/66 8/66 0 97  Miscellaneous maintenance
9/66 2/6T 9 206 . Off-ges line maihtenghce.

3/6T  8/67T  <0.3 ho . -

9/6T  2/68  <0.3 2.3 |
3/68  8/68  <0.3 33 Off-gas line maintenance
9/68  2/69  <0.3 ok o
3/69 8/69  <0.3 12.5 OFT vent maintenance |

<0.3 8.2 Removal of core speciméhs an

primary system leak

 
 

 

237

lEJ WATER SYSTEMS.
P H. Harley

Water for various typeS'Of-usage5Was‘supplied at the MSRE by the fol-
lowing systems: (1) the potable water system distributed ORNL water for
fire proteCtion, general building services, and was the supply for the pro-
cess water system, (2) the process water system distributed water (potable
water which had gone through a backflow preventer to provide isolation) for
process usage,"such as dilution of.the'liquid'waste, for temporary or emer-
gency cooling and for meskeup to the cooling tower, (3) the cooling tower
water system recirculsted treated process water through various out-of-cell
components and provided cooling for the treated water, (4) the treated wa-
ter system recirculated treated steam condensate in a contained loop to
provide cooling for all in-cell or contained components, (5) the steam con-
densate system condensed ORNL steam to produce pure condensate which was .
used in the drain tenk afterheat removal sYstem_and for treated water sys-
tem mekeup, and (6) the nuclear instrument thimble water system recircule-
ted¥treated,steam.eondensate to provide cooling for the nuclear instruments
and biologicel shielding from the reactor vessel.

The operating experienee, difficulties encountered, and recommenda-

tions are given separately for each sub-systenm,
f.lE.l'_Potable Water System

Two six-inch cast iron lines supply water to the MSRE area from the
supply main, One line suppliee'fiater to the fire sprinkler system and

- general building services and the other supplies makeup to the process wa-

 ter system. T s

 

| The underground supply 11ne to the fire sprinkler system and building
~ services ruptured in the fall of 1967 ‘Repairs required: shutt1ng off both
.potdble water supplies to the area 80 process water makeup was supplied by
8 fire hose from a vent line on’ the valley water main, Fire protection was
':prov1ded by ml/h mile of flre ‘hose from the Nuclear Safety:Pilot Plant and
posting of a fire watch at the local fire alarm box in the bPuilding. Iso-
latlon valves where the lines teed off the water main would have 51mp11f1ed
fihe regalrs. In December 1969, the other 6-in. supply line ruptured under-

’_ ground. Process water was supplied from the other 6-in. main through a

 
 

 

238

temporary backflow preventer during repairs. In both cases, repairs were
- made using a stainless steel Adams clamp. '

Although no serious malfunctions have occurred in the automatic fire
sprinkler system,  a better designed system could have been provided. The
MSRE system was installed from nationel fire profection specifications
without regard to area services. In some areas containing electrical equip-
ment, more Hanger existed from possible water damage than from the possi-
bility of fire. In these areas the sprinklers were disconnected or iso-
r'lated'with manual valves. Portable extinguishers suitable for electrical

fires have been provided in areas containing electrical equipment.
12.2 Process Water SyStem

This system has performed very well. On a few occasions, process wa-
- ter was used for cooling critical equipment when repairs necessitated shut-
ting down the cooling tower water system. Process water which was ~20°F
cooler than the cooling tower water was effective in cooling the fuel pump
and coolant pump oil tanks when the coils became fouled with scale, The
use of process water tended to flush the coils so that adequate cooling
could be obfained'using'cooling tower water. |

The process water system was supplied by ORNL potsble water through
one backflow preventer. Another backflow preventer was installed in series
with this in the process water Suppiy to the liduid waste tank, These
backflow preventers were scheduled to be tested semiannually. Actual test-
int occurred at intérvals of 3. to 14 months. The main backflow preventer
.never required any repairs. The one in the line to the waste tank réquired
minor repairs on”two occasions. Velving was provided to install a second
- unit in parallel if repairs of the main operating backflow preventer were
required. Howefier, the second unit was never installed, Instead the pro-
cess water flow was shut off to make the periodic tests. For a more com-

plex Watér system, two backflow preventers should be installed.

i
239

12.3 Cooling,Tower Water System

Standard centrifugel pumps were used to circulate the cooling tower
water. Flow was adequate and no pump performance tests were made. The
cooling tower was a packagedrcommercial'unit and_provided adequate cooling.
Very adequate temperature control was obtained using a temperature con- -
trolled valve which allowed part of the water to bypass the cooling tower.
One major pipe leak occurred whlch required shuttlng the entire system down.
This leak was in the underground six-inch return line to the cooling tower.

Nalco-360 balls, a mixture of sodium and potassium chromate containing

some phosphates, were used.as a corrosion inhibitor.
| In the summertime, & one-pound bell of Nalco-21S, an algae inhibitor,
was added daily and was satisfactory in preventing algae formation on the
cooling tower. Hardness was controlled by bleeding off about 2 gal. per
minute of cooling tower water. This water plus about 8 gallons per minute
of process water was used to cool the charcoal beds.

Over 1500 samples of CQoling tover water were taken and analyzed
Most of these vere analyzed at the site to assure adequate treatment. Peri-
'odlcally the samples were sent to the laboratory for iron analy31s. This
remained less than 1 ppm throughout the operation. Due to the continual
nwkeup'and bleedoff from the?syStem, good corrosion rate figures are not
available, however, no.leak has_occurred which could be attributed to cor-
rosion. o - , |

A con51derable ‘amount of calclum phosphate scale coated the . rotameter

'tubes and cooling coils and . left sludge in the coollng tower basin and low

o veloc1ty areas of the treated water cooler. Equlpment was 1nstalled to me-

 

"ter & solution of pota551um dlchromate and zinc-sulfate inhibitor into the
' cooling tower water by a proportionlng pump powered by the coollng tower

"makeup water flow. When thls pump failed to operate propertly, the change
- in inhihitor was abandoned nghts were installed behlnd the rotameters

' to facilitate readlng and mlnlmlze frequency of cleanlng.,

 
 

240
12.4 Trested Water Systems

12,4,1 Initial Fill and Operation

The treated water system’uas filled by adding steam condensate and
“corrosion inhibitor batchwise. The total volume was about hOOO gallons
with about 2300 gallons in the thermal shield,

12.4,2 Circulating Pumps

Standard centrifugal pumps were used to circulate the treated water
in a closed loop. The capacity was adequate and no pump performance tests
were made. The pumps have operated satisfactorily with little meintenance
other than repairing the rotating seals, Leekage from these was collected
in plastic bottles due to the- induced activ1ty in the water.

12.4.3 Thermal Shield

Initial testing indicated that excessive pressure would exist in the
thermal-shield.and that the flow was not adequate'to the thermal shield
slides (sections of the sides which could be removed for replacing the re-
actor vessel). Piping changes were made to lower the back pressure &nd to
provide a separate supply line and pump to assure adequate flow in the
slides. During early operation, several rupture discs 1n the thermal
shield discharge lines were ruptured due to the supply and discharge valve
c1051ng simultaneously To prevent this, the supply valve was modified 50
that it would close before the discharge valve closed. |

- The water level was noted to increase in the'surge tank when the re-
actor power was increased and the level decreased when the reactor was sub-
critical. Decomposition of the uater was producing "2 cubic feet per hour
of Hy gas at full power and about 8 ft3 of this gas collected in one. of .the
TS slides in which it is postulated the water piping was reversed i.e.,
water entering the top end 1eav1ng from a dip leg near the bottom. A small
deaerating tank installéd in the water supply to the TS slide permitted
2 ft3 of gas to redissolve when the system reached equilibrium. Later a
large degasing tank was installed to deserate the water leaving the TS to
permit more Hy; to be dissolved. This reduced tHe‘accumulate-gas to ¢3 £43,

Initially a purge of N, and later air was introduced into the surge tank
241

and degasing tank to flush the gas space and dilute the Hy, to below the

- explosive limit., No further difficulties were encountered, however,

%3/h-gal. of water/day was lost by evaponation in the purge gas.
12 b, h Reactor and Drain Tank Cell Space Coolers

Two space coolers were ;nstalled-ln the reactor cell and one in the
drain tenk cell to keep the cell air'temperatures below 150°F. The meas-
ured capacity was only L0 te 60 kW compared tq the design valfie of T5 kW.
However, the cell temperatures normally ran at about 140°F for the reactor
cell and 130°F for the draln tank cell. |

The 3-hp fan motors orlglnally installed on the Space coolers, were
edequate-for the normal operating pressure of —2 psig but_were too small
for higher pressures such as could be encountered in the event of an acci-
dent. They were, therefore-replaced with 10-hp motors.'-One of these en
the drain tank cell cooler failed after 2900 hours of operetion due to the
rotor slipping on the.ehaft.',Since the trouble appeared to be caused by
faulty construction rather than poor design,_itflwas replaced by an identi-
cal motor. | '

- Before the cells were sealed, a reactor cell cooler (RCC-2) developed
a leak at a header-to-tube brazed fitting. Repairs proved to be difficult
because .heat required for brazing one lesk would cause an-adjaeent tube to
start leaking. After the cell was sealed, a 1-1/2 gallon per day water
leak was traced to another brazed fitting lesk on RCC-1l. ThlS was re-
paired. - In Juiy 1967, RCC~2 again started leaking and was replaced by one
of three newly. dbtalned units An which the copper tfibes were heliarc-welded

(1nstead of brazed) to the main header.
12.4,5  In-Cell Water Lea.ks

_ In-cell water leaks.have;persisted throughout the reactor operation.
Before the cells were sealed, components could be checked;viSually but af-
“ter pbwer operation,_locetinéfit?eated water leaks required hydrostatic
. testing of in-cell componente.} Although all known leaks such as at the
space - coolers were repalred, we continued to have g leak 1nto the cell of
1/2 to'1 gallons per day.. Water from this small leask did not collect in
the cell sump but evaporated in the 130 to 1LO°F cell gas and then con-

densed in a colder portion of the containment, the suction lines and gas

 
 

 

242

cooler of the component cooling system. This condensate was found to con-
tain considerable amounts of tritium. Drains were installed¥tofpefmit7
draining the condensate to a measuring tank in the sump roofi'fromjwhich'it
could be pumped to the llquid waste storage._' , , ‘

The nuclear instrument penetratlon was one of the suspected sources
of the leak. Since it could not be hydrestatlcally tested, 12 kg of Dy0
:was added as a tracer. This.increased-the deuterium concenfiretion to about
seven times the normal water concentration. Since the deuterifim'eoncéfitra-
tion did not increase in samples of the cell condensate, a leak from the

nuclear 1nstrument penetration was discounted.
_12 h 6 Treated Water Heat Exchanger

' The heat exchanger installed to transfer the heat from the treated wa-
ter to the cooling tower water was & surplus tube-and-shell unit. During
'-preoperational tests the tubes had to be rerolled and tube sheet gasket re-
placed to eliminate intermixing of treated water and cooling tower water.
After a year of operation, 17 tubes which had developed cracks where the
tubes were rolled into the tube sheet were plugged. This was probably
caused by the earlier tube repairs. When leeks again developed in the TW
" cooler, more tubes were plugged but when attempts to seal the flanged tube
sheet were unsuccessful,,the heat exchanger was replaced with another sur-
plus unit which was smaller but still had adequafie cooling capacity. Af-
ter replacing the heat exchanger, sodium from the cooling tower system had
to0 be removed from the treated water by dilution with steam condensate and
new inhibitor was added. The activation of the 1.6 ppm of sodium left in
‘the treated water after the dilution raised the radiation 1evel of the sys-
tem to 80 mR/hr at surge tank and 260 mR/hr at the fllter but did not seri-

ously affect reactor operation.
12.4.7 Valves
- Containment of the treated'water_System depended upon air-operated
block valves and check valves. None of the block valves leaked during the
annual containment checks. However, several of the‘piston—type'check valves

‘leaked excessively.  Foreign materisl and scale caused some of the valves

to bind in the open position due to the close clearance between the valve

\&/
 

243

- body and the plug. The valves were mounted in horizontal pipes which
probably magnified the problem."

Pressure rellef valves were installed to prevent damage to in-cell
components in case the block valves closed. These discharged to the liquid
waste tank. Leakage from these occurred several times. The first indi-
cation was usually an inorease_in rate of condensate makeup. Pinpointing
the leak involved. disconnecting piping from each relief valve. Power re-

duction was necessary to- aecomplish this.
12, h 8 Corrosion Inhibitor

The treated water system was first fllled with condensate containing
2000 ppm of a 75—25% mixture of potassium nitrite-potassium tetra—borate.

Induced act1v1ty l+2K at low reactor power weas extrapolated to radia-
tion levels of hOO mR/hr at the surge tank in the water room and filter in
- the diesel house when-the reacto: was at full power. To prevent this radi-~
‘ation, the potassium inhibitor was chenged to lithium., Lithium-T was used
~to minimize the production of tritium which would be formed in the thermal
shield.. The potassium was removed by dilution with demineralized water
- from ORNL but had to be repeated using steam condensate becsasuse the demin-
eralized water which contained:1 ppm of.sodium became radioactive during
-~ the next run, The new inhibitor was composed: of lithium nitrite, boric
acid, and.lithium hydroxide. :

During early pover operatlon, the treated water lost active inhibitor..
Samples indicated that 300—500 Ppm H202 was being produced by radiation

"damage which oxidized LiNOz to LiNOg. Additional L1N02 was added until

. equilibrium of ~NOp— N03 was ‘reached with ~T00 ppm of N03 in the water.
; Approximately 365 samples of the treated water were taken. ‘Most . of
”these were analyzed at the . site to ‘assure that adequate treatment was be-

iing ma1nta1ned.
:12 h 9 Corrosion

B Corr051on -in, the treated water system has not been 51gnif1cant. " Dur-
ing most of - the operation, samples were analyzed monthly for 1oss of in-
hibitor, pH and_iron.' In 1969, chromium and nickel analyses were added

to further check for corrosion on the in-cell stainless steel. The iron

 

 
 

 

 

 

244

analysis has remained near the lower detectable limit of the analysis indi-
cating no corrosion and the Cr on Ni analysis also indicates no significant
corrosion., Although the chromiufl and nickel snalyses haven't been followed
continuously, any corrosion would have caused an increase in the amount
present. The present chromium anslysis of .4 ppm is equivalent to 0.05
pounds of stainless steel. _ _

One factor which clouds the interpretation of the corrosion results
has been the continual leaskage from the system which has required periodie
mekeup. The decrease in concentration of inhibitor of 25% in two years
agrees reasonably with the known leakage. An increase of the iron, chrom-
ium, and nickel by the same ambunt would still indicate & negligible amount
. of corrosion. If corrosion produéts precipitated as oxides, they would be
caught on the treated water filter. This filter, located in the diesel
hbuée, built up excessive pressure and was cleaned twice during pfecritical

testing but has not required further cleaning in the past four years.
12.4,10 Recommendations

- The biggest problem in the treated water system was the result of the
low design pressure of the thermal shield (2030 psi). Care should be ta-
ken in future designs that all components withstand the dead~head pressure
of the circulating pump. The piping changes required for protection of the
thermal shield resulted in other parallel flow paths having marginal ca-
pacity.

Consideration should be given to the design of a system which is com-

pletely contained. This would eliminate the need for so many block valves.

Check valves, of the type installed at the MSRE, should not be relied on as
‘containment block valves. If similar valves are used, they should be in
vertical piping and have sufficient clearances to insure that they will
close when needed. |

A corrosion inhibitor should be selected to minimize induced activity.
Provisions for handling gaseous activity such as tritium and radiolytic
hydrogen should be provided. Lafge components should be.shielded or lo-

cated so as to minimize personnel exposure to the induced activity.
 

 

245

- 12.5 -Condensate System

Building,steam condensate.has been a‘very_satisfactory supply of de-
mineralized_water. The steam is condensed in a stainless steel tube-and-
c_shell_heat exchanger and stored in two stainless steel tanks,

.Initially condensatevmakeup was added by gravity feed through an auto-
mstic level control valve to maintain a constant level in the treated water
'surge tank. To obtain better treated water inventory data, this was
changed . to periodic manual -additions when the surge tank level decreased
 below 50 gal. . |
~ Condensate is used to flll fEedwater tanks which supply cooling water
- for recirculating evaporative cooling of the fuel drain tanks.' A corrosion
inhibitor in this service would plate out during boiling. The system has
worked - well.except for 2 occas1ons vhen leaking valves-between the feedwater
tanks and steam dome caused accidental addition of water to a steam dome.
At one time the high drain tank temperature (1300°F) interlocks actuated
the supply valves to FD-2 steam dome following a normal drain. This could
~ have been prevented by turning off electric heaters to the drain tank
shortly after the drain, DueAto refluxing,‘all of the water could not be
removedlfrom the steem domelunenrcooling was no longer required To cor-
rect this, 8 pump and storage drum were 1nstalled in the north electric

serv1ce area to permit complete drainage of the steam domes when required
12.6 Nuclear . Instrument Penetration

Instruments for monitoring nuclear characteristics of the system.were
located in & 36-1n.-diameter shaft which extended from,the high-bay floor ,
to the inner edge of the thermal shield. This was originally filled with '
about 1700 gallons- cf demineralized water to which a 75—25% mixture of po- .
tas51um nitrite and potassium tetraborate was added to a concentration of
2000 ppm._ As with the treated.water system, the 1nh1b1tor was changed to
lithium nitrite-lithium- tetraborate to lower the 1nduced actiV1ty. Cooling
was initially by natural circulation but -gamma. heating when the reactor was
‘taken to full power. increased the water temperature “T0°F. This decreased

the neutron attenuation causing a divergence bhetween the indicated nuclear

 
 

 

 

 

246

power and the heat balance power. A cooling system using a S5-gpm pump and.
a small available heat exchanger were installed which reduced the increase
in water temperature to A18°F and kept the nuclear power measurements and
heat balance caleulations within ~5%. | R
Initially treated water was used for mskeup to the nuclear instrument

. penetration. Since the main loéseS”were by evaporation, this caused the
inhibitor concentration to increase. Steam condensate was then used for
water makeup. |

- Approximately 215 samples were taken of the nuclear instrumént pene-
tration water. Most of these were analyzed at the site to assure that ade-
quate treatment was maintained. As in the treated water system, the iron
analysis of samples sent to the laboratory remained near the lower detect-
able limit and the chromium content-never exceeded 0.4 ppm indicating very

little corrosion.
12.7 General Water Systems Conclusions

The water systems were quite adequate afid_as a whole'presented very
few problems. The biggest problems resulted from the marginal préssure
rating of the thermal shield and piping error to one of the thermal shield
slideé which résulted in the formation of a large gas pockefi. The piping
changes réQuired to prevent over—pressurizing the TS and additiohal equip-
ment needed for gas stripping weakened the‘performahce of the treated water
systen.

Corrosion inhibitors in both dooling towér water and treated water
systems proved to be excellent inhibitors but they had undesirable side
effects — the scale and sludge in the CTW system and induced activity and
tritium formation in the TW system.

Equipment and instrumentation performance was very good. The‘only
‘equipment troubles were the treated water heat exchanger and brazed tubing
in the cell space coolers. The in-cell water leakage (preéumably from the
‘'space cooler brazed fittings), was annoying, but after installation of the
condensate drainage system, did not-interfere with performancé of the re-

actor.
 

247

13. LIQUID WASTE SYSTEM
o P H Harley

13 1 Descrlption

The liquid waste system consisted of eqfiipment which:handled the aque-
_'ous wastes from the area. Non-hazardous wasts were pumped to small catch
basins and then to & drainage field west of the area. This water was from
building foundation drains and various floor drains in the building, over-
flow from the charcoel bed cell, etc. Hazardous wastes, primarily the
'leakage into the reactor cells, and auxiliary cells were pumped or Jetted
to a waste storage tank for sampling and treatment and then transferred tc
the Melton Valley Waste Disposal. System.

The waste system design included a component decontamination facility
and a. water—shielded -cell in which maintenance of radioactive components
could be performed. Some of the equipment was installed but as an economy

“measure, this secticn of the system was not completed.
13.2 Preliminary Testing

' Piping and tenks were lesk-tested fihile running performance tests on
the pumps and'Jet ejectors used to transfer liquids. The 1iquid waste
'storage tank and cell sumps which have bubbler-type level indicators were
calibrated and the eJectors were then tested to remove the water.
| - No difficulties were encountered with the steamyoperated ejectors in
- the. auxiliary cells but the capacity of . the eir-powver Jets in the reactor
"and drainutank cell was exceedingly low. Attempts were made to 1mprove the

performance of the ejectors by changlng the venturi diameter but no sig-

”Tfniflcant improvement was noted Back pressure of the long'discharge line

'-fij_limlted the Jet capacity.j Modiflying piping in the waste cell helped some

but to. dbtaln more capacity, a steam.supply was piped to the jet supply

_leines at the penetration to the drain-tank cell.' Prov151ons were made to

~ use either air or steam.(60 psi) for the jet supply. Water capacity of the

Jets using air was 250—350 cc/min end nsing steam 8 5 gpm._ To keep mois-
“ture out of the cell, air remained the preferred supply. '

Although the decontamlnaticn facllity was notrcompleted, the waste
filter for clarifying shielding water was installed and tested. Filtering

 
 

 

 

248

required pumping 30 gpm through the filter using the waste pump but back-
flushing the filter required'mlho gpm. Process water was used for this
operation and the flow was calibrated versus the process water pressure
required (15 psi = 140 gpm).

Initial testing of the waste pump for circulating (mlxlng) waste tank
contents and transferrlng was tested satisfactorily but the pump hed a ten-
dency to gas bind at low tank levels and process water was required for
priming. | _

These initial tests were used to check and improve operatlng proce—‘

dures as well as test the equlpment
13.3 Operating Experiefice

The primary aqueous activity found in the MSRE waste has been 3H. A
continuing water leak into the cell since early power operation* has been
drained into a condensate collection tank and transferred to the waste tank
at least once a week. The water from the in-cell leak did not collect in
the cell sump. The water evaporated in the 135°F gas in the cell and then
condensed out after being compreséed to 6 psig from —2 psig and cooled to
%110°F in the component cooling system gas cooler. This water contained

up to 6 curies/gal. Most of the water in the waste tank was from rain wa-

’ter'which drainéd from the ventilation stack and filtered into a collection

tank and was pumped to the waste tank. Although this water could have been

contaminated from material on the filters, very little activity was de-

tected. Other sources of contaminated water in the liquid waste were from

the caustic scrubber during fuel processing and wash water from cleaning

the hi-bay floor and auxiliary cells following reactor maintenance periods.

Table 13.1 shows the calculated activity of material transferred to

‘the Melton Valley Waste Disposal System. The-3H7actiVity correlates to

reactor power fairlyVWell as being one curie of 31 for every 45 Mwh of
power operation. - -
The equipment in the 11qu1d'waste system worked quite well Some

difficulties were encountered in keeplng the bulldlng sump level switches

 

%
See Section 12 on Water System.
249

Table 13.1 Liquid Waste Tank Transfers
- And Total Contaminants

 

 

$ Tritium I FP Vol

 

Date - — -

Curies , Curies Gal.
2/5/66 - = 0.092 5030
T7/18/66 141 —— 1800
10/23/66 10 0 8850
9/3/67 45T —— 5060
3/31/68 868 — T000
6/24/68 _— 0.031 5020 -
1/10/69 71 0.005 Thko
7/30/69 488" 0.0002 6500
10/8/69 126 6400

 

aligned. These were operated_by floats with 5 feet vertical tube which
actuates an ON-OFF switch. Misalignment usually caused the float to "hang"
and the pump would fail to shut off on.low water level,

On several occasions, the water level in the charcoal bed cell was

lowered so the inlets to the béds-could be heated. This was accomplished

. . by draining NlSOO gal. of water into the sump and allow1ng the two sump

 

 

o pumps to pump the water to the catch basin.
.t : Dur1ng the resactor constructlon, there was only one bulldlng sump pump
- - and 1t falled causing the coolant drain cell to flood through a floor drain.
:(The CD cell and sump room floor are at the same elevation, 820 ft.) A
small sump was cut 1nto the coolant draln cell floor and-a float-operated
alarm and steam Jet were installed ‘The system was malntalned to warn of
: fallure of both sump pumps and prov1de the steam jet to keep the sump level
down during a possible electrical power failure.

A second sump pump was ordered to replace the unit which was left from

earlier building operations. Before'receiving the new pump, the original

 
 

250

one was overhauled and the new pump was installed to prévide an operating
spare. The tritium transferred_through~this system was mlo%_of the dis-

charge from ORNL and no effort was made to limit the rate of discharge.

‘However, for a large breeder reactor where larger quantities of 3H would

be involved, some control of the tritium release would be required.
If air-operated sump jets are required in other reactors, care should

be used to insure that the ejJectors have sufficient capacity.
 

 

251

14, HELIUM LEAK DETECTOR SYSTEM
,J' K. Franzred

14.1 Description and Method of Operation

A lesk detector_system‘wss-used.to monitor all flanges in the MSRE
system which could permit- the-escape of radioactive materials. In addition,
all flanges whioh had to be maintained by renotely-operafed'tooling were
provided with leak—deteoted'type Joints to serve as an indication of satis-
factory reassembly. The 119 lesk-detected flanges were pressurized with
helium by 65 lines from 8 headers. As shown on Fig. 14-1, some leak-
detector lines were used for more than one flange. The principles of con-
struction of a typical leak-detected flange closure is shown on Fig. 1h-2,

In normel operation, the entire system was 1nterconnected and the
helium supply was valved off. Periodic repressurization was necessary to
keep the pressure within the limits of 90 to 100 psig. In the event of a
lesk, helium flowed into the affected systems and the resultant loss in
pressure in thelleakrdetector headers actuated an alsrm in the control room.

If the records indicated that-the rate of pressure decrease was greater

 

~ than the 0.66-psi/hr arbitrary limit, the headers vere isolated to narrow
down the location of the leeklng flange. . The- leak—detector lines were then
-velved offrindividuelly orfin_groups until the location of the leak was de-
termined. A reference volume and a differential pressure transmitter was
provided to aid in.isoleting leaks end for.determining leak rates. The al-.
B loweble leak rate from each flange or group of flanges was set at 1 x 10-3

'flf'cc/sec.
.”dlh;éflcelibration

The volumes in the leak detector system were determlned by pressuriz-
fing various parts and then equallzlng these unknown quantlties of gas with

,~an accurately callbrated known volume ~ The unknown system volumes were

"f calculated by using the fOrmula Plvl + PoVy = P3(V1 + Vz) - Leak rates were

 

fceleulated using these volumes and the rate of pressure decay from these .
;sect1ons of the system._ The volumes of various perts of the system are

given in Table 1h-1,
 

 

1
:
!
1
]
1
]
i
j

 

 

 

252

ORNL DWG 64-8834

ANY LINE REQUIRING

 

REACTOR

 

/‘ LEAK DETECTOR SERVICE
COMPONENT

 

 

 

Fig. 14.1 Method of Utilizing One Leak Detector Line to Serve Two

 

REMOTE DISCONNECT FLANGE

 

 

! PERMANENTLY

i o e = = e — e —— INSTALLED PIPE
LEAK DETECTOR LINE
! FROM LEAK
l_.._—<-zoerecgon

Flanges in Series
 

LEAK DETECTOR
" LINE FROM LEAK
DETECTOR snmou] A

 

e

 

L ASSKSSS

 

 

ORNL DWG 64-8833 _

 

 

I/////// —SEALING SURFACES
' _—0-RING GASKET
S SRt Ea S

 

 

 

 

 

 

 

 SEALING SUREACES‘ : ‘

 

 

 

 

 

 

FLANGES

Fig. 14.2 Schematic Diagram of Leak-Detected Flange Closure

£6cC

 

 
 

 

254

Table 14,1 Volumes in Leak Detector System

 

Volume of Line 400 = 184 cc |
Volume of High Pressure side of DP cell = T30 cc

Volume of Header L0l
Line 410
Line Ll1l
Line L12
Line 413
Line L1k
Line 415
Line 416
Line 417
Line 418
Line 419

Header 402
Line 420
Line 421
Line 422
Line 423"
Line L2k
Line 425
Line 426
Line 427
Line 428
Line 429

Header 403
Line 430
Line 431
Line L432
Line 433
Line 434
Line 435
Line 436
Line 437
Line 438
Line 439

Header LOL
Line -440
Line L4bL1
Line bLk2
Line L4L3.
Line Lul-
Line uUlL5S
Line LuU6
Line LhT
Line 448

‘Line Lho

nnanmaunnnnn

nmn o unnrnnnnn

HuwRnoD nmnunnnn

nmwwwnnwnn wono

156
24y
216
224
144
65

- OPEN

OPEN
88
OPEN

OPEN .-

1l

130

OPEN

111

125

120
115 .

115

1%9

115 -
138
135

ce ’ Volume of Header
: Line

Line
Line
Line
Line

~ Line
Line
Line
Line

_ _ Line
ce . ] Heeader
. ' : - Line

Line
Line
Line
Line
Line
Line
Line

- ‘Line
. Line

cc - ‘Header
| : Line
Line

Line

Line

Line

Line

- Line

Line

© Line

-Line

Header
Line
Line
Line

" Line
Line
Line
Line
Lyne
Line
Line

405
450
451
Ls52

453"

L5l
455

456

457

1458

459

406

460
461
462

463

L6k
465
L66
L67
L68
L69

4OT

470
471
472
473
uTh
L75
476
LT

478

k79

Lo8
480
481

k82

183
L84
485
186
487
488
489

HNWoR NN nuNnoN

e nnu

nmunwonanuwnwunan

HM BN NN

1hh
OPEN
113
108
106
109
104
109
102
99
98

148
154
99
148
99
107
108
OPEN
OPEN
OPEN
OPEN _

150
OPEN
291
235
518
143
135
138
109
305
L66

1h7
14k
159
123

123

310
210

SPARE
SPARE
SPARE
SPARE

 

NOTE: The "B" valves in the leak detector lines remained

all tests.

open during
255

| _lh.3”,0perating Experience

During construction and pre-operating leak testing, some difficulty

was encountered in attaining acceptable leak rates on some flanges, par-
'ticularly the very large salt freeze flanges._ During this and subsequent
"maintenance periods, the leak detector system proved very valuable in

-checking the progress and degree of tightness of all flanges.

Operating experience showed that once a flange was made tlght, 1t

: stayed tight during serv1ce.r During the entire life of the MSRE no leak

-ever developed that was serious enough to cause a shutdown of the reactor.

Some problem was posed by a freeze flange (FF-201) located near the
point where the coolant salt left the heat exchanger in the reactor cell.
It leaked more than the max1mum allowsble rate of 1l x 10‘3 cc/sec when salt

- was not in the piping and the coolant system had been cooled down. This

leak was measured at sbout 0 2 cc/sec on August 2, 1966 vhile the system

- was drained and cold. The flange always became acceptably sealed when the

- system was heated prior-tofiputting'molten salt into the piping and remained

so when the reactor was at power. The sealing action was probably due to

the thermal expansion of the 1nterna1s of the flenge being greater than

.that of the external clamp.m This would cause an 1ncrease 1n bearing pres-

sure on the seating surfaces of the metal "O" ring.

<

During normal operation with the entire system interconnected the

. pressure decreased at about 1 p51 per day, thus_requiring_repressurizing
at 1ntervals of one to two weeks. In a 52—day period of system operation
at temperature in October and November 1968 the pressure decay rate was -

v_il 2h psi/day. In a 60-day period of power 0perat10n in September and Oc-

tober 1969 the decay rate was 0 97 psi/day. During a system cooldown on

:f_November 30, 1968, the 1eak rate daid increase to 2. 15 p51/hr._ Differential
. ;,thermal expansion combined with thermal pressure decay probably caused the
"thgh indicated leak rate.ifpjff" S

- ‘When an sbnormal leak rate occurred it would take about L to 8 hours

:ffto perform Aan entire survey which consisted of 1isolating each header and
- then each line of the. header found to be leeking., . Determining the leak

rirate of an indiv1dual line took about an hour or two, however, 8 rough

check to guide maintenance personnel could be obtained in about 10 minutes.

 
 

 

 

 

 

- 256

The individual lines, which monitored more than one flange, gave an
interpretation problem. It was not possible to tell at the outset if all
the ieak wes through one flange or appprtioned between multip;e flanges on
fihe same line. When this situation arose, the practice_beeame thdtfdf 51_
.ternately tightening one,flahge, then the other, incrementally, and running
a combined leak rate on the line,betweeq'these_operations. By ebserving
when improvement was made, the apportionfient of the_leak‘between-flangee
could be estlmated | |

It was necessary to exceed the maximum recommended torque 1n some in-
stances. The technique was used to overtorqne where necessary to seat the
sealing faces, then to 1oosen‘the'bolt5'and retorque them to the nominal
or, if necessary, to the maximumrvelue listed in Table 1h.2,..in a few ex-
| ceptiohal cases, it was deceséary-to exceed this maximum, Some_of fhe
metal O-rings had to be replaced with.goldpplated O-rings in 6rder to ob-
teinlacceptdble leak rates. Some of»the 1/2=in. bolts ih small-flanges,
were torqued as high ae 150'ft-1bs and operated at about 80 ft-1bs. The
maximum recommended torque was TO ft-lbs, ‘ |

The record of measurements of system end flange leak rates was kept
~ in a "Leak Detector Log Book" at the LKD control cabinet. Flange leak
rates, plus a reedrd of all work done involvingfany of these flanges, was
logged in the "MSRE Flange Log"_notebook. This latter log gave details on
each flange ofi a separate page. Information givefi was Flange No., lesk de-
tector Valve.No., size of bolts (or in the case of freeie flanges, the
force exerted by the removable clip in tons), riné_size, the maximum per-
missible and recommedded torque in ft-1bs., the location of the leak de-
tected 301nt .plus the date of action, the reason for the actlon, actual

torque applied, and any general comments.
1k.4 Discussion and Recommendations

The lesk detector system functioned setisfactorily as designed. Due
to the two sealing surfaces on the "O" rings, primary containment was main-
tained as long as the leak detector system overpressure was gresater than
any possible pressure that could develop in the primarydsystem. ‘Therefore,
leak rate limitations were set based on convenience inrmaintaining an ade-

quate overpressure.
 

 

 

 

257

Teble 1L.2 MSRE Flange Bolt Torque Chart

 

'Bolt  Threads - © Nominal . Maximum ~ Wrench

 

Dismeter Per inch’  Torque (ft-lb.)  Torque (ft-1b.) Size
1/2" 13 b T0 - 3/4"
5/8" 11 90 135 7/8"
3/4" © 10 150 225 1-1/8"
7/8" 9 240 - 360 1-5/16"

1-1/8" T 533 | 800 1-1/2"

1-1/2" 6 900 1200 2-3/8"

 

Tightening Procedure:

1. Torque bolts to nominal torque value following a crisscross
sequence. Approach final value in successive increments of 10 to
20 ft-1b. : ,

2. If leak exist, torque bolts to maximum torque value, back off
and retorque to nominal value per step 1.

3. Repeat step 2 as required, increasing nominal value by 10% each
successive step. NEVER EXCEED MAXTMUM VALUES without permission of the
MSRE Operations or Maintenance Chief,

 
 

 

 

258

Remote welding or brazing techniques should be used where maintenance
is infrequent. This should be more economical and would reduce the number
of details to be handled by the operating personnel.

When opening primary system flanges, it is desirable to have the lesak
detector vented to atmospheric pressure to minimize the spread of activity.
Each header should have a line which is connected to the cell'tqlfaéilitate

doing this since the gas in the line could be contaminated.
259

15, 'INSTRUMENT»AIRVSYSTEM
T. L. Hudson

-15:1 Description
The instrument air system supplied clean, dry, compressed air for
'pneumatic instruments and other spec1a1 uses. .Two Joy compressors, AC-1
and AC-2, were used to compress the air, which then passed through an after-
cooler and entrainment separator to a common line supplying two parallel
receiving tanks. From the receiving tanks, the air paSsed through one of
two parallel drying stations conteining Trinity heatless dryers. The dry
air was-distributed.through headers to locally mounted filter and reducing
stations. The serviceiair5compressor; AC-3, could be valved into the in-

strument air system upstream of the receiver tanks if required.
15.2 Operating Experience

The capacity of the instrument air system.withlone,air compressor in
.operation was edequate'at all times. Normal air usage was about 70 scfm.
- Of this amount, 18 scfm Was_used\for purging the instrument eir dryers, -
| 12 scfm for supplying noncritical instruments, and L0 scfm for supplying
. the ‘eritical instruments. The critical instruments were supplied by the
emergenCy instrument air supply, which consisted of two banks of six ni-
trogen cylinders, when -the normal instrument air supply was lost. |

The- capa01ty of the emergency instrument air supply was tested during
normal operation with flush saltrto assure an orderly shutdown of the re-
~actor in case of loss of:bothfinstrument air compressors (see MSRE Test
' Report‘2.3.ll.3)._,With an'emergency air usagelof hQ scfm and,one bank of
:cylindefs'at-their lowestHOpereting, 1500 psig (initial low-pressure alarm
point) and the other ba.nk fu.'Ll (meooo psig) at the time both air compressors
| stop would provide the following times:
| (1) Air in the-compressorrsurge tank would last i min,

(2) First bank of nitrogen cylinder (if used to TOO psig — at this
-pressure the control valve stopped controlling) would last V13 min,

(3) Subsequent full benks of nitrogen cylinder (if used to TOO psig)
would last m20 min.

 
 

260

From the test it was concluded that'the'emergency instrument air system
was adequate. It would have been necessary to install a larger control
valve to more fully utilize the nitrogen cylinders. |

-,'Leaks in ihe emergescy nitrogen headers were repaired several times
and the control valve was repalred once. On the average one bank of nitro-
gen . was changed out every two weeks to keep the cyllnders above the 1nitial

alarm pressure of 1500 psig.'
15.3: Conclusions -

~The operation of the‘sormal instrument air system was very satis-
facfory.._At so time was it necessary. to use the emergencyznitregen system
'dfi;ing eperstion of the reactor. One instrumeht air'compreSSor was oper-
ated with the other in.standby except fbr.during'repairs or maintenance.
Even with a complete power outage, it was possible to ‘start one air com-
pressor on diesel power before the receiver tank pressure dropped ‘enough
to cause a transfer to the emergency-system. Although some extensive main-
tenance was done on the air’ compressors, only one was ever inoperative at ' Q;j

g time.
 

 

261

'16.  ELECTRICAL SYSTEM-~ - -
T, L. Hudson.

16.1*-Deseriptionr

‘The- MSRE electrlcal power was supplied from the ORNL substation by
elther of -two 13.8-kV TVA lines, e preferred line or an alternate. These
| were separated by—two_interleeked-pole—mounted motor-operated line ‘switches,
{'-Automatic transfer was poryideg from the preferred line to the alternate
line. Both,transfertswitches-could-be manually operated remotely from-the
auxiliary control room. . | |

The ac power;enteredtheiMSRE building through two transformer sub-
stations. A 1500-kVa, h80-V373-phaSe substation served the process equip-
ment, and a 750-kVa, 480-V, 3-phase auxiliary station was initially for
‘building services, such as lighting, ventilation, ete. After the process
'SUbstation becane overlbaded bus 5 whieh_Supplies electrical heaters was
connected to the aux111ary system. Three diesel-generators were installed
for emergency ac use. _ '

There were two separate ares de systems, a h8—V system and a 250-V
system. These were normally operated by ac-dc motor—generator sets and
had battery supplies for emergency use.

The L48-V dc system,prov1ded power for electrical controls and one
channel of the reactor safety systems. The system consists of two 3-kW
~ motor-generator sets (each rated;at'53-ampere at-56 V) and & bank of 48-V
batteries (600 amp-hr). The two generators normally supplied the dc power
and- charged the'battéries.'rln an emergency, when no ac power was availsable
.—to run - the motor—generators, h8—V de power was supplled dlrectly from the
batteries. ' '77 o | | |

~ The 250-V dc systemfprevidedlpower=to the relisble power system, pro-
cess breakers trip power, 13'8ekV transfer control pcwer;rand'emergency'
lighting. The system consisted of one 125-kW motor-generator set and a
bank of 240-V batteries (36h4 amp—hr)
7 The relisble power system prov1ded power to certaln 1mportant process
equipment and one channel of the safety systems.' Inltlally, the pover was

supplied from a 25-kW dc-ac motor-generator set. The dc motor was supplied

 
 

262

from the 250-V dc system and the generator supplied the 120/2h0;V ac re-
liable power system. The operation-ofthe MG set was unreliable because
of failures in electrical control componefits,and.was replaced with a static
inverter (see 16.4).. o

- The MSRE electrical 1nstallation made extensive use of the electrical
facilities that were installed for the ARE and ART operations in the 7503
- area, - Many modi fications. were made.to-the existing equipment, but only a

1imite§'am0unt'of supplemental equipment was required.

16.2-‘A1ternatifig Current System

| The operation of the ac electrical system.was very . satlsfactory. There
rwere eleven important failures, ten of which caused unscheduled interrup-
tions in the power operation of the resctor:. five were-caused by electrical
stoims,flone by a cable failure at a componenp-cooling pump, one by an over-
load of the meih process-power breaker, one by an arc fietween a 13,8—kV_line
and en activator rod to the line fuse, one by a failure of the main trans-
fbrmer prifiary fuse, one by a failure of the auxiliary transformer primary

fuse, and ofie by a failure of the drain-tank space'copier motor. These
| interruptions varied in. length from a few minutes to several days. In all
cases the necessarj repairs or modificationsrwere mede, and no damsge to
reactor equipment was incurred. These are discussed in more detail below;
On two occasions while the reactor was operating at power, momentary

electrical outages during storms caused control—red screms by range-
switching the nuclear power safety channels because of dips in the fuel-
pump-motor current. In both these cases the nuclear power was quickly re-
stored to the value thet existed just prior to the outage. Since the
eafety analysis ihdicated that low range of the safety channels was needed
only while filling the reactor (when the fuel pump was off), rapid response
- of the range-switching fias not required Therefore, time-deley relays were
‘subsequently 1ncorporated in these circults to prevent their activatlon on
momentary power dips. | . ,
- Another storm-induced failure occurred when lightning struck the main
ORNL power eubstation and parted one wire of the normal MSRE feeder. 1In

this caese the electrical load was automatically transferred to en alternate
263

feeder, and all essential'equipnent'was restarted in time to prevent drain-
ing either the fuel or coolant system. However, operation'at'high nuclear
power could not be resumedluntil service was restored on.the mein  feeder.
The entire 13.8-kV supply system was subsequently revieuEd‘and improved bo
make it less susceptible to damage by storms.

During a thunderstorm in June 1967,_the reactor operation was inter-
‘rupted by the loss of both feeders. The reactor period safety system func-.
tioned properly and,scrammed the reactor but bwo amplifiers were damaged.
Approximately 32 hours later, the reactor_was returned to critical operation
after the period safety amplifiers had been repaired.

In July 1967, an interruptlon‘was caused by the loss of the preferred
feeder during another thunderstorm. Emergency power from the diesel gener-
ators wasuin‘service'uithin 2 min, and low-power nuclear operation was re-
'sumed‘in about 13 min. After repairs had been completed on the preferred
feeder, full-power operation was resumed in- approxlmately 6 hr.

The short in the component-cooling pump cable seal is deseribed in
 Sect. 10 entitled "Component Cooling System." ~When this short occurred,
there was a messive flow of current and the breaker supplying the entire
bus tripped before the individual bresker for the motor.

Durlng the initial full-pcwer operation of the reactor, the main
process~power breaker (breaker R) wes loaded to near its setpo1nt value.
When the load on the coolant-radiator main-blover motors was increased by
increasing the pitch on the blower fan blades, the increased load on
bresker R caused it to trlp,on_overcurrent. Th1s condltion was relieved
by transferring bus 5 (v200 kW) from the main process transformer to the

auxiliary power'transfbrmer."fThe load transfer decreased the current

- through bresker R from sbout 1700 amp/phase to about 1550. =

A building power fallure occurred on a foggy mornlng in October 1967,
when an arc developed between a 13. 8—kV line to the main transformer for
“the building and parallel: metal activator rod to the 13. 8-kV: line fuse.

The cause of the arc is unexplelned Although the weather was foggy, 1t
was not unusually wet, ‘This- fault caused the operation of the preferred
feeder overcurrent.ground-relay located at the ORNL_substation, vhich
tripped the feeder breaker before the line overcurrent relay located at the

“MSRE -could operate to prevent an automatic transfer to the alternate feeder.

 
 

264

When the MSRE was transferred, the alternate feeder overcurrent ground Te-
lay. operated and the alternate feeder bresker also tripped. (To prevent |
an automatic transfer to the alternate feeder on a similar fault, an over-
current ground relay was later installed at the MSRE.) After an interrup-
tion of 37 min, low-power nuclear operation was resumed on emergency elec-
trical power from the diesel generestors. The_damage:wasrrepaired, the
spacing between the line and the ectivator=rod‘was increased, and normal
service was restored after an interruption of 2 hr, I

In April 1969, there was a failure of a fuse on the primary side of
the MSRE main power transformer. This resulted in the loss of voltage on
one of the transformer windings, The power supply was ‘changed from a three-
! phase to-a single-phase supply. This caused a drop in secondary voltage
ecross two windings and increase-in voltage across the other winding. - The
motor-operated.equipment continued'to operate until the individusl breskers
were tripped by the high current in two of the lines. All breakers did not
trip at the. same time vhich made it difficult to locate the trouble im-
m.ediately° The reactor was scrammed from full power and the diesel gener=-
ators were started, but both the fuel and coolant systems were drained
- before cooling air could be restored to the freeze valves. No apparent
. cause for the fuse failure was found other then possible overheating from
| poor electrical contacts..

About the middle of February 1966 a ground developed in the drain-tank
cell space cooler 10-hp motor after about 2900 hr of operation° “An iso-
lation transformer was instelled to isolate the ground from the electrical
system. After a few hours of operation, the motor stopped and couldn't be
-restarted. The reactor was teken subcritical and the fuel system was
drained. The drain-tank cell was opened to provide cell cooling. The
- cooler and motor were removed,'and-examinetion'showed:that the rotor had
slipped. along the shaft until the rotor fan blades rubbed, causing the .
stator windings‘to‘be\destroyed. The feilure appeared to be due to im-
proper manufacture and not e design weakness. Therefore, the damaged motor
was replaced with one prectically identical. | )

1 On one occasion the auxiliary power system was out of service for
about 2 hrs due to & failure of the auxlliary power transformer primary

fuse.. During the fuse replacement the operation of the reactor continued
265

at power with bus 5 supplied from diesel-generator 5. The fuse failure
apparently was caused by ovefheating from poor electrical contact. Due to
low resistance readings, the three auxiliary power transfers were replaced

in 1968.

16.3 250-V dc System

The operation of this system has been very satisfactory. There have
been some tube failures in the automatlc voltage control system for the
generator, but each time the unit was operated on manual control until the
tubes in the,automatic'controls could be replaced.

Once a month the charging voltage was incressed to 275 V for 2k hrs to
give the batteries an equalizing charge;

The batteries were given a 3-hr life test under full load. The vcl-
tage dropped to 215 volts over_the.three-hour period. Extrapolating the
voltage curve indicated that the lower limit of 210-V would have been
reached in %3.5 hours which is well sbove the 2-hr 1ife upon which the de-

sign was based.

16.4 Reliable Power System

The 120-V ac relisble power system was originally supplied by an MG
set., In early operation, this MG set (originally installed for the ARE)
proved very unreliable due to fallures in the electrical control components.
To 1mprove the rellabllity -and -increase the capaclty, the motor generator
was replaced with & new 62—kVa three-phase static 1nverter. The total load
on the reliable power supply, 1nclud1ng_the on—llne computer, which has
some three-phase 1oad ‘was . hO kVa. The‘etatic inverter offered & number
| of adventages over the old rotary inverter, including higher efficiency
'?(1ncreased 35%), smaller size, no moving parts, quiet operation, and ex-
cellent voltage and frequency regulatlon. Prior to the installation of
~this equlpment both ‘the capacity and voltage regulatlon of the reliable
power supply were 1nadequate for;the operatlon of the on-line computer.

a During the'checkofit:and”testing_of the static inverter in April 1966

after the installation had been completed, trouble occasionally developed

 

 
 

 

266

that blew the load fuses. Two times the manufacturer's field engineer made
exhaustive tests and could not find any trouble but all symptoms indicated
that the trouble was in the 10w—voltage logic power supply, ‘which supplies
2h—V dec control power. On April 25, the inverter failed again whlle the
reactor was operating at low power. Apparently the transfer caused a tran-
-sient on the ac system, thus also dropping out the rod scram relay supplied
from the ac system, causing a control rod scram. The inverter load was
automatically transférred to the normal ac pover supply Aftsr this failure
& new power-supply module was installed 7' -

The inverter output voltage regulatlon was 208 1/2 v, ané_the output
frequency better than 60 cps 0. 01%. The inverter was tested for 2-1/% hr
‘operating from the 250-V battery system.without the dc generator operatlng°
Although the input to the inverter varled 20 V the output varled only
0.5 V. _

There have been three other failures associated with the static in-
verter that have aufiomaticaliy transferred the normal inverter load to TVA
supply. The first was caused by failure of load fuses which rssulted in an
unscheduled control-rod seram. The second was caused by a failure of &
thyrister in the inverter logic circuit. The third was caused by & momen-
tary ground during meintenance on the com@fiter.

When a transfer occurred, manyfqlarms vere received in the control
room. Somslwere caused by loss'of'oPsrsting equipment, and others were
due to the momentery loss of control voltage to_cerfiain monitors. These
had ts be reset to stop thé alarm and control action. After the second
invepter_failure, all‘operating_equipment was restarted, but an emergency
fuel drain occurred before the reactor-cell air activity monitors were re-
set. No.drain or other seriqus condition resulted from the .other inverter

failures.

16.5 L48-V dc System

The 48-V dc system was very reliasble. The only maintenance on the
system was caused byrexcessive arcing_of the generator brushes. ‘Initially
- one generator was operated near rated capacity to supply the load., Filters

and viewing windows vere installed on the generators and both generators
 

267

were operated in parallel"to reduce the individual generstor current after
the load: sllghtly exceeded the capacity of one generator. These changes
greatly reduced the generator brush prdblems. | |

Once ‘a month, the charglng voltage was 1ncreased to 5k volts for m8 hr
to give the batteries an equa1121ng charge. _ .

The batteries were given a 6-hr life test. The first 3.5 hr was at
47.5 amps and the last 2.S_Hr-et 19.5 afips. The batteries expended 215
amp-hour of the 600 amp—hour life with essentia;ly no voltage drop.

16.6 Diesel Generators

 

The three diesel-power generators, each With.a capacity of 300 kW,
supplied emergency ac power‘to motors and heaters in the MSRE. -They proved
to be quite reliasble. They were started and operated unloaded for sbout an
hour each week. Once a month theyrwere tested under load. They never
failed to start when required during a power outage. Under emergency con-—
ditions, they were usually running within 2 min after a failure of the nor-
‘mal power supply. o

In February 1966 a crack was found in the block of DG-3 at one of
the bolt holes. Attempts at repairs were not successful, but the unit was
operable: Later DG-3 waé-replaced with a surplus diesel generator from
another facility. Thegreplacement unit was similar to the original except

that it was started by compressed air instead of by an electric starter.

16.7 100-kVa Varisble Freqpency Motor-Generator Set

A temporary-installatiofi'of a used variable frequenoy MG set was made
at the MSRE to supply the”ffieiépufip motor during special"tests; The fuel~
pump motor was operated from 610 rpm to 1165 rpm (maximum speed.w1th MG
set) over a perlod of several weeks.

The automatic controls for the variable speed and generator voltage
did not vork properly durlng the 1nitia1 checkout u31ng a dummw resistance
load. After replacement of some ‘defective components in the control cir-
cuits, the fuel pump was operated from the MG set. The operatlon of the

.automatic controls were still not completely satisfactory and some manual

 
 

 

268

adjustments were required. Frequently the fuel-pump motor waé”stopped,be-
cause of over adJustment of the manual controls. Also Qufing_phe lower-
SPeed opération,rfihe speed of the fuel pump wasrnot:constafiéaenough tdlfiaké
the desired tests. Therefore, the automatic controls werenfiypassed afid‘;
manual controls of speed and voltage were installed.. Satisfactory operec
tion with manual controls was not dbtained-until the fuel pump_fiaSaoierated
at about.35 V above fated volfage,'assuming the fuel-pump motor voltage is

directly proportional to speed.

16.8 Conclusion

Considering the agefand_the'numbér_of modifications made to the exist--

ing electrical system within the MSRE building, the performance has been
very satisfactory. The main weakness of the entire electrical system was
with thekl3°8—kV feeders and transformers supplying the MSRE. Most of the

eqnipment'wasiover 15 years old. - During 1966, there were three momentary

_interruptions of the power to the MSRE caused by thunderstorms. The entire

13.8-kV supply system was reviewed and improved to meke it less susceptible
to damage by storms. Since the improvements were made, there have been

only two interruptions due to storms in the last three years.of operation.

i
;

¢
 

 

17. HEATERS
T. L. Hudson

17.1 Description

Electrical heaters were provided for all parts of the salt system to
permit preheating of the circulating loops and to maintain the temperature
of the salt at 1200°F during zero-power operation. (In practice the
heaters were kept energlzed durlng pover 0peratlon as well, ) The total in-
stalled capacity was 900 kW the actual pover requirement was less than
half this amount. It was p0551ble to supply all heaters essential to an
orderly shutdown, including enough to keep thy salt molten in the drain
tanks, from a 300-kW diesel-driven generator. A summary of the heaters
and insulation on the salt systems is given in Table 17.1.

The type of installation was determined by the location. In the re-
actor and drain-tank cells, radiation from long-lived fission products pre-
cluded direct maintenance after start of power operation, There, the heat-
ers were deéigned for removal and replacement using 1ong¥handled tools in-
serted through the cell roof. In the coolant cell, the piping could be
approached within a few minutes after a shutdown, and the heaters and insu-
lation were the conventional type used in component test facilities. -

Removable heater units 1n the reactor and draln-tank cells were of two
types. On the piping and heat exchanger, the heaters were nichrome ele-
nmnts in ceramic plates, located on the inner surfaces of the removdble
units that also insulated the sides and top of the equ1pment. On the re-
actor vessel, fuel pump,,and drain tanks, the removsble heaters were in-
‘serted into penetrations in pérmanently mounted insulation. All electrical
insulation.was ceramic tyfie;jbééause.of themhigh gamma~radiation fields (up
to 105 R/h in the reactor cell Quring operation). |

Thermal insulation 1n the removable heater-insulation units was of the
metalllc mnltiple-layer, reflectlve type, which could withstand handling
w;thout dusting. The permanently mounted insulation was 1ow-conductiV1ty,
high temperature low sullfur contént, ceramic fiber of expanded silica,

with a thin metal sheet between the insulation and the pipe surfaces (see

ORNL Drawing E-MM-B-51676).

 
 

 

Table 17.1

Heaters and Thermal Insulation

 

 

 

Type Number | Rumber
Type Thermal Heater  Heater d Number Pover
Location Systenm Heaters Insulation Units Elements Type Heater Elements Controls  Supply
Reactor Cell Reactor Removable Permanent? 9 9 1135 £t of 3/8-in. by 0.035-in. wall 3 N
: Inconel tubing. Each removable
unit consists of T U~tubes.
Fuel Pump Removable Permanent® 5 14 3/h—ih.-d:lam, atainless steel sheath 2 N
5-ft long with ceramic insulation
~and nichrome elements.,
S5-in.-diam Piping Removable Removableb 31 17T Cér_amic plate with nichrome elements. . 23 N
. Permanent  Permanent® 3 8 Tubular 0,315-in. OD Inconel sheath 3 N
vith nichrome elements.
Heat Exchanger Removable Removableb 3 2h Ceramic plate with nichrome elements. 3 N 8
Freeze Valve Removable  Removable® 1 1 Tubular 0,315-in, OD. Inconel sheath 1 N
with nichrome elements.
Reactor Neck Permanent Permanent? 2 2 Tubular 0.315-in. OD Inconel sheath 2 N
‘ with nichrome elenments,
Drain Cell Drain Tanks Remva’ble Permanenta 23 156 Cc_era:nic pla.tes with nj.chrome elemenis 6 N&E
1-1/2-in.~diam Removable Remova.bleb 13 | 7'5 Ceramic plates with niéhrome elements 15 N&E

Piping (Fill
Lines)

 

%Ceramic fiber of expanded silica.

bMeta.llic, multiple-layer, '

dCere.mic plates — maximum install watta.ge - 30 wa.i:im/in2 of surface. Tubular 0 315-1\1. OD - normal . ra.ting 500 watts/ft of heater,
N - noyrmal power supply; E ~ emergency pover supply.

-
Table 17.1 (Continued)

 

 

: ' Type ° Number  Number
. . . . Type Thermal Heater Heater o ' d Rumber Power e
Location - Systen Heaters Insulation Units Elements Type Heater Elements . Controls Supply
Freeze Valves Removable Remvableb 3 27 Ceramic plates with nichrome elements 9 N&E |
{Fill Lines)
Fill Line See Note ¢  Permanent® 1 1 Direct resistance heating of 1-1/2-in. 1 NSE
' ' ' by 0.195-in. wall INOR-8 pipe 65 ft
. long.
Drain Cell 1/2-1n.-diam ©.~  Permanent Permanent® 12 32 Tubular 0.315=in. OD Inconel sheath 12 RAE
: Piping (‘l‘ra.nsfer : T ‘ with nichrome elements
11nes) ‘
Freeze Valves - . ‘Permaneént Permanent? 3 32 Ceramic plates with nichrome elements 6 FSE
: ('I'ra.nafer li;xes) ‘ :
Coolant Cell Cdolant Pump Pema.flént Permanent? .2 1L . Ceramic plates with nichrome elements 2 N
5-in.-diam Piping Permanent Permanent? 10 87 Ceramic plates and tubular 0.315-in. 10 N
R ' : 0D Inconel sheath with nichrome
elements,
Rédia’tor ‘ Permanent Permanent® 8 162 Ceramic plates and tubular 0.315-in. 8 N&E
OD Inconel sheath with nichrome
elements,
Drain Tank Pei'ma.pent Permanent® 3 2y Ceramic plates and tubular 0.315;-in. 3 N&E

 

%ceramic fiver 6f'expanded silica,

bMetallic, multiple-layer.
®The £111 line was heated by passing current through the wall of the pipe,

dCera.m:lc plates — maximum install wattage — 30 watts/in of surface. Tubular 0.315-in., OD = normal rating 500 watts/ft of heater.

®N — normal power supply; E — emergency pover supply.

0D Inconel sheath with nichrome
elementg.‘

The pipe wes removable.

Tl¢

 
Table 17.1 (Continued)

 

 

Type Number Number
Type Thermal Heater Heater d Rumber Power e
Location ‘System ‘ Heaters Insulation Units Elements Type Heater Element Controls  Supply
Cooie.nt Cell Freeze Valves Permanent Permanent® 2 8 Ceramic plates with nichrome elements. h NAE
(continued)
Fill Lines Permanent Permanent® 6 30 Ceramic plates and tubular '0.315-131. 6 NAE
0D Inconel sheath with nichrome .
elements. - ‘. ‘
Flow Transmitters Permanent P’ern:lanem;‘:z L 8 Tubular 0,.315=-in. 0D Inconel sheath 8 N
with nichrome elements.
Level Element Permanent Permanent® 1l Y Tubular 0,315-in, OD Inconel sheath 2 N

with nichrome elements

 

aCeramic fiber of expanded.asilica.

dCeramic plates — maximum install wattage -~ 30 watts/in? of surface. Tubular 0.315-in. OD — normal rating 500 watts/ft of heater.

®N — normal pover supply; E — emergency power supply.

C

¢l

 
Maintenance requirements were minimized by designing the heaters with
excess capacity and operating them at reduced voltage. If some of the
heater elements failed, the other elements could be operated at a higher
voltage or the power could be increased to the adjacent heaters. Heater
units that were less accessible had spare heater elements installed and -
connected, ready for use with only minor, out-of-cell wiring changes.

Control of the electrical input to the heaters was entirely manual, in
response to system temperaturee displayed on either a multipoint tempera-
ture recorder or a csthode-ray tube by a multipoint scanner. At least one
thermocouple was attached to the piping or equipment under each heater unit,

with additional couples at potential cold spots.

17.2 Preoperational Checkout

To assure that the heaters would perform'properly and to provide ref-
erence data for future trouble-shooting, a number of tests were done before
and after installation. A visual examination of the hesters and the wiring
in the Junctlon boxes was made and all obvious errors were corrected. Con-
tlnulty checks were completed_and measurements of the heater circuit re-
sistences and resistances toiground were recorded. Each heater circuit was
then energized‘individually to assure that the proper thermocouple or ther-
mocouples responded. |

The effective resistance of each heater circuit was calculated and
these were compared with'measured4resistance of the heaters before and af-
ter being installed in the cell Several of the resistance measurements
were made u51ng 8 volt-ohmymeter but the accuracy of these readings was usu-
ally not good enough. The flnal resistance measurements were made with &
wheatstone bridge and were recorded to the nearest one—hundredth ohm,
| Teble 1T.2 is a summary of repairs mede to the heaters during the pre-

operational checkout. .

17.3 System Heatup

The heatup of the coolant, fuel, and drain tank systems was started as
soon as construction and checkout of the systems were completed. During

the first heatup, &ll heaters operated satisfactorily and there were no

 

 
 

274

Teble 17.2 Summary of Preoperationel Hester Checkout

 

Repair or Replacement
Of Hesaters '

_Reactor Cell

Drain Cell Coolant Cell

 

Corrected wiring

Replaced open heater element

Replaced broken ceramic beads

on removable unit - |

odified removaeble heater
box to eliminate possible
short circuit in ceramic
bead insulated leeds

Modified heat box to permit
- remote meintenance

Replaced broke MI cable end
seal

Replaced MI cable

4

- 3
1 - 3
16 ¥
L ¥
2 *
1 *

 

*
-Does not apply to coolant cell,
 

 

275

heater failures, but there were a few areass where the temperature was low
 because of air leskage and lack of thermal insulation. These areas were
the inlet to the fuel-pump furnace, north end of heat exchanger, transfer
line entrance to the drain-tank furnace, and the coolant radiastor. After
the air leakage was reduced and thermal insulation installed, satisfactory .
temperatures,were‘Obtained'nith the heaters as installed. The voltage set-
ting required.for each heater was determined during the startup testing and
stops were set to 1limit the controls to hold the system temperatures to
1300°F or below.. The general:heatup of the systems and the difficulties
are discussed in the following three sections. '
17.3.1 Coolant and Coolant Drain.Tank Systems Heatupfi
| The initial heatup of the coolant system in. the coolant cell and :cool-
ant ‘drain cell was started on October 13, 1964. The coolant pump was run
during the heatup with normal. helium.purgewand,the heaters were adjusted
| to llmit the heatup rate. to: about 100°F/hr. Heaters located in the cell
, penetratlon sleeves on lines 200 and 201 were operated at a reduced set-
ting to provide a temperature gradlent untll the heatup of the reactor cell
- was started so as to reduce ‘the thermsl stresses at the anchor point,

Heater and thermocouple leads on top of the radiator enclosure were
observed to be overheatlng as the system was being leveled at a reference
temperature of 350°F. The: radiator was cooled to ambient temperature and
a section of the top of the radiator was found that had not been insulated.
| The insulation was installed and a second heating of the radiastor was
started. Some of the. radiator heaters were disconnected in an unsuccessful

effort to achieve a satlsfactory temperature dlstribu.tlon° The radiator

.. was 1073 to l2h5°F and the outlet tubes were 620 to 1000°F. This condition

f'was reached nine d&ys after the heatup was started. During this time the
'remainder of the coolant system in the coolant cell was: heated to 1047 to
_1259°F except on 11ne 202 between the radiator and the coolant pump. There
was hOO°F dlfference among temperatures along this sectlon of line under a
,31ngle heater control. The spread was reduced to sbout 175°F by replac1ng
part'of the permanent 1nsulat10n near the radiator outlet to reduce inleak-
age. The coolant drain tank‘was heated to sbout 1200°F without eny diffi-
culty with a temperature dlfference of 30°F between the highest and lowest

thermocouple readlng°

 
 

 

276

The radiator was cooled to room temperature for inspectibn end to re-
pair the doors and seals. After répairs were made, the radiator wés again
heated. With the best temperature distribution obtained, there was sbout
'8 500°F spread on the radiator outlet tubes. This indicated that a rela-
tively large heat leak existed. An inspgcfion wes made of the radiator
enclosure near the‘outlet heater and'sii compartments in the enclosure wall
 were found to be without insulation. After the propér insulation was_ih—
stalled and gaps in the vicinity of the outlet piping were ééulked, satis-
factory temperatures were obtained'with all hesters in serviée as origi-
nally designed. Localized overhesting in the areas_above the radiator en-
closure was again observed. The heat;;eaks‘were primafihy at'dfienings
through which the ceramic-beeded insulasted electrical'leads entered the
enclosure., Some minor insulation:damage was sustained on the wiring inside
the junction boxes mounted sbout 12 inches sbove the tqp'of the radistor
enclosure. m | | | | | | |

During the following shutdown,.laté in 1965, radiator doors of im-
proved design were installed. Heat losses due to air leskege around the
doors continued to present a problem. To 6btain ap-déceptéble'tefiperature
distribution, it wés necessary to shim the gaskets:and diSconnect some of
‘the heaters, so as to give a nonnuniform.heat input ovér'théffdée_of the
radiator. When the main blowers. were operated to test the doors, other
difficulties became apparent. One of these was leakage of hot air into the
region just sbove the radiator enclosure which caused ofierfieating of elec-
trical insulation on the numerous thermocouples and heater leads in junction
boxes located in this vicinity; It was necessary to relocate junction boxes
in coglér locations slong the cell wall and install higher tempéfature elec-
trical insulation on the thermocouple and heater ieads'over.thé énclosuije°

buring'the approach to full reactor power, heating the efipty radistor
to an acceptable temperature distribution prior to a fill hafl.bécome in~-
creasingly difficult, end on the last fill, some of the tubes could not be
heated above the mélting point of selt until the radiator was filled and
‘circulation was started. After the shutdown in July 1966, the radiator
door seals were modified. The electrical heat'required to hesat fhe empty
radiator was reduced from 108 kW to 62 kW by the door seal modification.

1
 

277

The four flow transmitters located in the enclosure above the radiator
required a higher heater setting when either or both blowers were on be-
cause of air leakage past the transmitter. |
17.3.2 Fuellsystem Heatup ,

The initial heatup of the fuel system and the coolant piping in the
reactor cell was started November 1, 1964. At this time the coolant system
in the coolant cell had already been heated.

About nine days were spent on the first heatup and leveling of the
temperatures. The heatup rate was limited by the heatup rate of the re-
actor because there is about 8000 pounds of INOR and 8000 pounds of graph-
ite to be heated. To aid in heating up'the graphite and to help purge out
any remaining oxygen, the fuel pump was run during the hesatup with normal
helium purge. Also the fuel system,plplng and heat exchanger were operated
about 150°F higher to help heat up the reactor.

The system temperatures were brought up to an acceptable range (1000
to 12SO°F) with a few exceptions. Temperature at the freeze flanges and
on the coolant lines at the cell penetrations were, as expected, beidw the
liquidus temperature of the salt, but only over short sections. Thermo-
couples under the hesater loeated between freeze flange 100 and the fuel-
pump furnace could be heated only to TOO to 900° until heat losses were
reduced by closing gaps in the heater base. There was about 300°F spread
between the coolant inlet nozzle and the fuel outlet nozzle on the heat
exchanger. The spread was_re@ueed_by closing gaps between. the heat-exchanger
'Vbase and adjacent heater. rThe;time required to heat up the reactor and fuel
system after the initial heatup was sbout three days.

17 3.3 Fuel Drain Tank System Heatup

The initial heatup of the draln-tank cell was started November 12,
1964, This heatup included drain tanks, FD-1, FD-2, FFT and lines 103,
| th 105, 106 107, 108 109, and 110 Thls was the last of the systems
heated. | | .

_  The helium purge was set up from the fuel system through line 103 to
the drain-tank system and then exhausted at line 110. B

| . The heaters were adjusted to limit the heatup rate to less than

100°F/hr. All of the system was heated to an acceptable range (1000 to

 
 

 

278

1250°F) except where lines (1ines7107, 108, and 109) penetrated the tank
furnaces. On line 107, the temperature was brought up to 950°F, but only
by operating the heaters in‘the_penetration”at 220% of_their_rating° _Ehe
other two line penetrations were somevhat better, but stil]‘.runsatisfactbry°
The insulation on these lines near the penetrations ?as removedt Then the
heaters were tied cldser to the pipes and additional insulation was insert-
ed between the heaters and the furnaee walls., Temperatures at these pene-
trations then reached 1200°F with the heaters at less than their rated

capacity.

17.4 Heater Performance

The system was operated at high temperature for sbout 10 months before
the prepower shutdown. Durlng the flrst 1500 hr of 0perat10n, only two
heater elements failed. Both of these were.in removdble,heater units on
the main fuel circulating lines and had spare elements in the ceramic heat-
er plates. The spare elements were out into service by out-of-cell wiring
changes., A summary of the heater failures and repairs are included in
Tables 17.3, 17T. h and 1T7.5. Slx add1tional heater-element failureg and
an electrical ground at s disconnect were discovered durlng the cheékout
for Run 2. In four of the heaters, the failures occurred at the junctlon
of the exten31on lead and the heater element inside the ceramic. pla.te°
These elements and the electrical ground were repalred'before startup, and
the other two elements were left out of service for Kuns 2 and 3.

During subsequent operation, one in-cell heater failure was noted, and
one failure occurred at a heater power supply. " (The latter was repaired
immediatelyo) .It may be noted that & number of heater failures could occur‘
without significantly affecting the operation of the reactor system.

| Three more heater-element feilures were found during the prepower shut-
down., In addition, a number of minor defects (grounds damaged connectors)
were found. All the defects, 1nclud1ng those left from earller operations,
were corrected.

An important cause of ceramic heater-element failure has been separa-
tion of the extension lead from the heater-elementor This is apparently

due to & combination of a design weakness and excessive flexing during
_Table:lTQB Summary of Heater Failures in the Reactor Cell

 

 

 

 

elements failed.

Date System Heater Type Failure Repairs Made
CRITICAL EXPERIMENT PERIOD

'11/9/6h4 Reactor R-1-2 One U-tube grounded U~tube shortened 3 in.

12/31/6k4 5-in. piping H-100-1 Failure of top element Connected up spares to continue Run 1.

1/5/65 | 5-in, piping H-102-5 Failure of top element Connected up spares to continue Run 1.

3/22/65 5-in. piping H-100-1 Replaced ceramic elements before start of Run 2.

3/22/65 S-in,;piping‘ H-102-5 Replaced ceramic elements before start 6f Run 2.

h/30/6$f  5-in. piping H-201-4B Lead grounded . Realigned disconnect.

T7/20/65 . ~ Heat pxchanger  HX-2 Lead grounded Removed sharp edge in lead.

| PREPOWER SHUTDOWN

T/30/65 Fuel pump FP1&2 Two fuel-pump heater units resistance increase from 12 to 1T ohms
since the preoperational checkout. These were removed from the
reactor cell and the mechanical Joint between the solid and
flexible nickel lead was highly oxidized. All five fuel-pump
heater units were removed from the reactor cell and a welded

: | connection was made at each lead extension.

7/30/65 5-in. piping H-102-1 Broken disconnect replaced.

7/30/65 5-in. piping H-200-10 Broken disconnect replaced.

T7/30/65 S5-in. piping 'H-200-11 Broken disconnect replaced.

7/30/65 WS—ifi.‘piping H-201-6 ' Broken disconnect replaced.

- POWER OPERATION
12/29/65 Heat Exchanger HX-1 Heater leads to 2 out of Operation continued. Repaired
| 8 elements failed 4/9/68 (see below)
9/15/66 Heat Exchanger HX-1 Heater leads of 2 more Operation continued.

Repaired
4/9/68 (see below).

6L2

 

 
Table 17.3 (continued)

 

 

Date System Heater Type Failure Repairs Made
POWER OPERATION
| (continued) .
10/13/66 Heat Exchanger HX=-1 Heater leads of 2 more Operation continued. Repaired
R elements failed. 4/9/68 (see below).
10/28/66 Heat Exchanger HX-1 All elements out of service Operation continued. Repaired
\ : due to heater leads. 4/9/68 (see below).
9/ /66 S5-in. piping H-102-1S 1 of 3 installed sPares No repairs required. Adequate
fajiled ~ capacity in others. | :
9/13/67 Freeze valve FV-103 Complete failure No repairs required. Adequate
temperature without heater.
12/12/67 Heat Exchanger — HX-2 Open lead — all elements Operation continued. Repaired
\ good | - 4/9/68 (see below). ‘
12/19/67 Heat Exchanger HX-2 Heater leads of 6 out of Operation continued. Repaired
' 8 elements failed 4/9/68 (see below).
4/9/68 Heat Exchanger HX-1 Heaters were removed from reactor cell and were repaired by
HX-2 welding the heater leads to the termlnal strips. All
. heater elements were good.
10/1/68 - 5~in. piping H-100-1 Failure of 1 side element Connected up spares.
10/21/68 Heat Exchanger HX-1 ‘Heater grounded Taken out of service.
8/2/69 S5-in. piping H-102-5 Failure of 1 side element Connected up spares.

 

08¢
 

 

 

 

Table 17.4 Summary of Heater Failures in the Fuel Drain Tank Cell
Date ‘V'Sjstem Heater Type Failure Repairs Made
CRITICAL EXPERIMENT PERIOD
by /65 i  FiiiflLihe-_ H-106-1 Failure 1 side element Replaced element before Run 2,
5/65 - Fill Line = H-th—l Failure 1 side element Replaced all elements — new design
S - Lo during prepovwer shutdown.
5/18/65  Fill Line '”H—th—G . Bad MI cable ‘Replaced MI cdble during prepower .
e - - S shutdown.
5/65 T‘ij‘Freézé'Va;fiés'"""FV4105+1 Top element — loose Repaired 8/6/65 (see below).
y“‘,'“  o u E connectidn . _ _
5/65 =~ Freeze Valves = FV=105-3 Top and 1 side element — Repaired 8/6/65 (see below).
| * - | loose connections o
5/65 Freeze Valves FV-106-1  Top elements — loose Repaired 8/6/65 (see below).
o : connectlon
| PREPOWER SHUTDOWN | / |
8/5/65  Drain Tank FD2-1 Female disconnect Added v 3 in. of ceramic beads’
S : grounded to the 3 leads
8/6/65  TFreeze Valves Replaced all heater elements (new design) in freeze valve heaters FleOh-l,
L - FV-10k-3, FV-105-1, FV-105-3, FV-106-1, FV-106-3, and H-10k-5.
10/10/65 'Fill Line “H-106- Failure 2 elements due to Replaced all elements — new design.
L o ' bad thermocouple
L B o 'POWER OPERATION |
1/21/67  Fill Line H-106-4 Failure 1 side element Continued operation using two good

elements and additional heat from
adjacent heaters.

 

 

T8¢

 
Table 17.5 Summary of Heater Failures in the Coolant Cell

 

Date System "Heater Type Failure Repairs Made

 

PREPOWER SHUTDOWN CHECKOUT

8/12/65 5-in. penetration H-200-14S L4 out of 8 elements Replaced all heater elements (new
failed — loose design) in heaters H-200-1h4,
connection - ‘ H-200-15, H-201-10, H-201-1ll, and

spares. '

B8/12/65 S5-in. penetration H-201-11 1 out of 8 elements failed —
loose connection

8/12/65 S5-in. penetration H-201-115 3 out of 8 elements failed —
loose connection

8/12/65 Radiator CR-2 1 element failed — Replaced element.
loose connection

- +8/12/65 Radiator CR-3 1 element failed — Replaced element.
: loose connection
8/12/65 S5-in. piping H-201-12 Element grounded Replaced element.
8/12/65 S5-in. piping H~202-2 Failure 1 element Replaced element.

POWER OPERATION

 

1/17/66 5-in. piping H-202-2 Broken lead Reconnected.

Checkout after Run T

8/17/66 = Radiator CR-3 Broken extension wire Replaced element.

8/17/66  5-in. piping H-202-2 Failure 1 element Replaced element.

8/17/66 Radiator inlet CR=-T Failure 1 element Disconnected element — adequate
header capacity in others,

8/17/66 Coolant Pump CP-2 Failure 1 element Disconnected element — adequate
_ capacity in others.

8/17/66 Fill line H-206-18S Failure 1 spare element No repairs necessary. Adequate
capacity in others,

C C

 
 

 

Table 17.5 (continued)

 

Dateffl_* ~ System

Repairs after Run 11
3/30/67‘ i; Rediator.

‘Repairs efter Run 13

5/23/69 Radiatori

35/23/69 wijadiator:f_;;i;CR-S 5

1/21/69 ' Redtator

- Heater
'CR-6
‘;CR~3-5

CReT

Type Failure

Failure 1 element

Broken ceramic

l_2 broken 1ead wires..

" Nickel lead wire had

o crystallized

1 element grounded

Repairs Made

- Replaced element.

“Replaced'element

-~ Removed defEctive sections and .
welded 1eads and dyewchecked -

Disconnected element.f Adequate .

capacity in others,

£ge

 

 
 

 

 

284

installation of the elements. The joint was redesigned and new elements,
which incorporate the change,‘were installed when such failures occurfedn

Two fuel-pump heater units resistance increased from 12 to 17 ohms.
These were removed from the reactor cell and the meéhanical Joint between
the heater solid and flexible nickel extension lead was found to be highly
oxidized. All five fuel-pump heater units were removed from.fhe reactor
cell and a welded connection was made at each lead extension. No other
trouble was encountered with these heaters.

Before and during the prepowér shutdown, direct maintenance was per-
formed on the heaters in the reacfor'and drain-tank cell.

17.4.2 During Power Operation .

All heaters performed properly during the heatup.of the fuel and
coolant loops in December 1965 for the start of power operation, with one
minor exception. There was a broken lead on a heater between the radiator
and the coolant pump. It was simply reconnected. After the startup, one
of three groups of elements in one of the héat—exchanger heater units
failed. No action was taken at this time because the other elements were (ii
enough to produce the desired temperature. A summary of the heater failures
and repairs made during power operfition is inclfided in Tables 1T7.3, 1T.k,
and 17.5. | B

During the shutdown in September 1966, a resistance check was made on.
all in-cell heaters at junction boxes located outside the reactor and
drain-tank cell and compared with initial resistances. Other than the
heat-exchanger heater, only one heater was found that had f&v.il’ea.‘7 This was
‘one of three installed spare heaters on the vertical section of pipe under
the heat exchanger. No repairs were required; since there is adequate ca-
pacity in the other heaters on this section of pipe.

The resistance of the coolant-system heaters located outSide of the
reactor cell was also checked. Five heaters were found'that had failed.
One was on the radiator, one on the coolant system 5~in. piping, one on the
céolant-pump furnace, one in the radiator outlet header, and the other one,
s spare heater, was on the fill-line piping. The heaters on the 5-in.
piping and on the radiator were replaced, there was adequate heater ca~

pacity at the other locations. { “j
 

 

 

285

~Aside from heater-element failures, a remote disconnect for one of
the fuel-pump heaters was damaged during remote operation associated with
-the thawing of a saltAplugrin'the fuel-pump offgaé line. The heater was
plugged.ihto & space .disconnect to restore it to service. ;'

By the last of October 1966, all of the heater elements in one of the
heat-exchanger heater units had failed. Satisfactory heat-exchanger tem-
peratures were maintained without this heater, even with the erl.loop emp-
ty. However, continuous_cifeulation of the coolant salt was maintained
until after Run 11, so the full effect of the heater failure could not be
determined. Tests wefe performed with both the fuel and coolant loops
empty after Run 11 to determine the need for this heater in.preheating the

system from a cold-conditien.' When.helium circulation was stopped in the
fuel system, the tempereture distribution was satisfactory fof a.fill with-
out the heater operating. With helium circulastion in the fuel system, one
temperature decreased to_belowr800°F; Since satisfactory temperatures
could be achieved without it, the failed heater was not repaired at this
time. ) : | . | Lo |
_ .While the reactor system was being heated fip_for Run 13,_Heater FV-103
-fias lost by'an open circuit. This 1.5-kW bent-tubular-type heater normally
supplies about 200 W of heat to a b-in. section of the fuel drain line
within the reactor vessel furnace, between tfle freeze valve and the

~ resistance-heated section of the line. The purpose of the heater was to
“help control the temperature profile through the freeze valve. After the
'-fieater failed, tests with flush salt showed that by,propef adjfistment of
the cooling air controls, theffreeze valve could be maintained feliably
'with thaw times in anpaecepteble range. (Thaw time waS-around 12 min when
'the'reactor vessel was at'll80°F'anduthe'center of the freeze valve was at
h95°F ) Therefore, this heater was not replaced. | | _
| In December 1967, a lead to the adjacent failed heat-exchanger heater

| L opened, thereby reducing the output ‘of this heater by 50%. A week later a

second partial failure further reduced.the output to only 33% of normsal.
Although these failures did;notVaffect operetions as long as salt was kept
 circulating, the lack of heet .in the two adjacent heaters made it necessary

to repair them during the shutdown afte/ Run 14. These two units were

 
 

 

286

removed from the reactor cell early in the shutdown (on the fourth and
fifth day after the end of full—power operstion) to see if repalrs were
possible or if replacement units would be required. The trouble was found
in the junction boxes mounted on top of the heaters, where the lead wires
from several of the heater elements had burned in two at their screwed con=-
"nections to the terminal strips. Repair was complicated by the induced ac-
tivity in the assemblies, which produced a gamma radiation field of about

3 R/hr at 1 ft. A temporary work shield was set having concrete block walls
and a top of steel plate and lead block. Direct maintenance on the junc--
tion was done through a small opehing in the'top} The‘terminal strips and
connections were severely oxidized, apparently due to high temperattres
during operation. This situation was corrected by installing nickel term-
inal strips and welding the heater leads directly to them, The  copper wire
from the terminal strip to the disconnect was also replaced with No. 12
nickel wire welded to the terminal strip. Aside from the damégé in the
Jjunction boxes, both heat-exchanger heaters were in good cofidition, and

~ after the repairs, both were reinstalled thrpugh the maintenance shield
without unusual difficulty. ' -

| After the units were reinstalled in the cell, it was found that the
current on one phase of south end unit was zero. This proved to be due to
a fault in the permanently mounted lead wire in the cell, "Tfié fault was
circumvented by installing a Jumper cable between the heater disconnect and
a spare disconnect. The heaters functioned satisfactorily durlng the sub-
sequent heatup of the system.

One of the elements in the heater unit on line 106 adjacent to the
- junction of line 103 with lines 104, 105, and 106 failed during the middle
of January 196T7. ‘This was not replaced since it was possible to maintain
adequate temperaturelusigg the other elements and adjacent heaters.

There have been three failures of radiator headers due to broken nickel
lead wires. The leads look like the nickel had crystalllzed Several bro-
ken ceramic bushings have been- replaced.

The heater located on the south end of the heat exchanger developed a
partial ground in the latter part of October 1968. In order to clear the

ground detector lights on Bus 5, the heater was taken out of service. This
 

 

 

287

was possible because tests had been performed after Run 11 which determined
that this heater was not essential for preheatlng the system from a cold
condition. The cause of the- ground was susPected to be a broken ceramic
bead on the lead wire between a heater element and the junctlonrbox mounted
on top of the wnit. N | B

About a 30% drop in current on one heater on the fuellline between the
reactor and the fuel pump (H-iOO-l) occurreddiate in September 1968. After
a resistance check indicated 8 side heater element had failed, the spare
‘heater elements were placed in service. | _

About a 30% drop in current on the heater on the fuel line between the
~ heat exchanger and the reactor (H-102—5) occurred in the early part of Aug-
ust 1969, After a resistance;check indicated that one of the three heater
-elemente had failed, the spare heater elements were placed in service.

The individual ammeters for each heater circuit were read and recorded
daily. Since operation started, most heater failures caused a sufficient
. change in ammeter reading so as to be noticed., Variation in voltege sup-

' plied to the MSRE was usually less than 2% over a 2h-hour period. A total
2% change caused less than a12°F/hr change of the system temperature under
jsothermal conditions. This was easily adjusted by Changee in heater set-
tings; After the system had been brought to equilibriumiet temperature it
was not ueuallyineceseary~to meke - further adjustments. Typical heater data
including power reduirements for 1200°F isothermal conditions are given in
Table 17.6. | |

The four types of failures experlenced with the ‘heater leads are given

i

~in Teble 1T7.T along w1th_metnods used to make repairs.

 

17.5 ”S§stems Cooldown Rate

:éBefore nuclear power'operetion was started, sjstem cooldown tests were
~run to determine the effects on the system if eiectricai power was lost.

- The cooldown of the fuel nnd eoolant sYetems was done simultaneouely,since
f'these are connected at'the"heat'exchanger. These teets”whichdwere.done
:rw1th the radiator doors closed did not show any dlfflculties. A much more
serlous condition occurred on two occasions after nuclear power operation

was started when the cold rad;ator doors were_scrammed and the coolant pump

 
 

<

Table 17.6 Typical Heater Data

 

1200°F Operating Data

 

 

 

 

Average Calculated
. a ‘ Heat Loss Heat Loss
Insulation ‘Maximum Heat Loss Per ft2 of Per ft2 of
Heater Thickness Installed % of Max Surfacé Surface
Numbers Equipment in. kW kW Installed Watts Watts
Rl, R2, & R3  Reactor Vessel 5 69.6 29.9 43 81 ‘146
FP-1 & FP-2 Fuel Pump 5 ~18.2 12 66 112 82
FFT-1 & FFT-2  FFT L ks 12.8 28.5 59 N7’
FD1-1 & FD1-2  FD-1 4 b7 16.1 34.3 Th.2 . 57
FD2-1 & FD2-2 FD-2 L 4s 1k.9 33.1 - 68.7 57
| - Per £t of Per £t of
Pipe ‘ Pipe
H200-13 30! of 5" piping b 23.75 5.4 22.8 . 150 126
H201-12 23.5' of 5" piping 4 21.6 3.3 15.3 140 126
H202-2 29' of 5" piping Y 23.75 4,7 19.8 162 126
RCH-1 T' of 5" piping Metallic
Multiple- - - — '
: Leyer 13 2.5 19,2 357 . b
RCH-2 5! of 5" piping Metallic
, L ' Multiple- . - »
| Layer 10 2.0 20 400 b
RCH-3 7.5' of 5" piping  Metallic '
| o - Multiple- o ] | C
Layer 15 3. 22,7 453 b

 

~ %Ceramic fiber of expanded silica.

bDesign based on heat loss of 600 watt/ft of pipe on test unit.

C

 
 

 

 

 

289

Type of Heater Lead Failure and Repairs

 

Table 17.7
Heater Failure | Correction Made

(1) Ceramic

(2) Fuel pump tubular
(3) Heat exchanger

(4) Radiator |

Separation of extension lead
from heater element

Excess-oxidation at mechan-
ical Joint between solid
and flexible head

Excess oxidation at mechan-

ical Jjoints at terminal
strip in junection box

Nickel extension lead
crystallized

 

Designed leads with a
cross bar

Made welded joints and
dye-checked

Made welded joints and
dye-checked

Remo&ed defective sec-~
tion, welded and dye-
checked

 

 
e o it s g 17 L it e s b 45 4 £ e Ak e e e

 

290

stopped. Salt was frozen in some of the radlator tubes but was succesfully
melted out with no apparent damage to the ra.dlator° These coolant tests

and radiator difficulties are discussed in the follow1ng four sectlonso

17. 5 1 Fuel and Coolant Systems Cooldown Rate Radiator Door Closed

Three cooldown tests were run. In the first test the cooldown rate

was determlned for the fuel and coolant systems with salt cireculating in

 both’ systems and with all the reactor and coolant cell heaters off except

the heaters on the fill lines. The fuel and coolant system reference tem-
perature dropped from 1160 and 1169°F to 1105 and 1115°F respectively in
one hour and 52 min. This is a cooldcwn rate of 0., S°F/m1nu

In the second test the same heaters were off as in the flrst test and

‘the fuel and coolant pumps. were turned off. After ten minutes, the radi-

~ator heaters and the fuel and coolant pumps were turned back on. This sim-

ulated a power failure followed by the startup of the diesel genersators.
The fuel and coolant system reference temperatures dropped from 1156 and
1185°F to 1089 and 1112°F respectively in 2 hours and 50 minutes. fhis is
a cooldown rate of 0.4°F/min. |

In the third test, all the heaters and the fuel and coolant pumps were
turned off and remained off for the dufation of the test., The reference
temperatures dropped from 1150 and 11T0°F to 1120 and 1141°F respectively
for the fuel and coolant system in 38 minutes. This is a cooldown rate of
0.8°F/min. The reference temperature on the radiator dropped from 1170 to
1090°F in the same period. This is a cooldown rate of 2.1°F/min.

In all three tests the temperatures of the coolant float level indi-.
cators dropped rapidly and their heaters were turned back on. o
17.5.2 Fuel Drain Tank Cooldown- Rate _

With flush salt in FD-2, e test was made of the cooldown rate by turn-

e

ing off all FD-2 heaters. The average temperature at the start of the test
was 1085°F. Ten and one-half hours later, the average temperature was 1020°F,
From this date, it is estimated that it would take at least 56 hours for

FD-2 to cool down from 1200°F to 850°F upon loss of electrical power to the
tank heaters., Decay heating would extend this time.
 

 

 

 

291

17.5.3 Coolant Draln Tank Cooldown- Rate

With coolant salt in CDT, a test was made of the cooldown rate by
turning off all CDT heaters. The average temperature of five thermocouples
dropped from 1160°F to 1085°F in sixteen hours. From this data it is esti-
mated that it would take at least 70 hours for CDT to cool down from 1200°F
-to_850°F.upon loss of electrical power to the tank heaters.

17.6 Discussion

The fuel loop operated above 900°F for 30,8&8-hours and the coolant
loop above 900°F for 27,438 hours. There were 13 heating and cooling cy-
cles of fuel loop and 12 for the coolant loop. The four drain tanks, in-
cluding the drain and £i11 lines, were operated near 1200°F continuously
since the salt was added to the tanksiexcept for two times when FD-2 was
cooled down for installing and rémoving equipment for the 233y addition.
There was not an 1nterruptlon of the power operatlon of the reactor due
to a heater failure.

The main weakness'of the ceramic heaters was the éeparation of the
extension leads from'thé heater elements. To correct this, a total of TO
(30 in the reactor and fuel drain tank cell and. hO in the coolant cell)
ceramic elements of an approved design were installed durlng the prepower
shutdown. Performance since then has been excellent. There were only
5 failures which could be attributed to the ceramic elements ou£_of 221
installed units. | |

 
 

 

292

180,' SAMPLERS S I A
A I Krakov1ak

......

.pllshed by a power-drlven cable which lowered a small capsule 1nto the cir-
culatlng l1qu1d of the pump bowl, The capsule was then ralsed and the salt
vas permltted to solidify before final retrieval and shipment to the Ana—
lytical Laboratory for .anelysis. Lead shielding (~10 inches) was installed
around the major components of the -fuel sampler and a stainless-steel-clad
lead shlpping cask was used to transport the sample° Double containment
was accompllshed by mov1ng the sample (after isolation from the pump bowl)
through a series of conta1nment areas while malntalnlng at least two valves

or membranes between the fuel system or sample and the env1ronment

 'Since the induced act1v1ty in the coolant salt is ~very. short- llved

no sh1e1d1ng was requlred durlng sampllng and only single contaanment was

necessary during semple transport ) '

A total of 81 coolant salt samples were taken and a total of 7h5 fuel
sampling cycles vere completed durlng the five years of MSRE operatlon°
Included in the fuel sampling total were 152 fuel enr1chments of either
uranium or plutonlum, In addit1on to sampllng and enriching, the fuel sam-
~pler was also used to sample the gas. space above the liquid in the pump
bowl and - to make chemical addltlons such as beryllium, niobium, and FeF
to the fuel.

18,1 Fuel Sampler-Enricher .

Since the fuel-sampler usage was more frequent and the eomplexity of
its components were greater than those of the coclant sampler, the major
part of this report describes the operating experience with the sampler;
enricher. A brief description of the coolant sampler and a discussion of |

its operation are included at the end of this report.
18.1.1 Description of the Fuel Sampler-Enricher

The component arrangement of the fuel system sam.pler—enricher50 is
shown in Fig. 18.1. The vacuum pumps and helium purge gas system reduired

for pressure and atmosphere control are not shown., The drive motor for
 

 

 

 

293

GRNL-DWG 63-5848R

 

Fig. 18.1 Compbnefit"Arrangement'for.Sampler-Enricher

 

PRy -
 

294

the cable, the latch onto which capsules were hung were located in primary
containment area 1C which was approximately 17-1/2 feet from the pump bowl.
Critical closures such as the'operational and maintenance valves (motor-
driven gate valves), access port to area 1C (doubly-sealed door), and the
removal valve (ball valve) were buffered with helium to assure that.if any
sealing surfaces leaked, the leakage would be inert gas into the system.
Metered quantities of helium were supplied to these seals and the leakage
rates across the seals were measured by the final equilibrium buffer seal
pressure., - Excess seal leakage into the system or to the containment area
~could be detected by a decrease in buffer seal‘pressure and a pressure in-
crease in the adjacent volumé. To prevent the accidental opening or closure
of compartment doors or valves in improper sequence during sample transfer,
a system of electrical interlocks based on buffer seal pressures and-me-
chanicel signals along with a detailed check list (6A-3) (Ref. 51) was used.
In general, a sampie capsule, similar to that shown in Fig.418.2; wvas load-
ed into a doubly sealed transport-tube which was lowered through the re-
moval valve into area 3A. The_lower part of the transport tube containing
the capsule was unscrewed and left in area 3A whereas the upper part was
withdrawn above the removal valve and the valve closed. The access port
to area lc.was then opened and the capsule was hung on the latch with the
manipulators (the latch and latch pin were so designed that the pin could
not be disengaged during sampling excebt in this area). The access port
was closed and area 1C was first evacuated then pressurized to equal that
in the pump bowl before the maintenance (MV) and operational (OV) valves
were opened. The cable drive motor was actuated until the capsule was im-
mersed in the salt in the guide cage of the pump bowl (1dwer part of Fig.
18.1). The sample was withdrawn 18 in. where the salt was allowed to freeze
in the vertical section of unheated 1-1/2 in. Sched-k40 pipe before final
retrieval into area 1C. As the capsule or sample was moved. from one com-
partment to another there were evacuation and pressurization steps which
removed adsorbed oxygen during capsule insertion and removed gaseous fis-
sion products during sample removal. Capsule transfers inside the sampler
were viewed with a periscope. | |

Enriching was accomplished by lofierihg an enriching capsule (Fig. 18.2)
~into the pump bowl and fillowihg the frozen salt in the capsule to melt and
 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

vl

 

 

 

N
N =)
Ln

 

 

 

 

 
 

 

 

- 296

‘be washed out before retrieval of the empty capsule (the walls of the en-
riching capsules were drilled through to the frozen salt in several places
Just prior to loading the capsule into the sampler).

The sampler-enricher was designed to make 1000 sampling or -enriching
cycles for a period of one year;>? however, the MSRE and sampler were op-

"erasted for 20 runs over a period of 5 years.

18.1.2 Eiperience with the Fuel Sampler-Enricher During Pre-Critical and

Low-Power Operation

i

The sampler-enricher was installed and put in operation in May 1965.
Sampler testing, shakedown, and operator training sessions were conducted
concurrently with.the sampling and'énriching operations of Run 2. During‘
‘this period, 53 fuel samples were fakeh, 87 enrichments made, and 20 oper-
ators trained in the use of the sampler. Some of the problems encounfiered
during this period were:

1. Buffer gas leaks at the operational (OV), maintenance (MV),

and removal (RV) wvalves. '

2. Failure of a solenoid valve on the removal valve actuator.

3. Failure of the removal valve to close completely requiring

manual closure. |

k., Failure of the access port (AP) to close completely.

5. Mbmenfiary failure of cable drive motor on capsule retrieval.

6. Rupture of the flexible containment membrane (boot) at the

menipulator. '
- T. Deformation of the manipulator arm and fingers.

8. Accidental drop of an empty capsule onto the operétional valve.
These and additional problems encountered with the subsequent 86 sampling
cycles made during the approach to power runs are discussed in more detail
in the following paragraphs. , _

Buffer Gas Leaks at the Operationél and Maintenance Valves —Fthring
the sampling operations of Run 2, leaks developed through the upper metal-
to-metal seats of both the operationél and maintenahce valves. The oper-
ational valve was subsequently removed; examination showed that a thin
black ring, which was easily removed, had formed at the upper sealing sur-

face of the valve gate and a small quantity of salt spheres (<1 gm) had
 

 

 

 

297

collected between the seats of the gate. Apparently salt particles were
dislodged from the capsules during the operation where the capsules (after
~ having been dipped in the pump bowl) were disengaged from the latch above
the valve. When the stem and~gste were lubricated, the valve sealed almost
completely; however on removal of the lubricant, the buffer gas leak rate
increased to 2 ce/min through the upper gate seal and remained at zero
through the lower seal.

A few sampling cycles after the valve was reinstalled, the leak rate
again increased to 20 cc/min through the upper seal. Apparently more salt
particles had lodged between the buffer-gas sesling surfaces. Repeated ef-
forts to blow the particles from the sesaling surfaces failed. Since there
were three other sealing surfsces between the pump bowl and the sample ac-
cess area and since the leak did not increase with continued sampling, the
valve was not replaced. | | |

After approx1mately 86 additional sampling cycles, an opportunity to
clean the operational valve arose when the reactor was drained to repair
the cable drive motor in April 1966 (to be discussed later). Cleaning the
seating surfaces decreased the leak rate for a while until an empty cap-
sule was accidentally dropped on the gate. Subsequent valve operation
again resulted in a high buffer-gas leak rate. The buffer gas seal at this
surface continued to deteriorate slowly throughout the remainder of MSRE
operations; the only consequence was a slow pressure buildup in the con-
tainment volume sbove the valve which required periodic venting. From
February 1967 until final shutdown, the buffer supply pressure to this seal
was used only when the sampler was in operatlon thereby dbv1at1ng the perl—
odic venting of area 1C. o ' o

During the latter part of 1966 there had been a gradual increase in
leakage of buffer gas through the upper seat of the maintenance valve also.
- Although the procedure specifled-that the maintenance valve be opened be-
fore.the operational-valve'is'ofiéned, and that the operational valve be
“closed before the maintenance valve is closed, on at least one occasion the
‘reverse sequence was followed during closure and thus foreign matter could
have been dislodged from the upper velve onto the lower valve, While the

lesk rate remained relatively small (v25 ce/min), it was sufficient to

 
 

 

 

298

cause difficulty with proper operation of the interlocks (an electrical
signal which indicated that the valve was indeed closed was based on the
ability of seating surfaces to maintain a specified buffer pressure). Since
'~ cleanup would have been difficult and replacement of the valve was not wér—
'ranted by the lesk rate alone, a mechanical.method of assuring that the
valve was closed was sdbstituted.for the pneumatic methoda No additionsl
problems with these valves were reported.

Area 1C Access Port — The access port door was operated by six clamps
vhich relied on pneumatic operators and a spring for opening and a sched-
uled time-delay of these pneumatic operators for closure;lhowéver,'once
closed, the mechanical linkage was such that the clamps no longer require
~gas pressure to keep the dobr tightly closed. The ability of twolconcen—
tric neoprene gaskets (between the access port door and housing) to contain
helium at 55 psia satisfied an interlock and indicated that a gas-tight
seal did indeed exist between areas 1C and 3A. On several occasions during
the pre-critical run, the access port failed to close properly.  Also, one
or two of the six access port operators ffiilgd to open and the manipulator.
wvas used to release the sticking operators. During the shutdown following
the low-power experiment, the clamps were readjusted and,an extension was
added to the pin of each clamp; this extension made it easier to open any
clamp manually with the manipulator in the evént of a pneumatic malfunction.

Flexible Containment Membranes — There are two 0.020-in. thick flexi-
ble urethane membranes (boots) which partially enclosed end provided mo-
bility to the manipulator arm used in transferring capéules to and from the
latch in area 1C and the bottom portion of the transport carrier in aresa
3A. A vacuum was maintained between these boots to .monitor the integrity
of the boots and- assure that double containment did indeed exist between
the sampler operator and the sample. A ioss of vacuum and a pressure re-
duction in area 3A would indicate a leek in the outer boot whereas a loss
of vacuum with no pressure change in area 3A would implicate the inner boot
(atmosphere side) or possibly both boots.

There were three boot failures durlng the first 1Lk0 sampllng cycles
of Run 2. One was caused by an inadvertent evacuation of area 3A while
the manipulator cover was off; the resultant pressure gradient ruptured

the boot. On another occasion, the boot was snagged on the bottom piece
 

299

of the transport tube in area 3A when the manipulator was used to release
-a stuck access port clamp. The third failure reSulted from pinching the -
boot between the manipulator arm and the housing. '

At the end of the next run (Run 3), a pressure switch (with an alarm
and interlock) was installed'to detect and prevent a negative pressure
gradient greater than 1/2 psi between area 3A and the manipulator arm.
Steel rings were installed in the convolutions of the inner boot to hold
it free of the manipulator arm and p0551bly prevent the pinching-type
failure. _

- Also to ensure thaf at-least two barriers existed.between.primary con-
tainment and sampling persoanel,_a pressure switch was installed on the
‘manipulator cover which required & negative pressure of k-in. Hg in the
cover before either the operational or maintenance valve could be opened.
| In Aprll 1966, the manipulator cover pressure was inadvertently evacu-
ated to v25-in. Hg w1thout 1ower1ng area 3A pressure simultaneously; the
12-psi differential caused a small puncture of the inner boot. This set
of boots had been used for 48 sampling cycles.eince,power operation was
- begun and the maximum power_reached at that time was 2.5 MW. The radiation
level of the manipulator assembly and boots was 10 R/hr at 3 inches from
the fingers, but was reduced by a factor of 10 by scrubbing with soap and
water. About 2 hours were'required for replacement of both boots.

Deformation of the Manipulator Arm and Fingers fi—-Durlng Run 2, the
manipulator arm and. fingers were bent which causedrdlfflculty in gripping
~ the latch cable and moving the manipulator arm. At the end of Run 3, the
arm was replaced and & l/h X l/h inch projectlon vas welded to the bottom
"of each finger to aid in grasplng the capsule ceble from the floor of area
3A§ The manlpulator arm was replaced and the clearances between the arm

and castle 301nt were increased to reduce the force requlred to operate

~ the manlpulator.

The manipulator and fingers operated satlsfactorlky for 421 addltlonal

- sampling cycles until August 1968 at which time the manipulator assembly

- was replaced because the tips of the fingers no longer closed tightly. The
old assembly was decontamlnated and repaired. The replacement fingers op-

erated satisfactorily untll September 1969 when the double metal bellows

 
 

 

 

300

 deve1oped a leak, The sampler on the fuel processing System'was canniba-
lized to make the repairs; this unit operated satisfactorily for the sub-
sequent three months of reactor opersation before shutdown.

~ Capsule Recovery from the Qperational Valve — On one occasion during
the precritical‘run; an empty sampling capsule was accidentally knocked
into area 1C before the latch pin was completely engaged in the latch, and
the capsule dropped onto the gate of the operational valve. The capsule
was retrieved by rémoving the manipulator assembly from 3A, opening the
access port, and snaring the capsule cable with a wire hook. Since this
type of accident could reoccur, the brass latch pins used to attach the
capsule to the latch were replaced with nickel-plated mild steel pins so
that capsule recovery could then be effected with a magnet.

After approximately a month of full-power operation in July 1966, an
empty capsule again was dropped onto the gate of the operational valve when
thé manipulator slipped during the latching operation. A magnet'was low-
ered into fihe transfer line and the capsule assembly was recovered as the
magnet was withdrawn. The radiation level of the magnet after the recovery
operation was 10 R/hr at 2 ft.

All sampling capsules of the bucket-type assembled since the latter
part of 1966 contained a nickel-plated mild steel top so that capsule re-
covery could be made with a magnet.

Removal Area, Valve, and Seals — During the pre—critical run, the
removal seal and valve required realignment and increased tolerances be-
fore the transport container and removal tool,assembly,would.slide through
these units freely without binding. In addition; the,rémoval valve failed
to seal properly even though  the ball and seals were replaced, Therefore,
at the end of the next run, the.valve assembly was replaced with a modified
version to improve the sealing characteristics and improve future access
to the valve. This unit served satisfactorily except that a buffer gas
leak developed in the top teflon seal of the valve dufing the latter part
of 1967° As a consequence of the leak, the equilibrium buffer seal pres-
sure feached only 30—35 psia instead of the normal 50-55 psiaj; the associ-
- ated pressure switch which pefmits the opening of the operational and.main—
‘tenance valves and the access port was lowered from-SO to 30 and finally
in December 1967 to 25 psia where it remained for the remainder of reactor

operations.
 

301

Latch, Capsule, and Csble Malfunctions — Early in Run 2, the cable

- drive motor stalled during the withdrawal of an empty capsule. The cap-

sule was reinserted sbout .12 ih. and then withdrawn successfully. The

reason fbr_thermalfunctioo;fias unknown., Approximately 100 sampling cycles
later in December 1965, the @otor stalled again while withdrawing a 10-gm
sample. After repeafied attempts, the capsule was withdrawn completely and

" the cable was examined., The capsule~was5empty_and»the'cable_was found to

be backed up into the drive'finit box and caught in the motor gears. It
was assumed. that the latch had hung on the gate of the operatlonal or main-
tenance valve causing the. cable to coil up inside area 1C, The cable was
straightened, the limit switches on the operatlonal ‘'and maintenance valves
‘were'reset_to_open;the velves wider and the diameter of.the latch was re-
duced. .

" In April 1966, after approximately a week at a power level of 5 MW,

- the capsule drive-unit motor stalled as the latch was being retrieved

through the pipe bend near the pump bowl. After several inserts end with-
drawals, the latch was retrieved. The latch and part of the cable were

T_then-pulled through the access port, the removal valve, and into the sample

transport cask which provided shielding while the latch was replaced with |
one designed torprovide-anfadditional‘l/8—inch'diametrical clearance,
Approximately 3 months laterithe capsule stuck again temporarily for

some unexplained reason on withdrawal and no sample was obtained. Subse-

'quent hangups - with ‘more serious consequences are discussed in Section 18.1.3.

Repair of an.Open Electrlcal Circuit — While testing the new latch,

the cable position, 1nd1cator stopped ‘when the latch was approxlmately two

feet from the pump bowl* the;upper'limit svitch also actuated at this time,

',Electrical continuity'checks showed open. circuits to the‘insertand with-

'n;,drawvwindiags'of the cable:drive'motor and also to the upper limit switch.

Allgthese-leads-pehetrate:fhe7con£ainment wall through a'cofimon 8-pin re-

'”ceptacle. ‘Since repalr could not be made while fuel was in the reactor

_Jl;without violating contalnment the ‘reactor was drained. The cable and
latch were then pulled into Area 3A using the manlpulator to pull and the

: transport container and access port 0perators to hold the cable’ so that

the cable could be regrasped w1th the manlpulators for another pull The

 
 

 

 

302

~ operational and maintenance valves were then closed. The assembly includ-
ing thé drive motor, cable, and latch were removed, partially decontami-
nated, and repaired with the use of shadow shielding techniques. (Three
connector pins in one of the 8-pin receptacles were burned off.) All six
similar receptacles at this location were filled with epoxy resin to ‘pro-
‘vide additional insuletion and strength., The cable which was bent in sev-
eral places during retrieval was decontaminated with soap and water to_re—
duce the radiation level to 5 R/hr at 3 'in. before it was straightened.
After reassembly and checkout, normal sampling was resumed. '
' Contamination of the Removal Seal — An area at the top of the Sampier
- was provided so that the lower part of thertransport carrier and capsule
could be evacuated and pressuriied several times to remove oxygen and mois-
ture from the capsfile and carrier before insertion into the dry box (Area
3A) of the sampler. The bottom seal for the removal areaiwaS‘provided by
the removal valve and a seal at the top was made when the transport tube
was inserted through a set of O-rings at the top of the.rembval-volume;
During sampling the shipping cask was aligned and placed over the removal
area, A small amount of vacuum grease was smeared onto the lower part of
the transport tube before it was lowered through the shipping cask and into
the O-rings of the removal seai° When the transport tube was further lo-
wered into Area 3A (a contaminated area) to insert an empty capsule and
again to pick up the sample, the ocuter surfaces of the transport tube and
removal tool became contaminated with solid fission products from the cap-
sule, floor of Area 3A, and/or manipulator. Some of these particles were
wiped off by the O-rings of the removal seal as the removal tool and trans-
port tube were withdrawn from Area 3A; that remaining on the removal tool
was wiped off with a damp cloth. The shipping cask was gradually contami-
nated as both the removal tool and transport tube were drawn ifito it, By
this process the removal seal becomes progressively more contaminated with
each semple. During pre-critical operations this transfer mechanism for
particulate matter was not obvious; however, after power operations had
begun and especially after the latch maintenance work, where the latch was
manually retrieved from +the vicinity of the pump and out thrbugh the re-
- moval aree for repair, the removel area and the fop of the sampler became

contaminated,

>
 

303

* The adoption of the'folldwing procedure'proved effective in preventing
~the spread of contaminatlon durlng sample removal and transport throughout
~ the remainder of reactor operatlons"

" (1) The top of the sampler was establlshed as a Contamination Zone.

(2) The removal tool was wiped with a damp cloth durlng withdrawal
from the removal area. 7 i

- (3) The shipping cask was wiped with a damp cloth prior to enclosure
in a plastlc bag for shlpment to the analytlcal laboratory.

(%) The top of the sampler was wiped with & demp cloth after each
sample. | _ - _

(5) An exhaust hood was erected near the removal area and.partially
enclosed the shipping cask'to'maintain & slightly reduced pressure in this
~area so that any airborne particles would be drawn into the fllter and ex-
haust system of the bullding.

(6) The removal seal" area was cleaned periodically with damp wipes.

- Control Circuitry Changes ——-Durlng this perlod five changes were made
' to the sampler control clrcults '
(l) A pressure switch was sdded to prevent evacuation of Area 3A to
‘more than 10 in. of water greater than the manipulator cover aresa.
(2) An interlock was*added'to'require that both the operational and
_'maintenance valves be closed before the access port can be opened.

(3) A permissive sw1tch and light were installed to indicate that
the access port can be opened only when Area 1C pressure is equal to or
"less than Area 3A pressure.a_,_~,f-,',, , | |
| (4) A fuse was added to the capsule drive motor circuit to protect
'the motor and the electr1ca1 receptacles from excess1ve currents. '

(5) Voltage suppressors were placed across the two motor W1nd1ngs
rto limit any high voltage peaks durlng startlng and stopplng.,- '

o Mlscellaneous Sampler—Enrlcher Prdblems_-—-Durlng the 1ow—power experl-

',dment a five-inch—long capsule W1th an opening nesar the top and capable of

~a1hold1ng SO-g of salt was lowered 1nto ‘the pump bowl but falled to trap a

 

_“sample because the assembly was not long enough to permlt total 1mmer51on.
-Subsequent longer assemblies (9) obtained salt except one which was believed
"to have hung on the latch stop ‘at the pump bowl and did not enter the pump.
The capsules were subsequently modified to allow them to hang straight.

 
 

 

 

 

304

*Most of the samples taken after the reactor exceeded 1 MW (nuclear
power) were highly radioacfive and exceeded 1000 R/hr at 3 in. Fission
products adhering to the outside of the capsfiles contaminated the,Bottom
cups -of the transport containers. For a limited time disposable plastic
liners inserted in the cups were effective in reducing the radiation level
in the bottom portion of the transport tubes. However, as more fission
'produets built in the fuel and the interior of the sampler became more
contaminated, it was more economical to fabricate disposeble,bottom,cups
of mild steel rather than decontaminate and reuse the stainless steel'eups°
The new cups were shorter than the previous ones and the plastic liner was
used to confine the capsule, wire, and latch pin so that the top part of
the trahsport carrier could be more easily slipped-over this combination
to engage the threads and double O-ring seals at the bottom of the mild
-steel cups. | ,

| On one occasion while the top of the transport containeriwas being
mated with the bottom which contained a 50-g capsule, the wire on the cap-.
sule caught in the threads of the matingfpieces and galled before the two
pieces were sealed completely. Thereafter the liners were lengthened to
accommodate the the 50-g capsules.,

In July 1966, the fuel pump was accidentally overfilled with flush
salt and salt was pushed up into the sampler tube approximately two feet
above the fuel pump where it froze and thus prevented the capsule and latch
'from being lowered into the pump bowl. The salt was melted and allowed to
drain back into the pump by energizing a set of remotely-placed heaters
around the sampler tube ‘and the normal heaters surroundlng the pump bowl.
Normal sampling was subsequently resumed ,

Contamination of the area surroundlng the sampler was mlnlleEd during
malntenance by designating the area surrounding the sampler a Contamina-
tion Control Zone and at times. by erecting a plastlc tent around the sam-
pler. Repair work was done inside the zone with approprlate protectlve
ciothlng, gloves, and shoe scuffs. Contamination was found outside the -
zone twice and on_both occasions it was attributed to contaminated shoes

which indicates that this type of contamination is not readily airborne.
 

305

18.1.3 Subsequent Operating Experience with.the,Sampler—Enricher-

| | The major sampler problems ‘encountered durlng the remainder of opera-
tions were. caused by the latch and/or capsule lodging at the entrance to
the sampler tube or at one of the isolation valves 1mmed1ately below Area
1C with subsequent unreellng and tangling of. the cable 1n Area lC and in
‘bhe gears of the drive wnit. ,

Ruptures or pln-hole leaks in the boots enc1051ng ‘the manipulator arm
were a chronic source of troublefrequiring perlodic replacement., Life of
the boots varied videly but avenaged L) sample cycles before failure. The
cause of most of theserleaks:was undetermined because the boots could not
be examined easily due to therhigh radiation level of the manipulator as-
 gembly, | . - | |

Gradual deterioration of the buffer gas seals at the operational,
malntenance and . removal valves and at the access port continued and on
occasions the time required to take a semple was lengthened because &
longer time interval was required for the buffer pressure to build up suf-
'ficiently to actuate the proper.relay for the next step inithe procedure.,

Loss of Capsules in‘fhe“Pump Bowl = In. ‘August 1967 during what ap-
peared to be 8 normal sampling operation the capsule and/or latch had ap-
parently lodged ‘at the maintenance valve while being lowered causing the
icable to coil up in Area 1C and into the motor drive area where it had be-
'come,snarled and kinked. The first indication of a melfunction was during
the capsule retrieval cycle when the motor stalled_with-B‘ft'3 in. of cable
off the reel.53 Repeated inserts and withdrawals recovered only 5 more in.

of cable before the motor stalled with only &n inch of travel in either di-

";)rection (thls had happened once-hefbre in 1965 but ‘the cable and capsule

~ ‘were-retrieved completely at that time) | |
. If the capsule and cable were lodged at the sampler tube entrance be-
:“'low Area 1C, then. the operatlonal and maintenance valves could be closed
;and sampler repairs could proceed without a reactor drain, otherw1se a
drain would be necessary. After con31deration of several poss1b111t1es, it
was decided to disengage the motor drive from the operational valve and
close the valve slowly with the hand wheel W1th the idea that if the cap-
sule were below the valve, the cable would offer enough resistance to clo-

'surerto be detected in which case the valve would be reopened and the

 
 

306

reactor drained before repair. No resistance was detected until the valve
was practically closed. Therefore,'bcth valves were closed and when Area
1C was opened, as expected, the cable was coiled up within;_hekever the
latch, capsule, and sbout 6 in. of cable were;missing.“The reecfior was
then drained and grappllng tools were fabricated of flexible tublng and
cebles which could be inserted through the sampler into the pump bowl.,
Four of the five tools which were most successful at grasping a dummy latch
are shown in Fig. 18.3. _v | ; -

When'the.noose tool was'lowered, it&snared-the latch but the 1/32-in.
steel cable broke because the latch was apparently glued in place with salt
at the latch sfopu‘ The_dislodgihg tool was then fised but apparently to no

~avail because & second noose grasped the laich but could not budge it with

a pull of 25 1b. The pump was then heated until the latch was at a tem-

'perature of TOO°F. It was then 1ifted several feet without difficulty be-

fore it lodged in the tube, apparently at the expansion joint between the
fuel fiump and the sampler. A corkscrew tool was next inserted but the
latch again hung on retrieval this time at the first bend near the fuel
pump. The latch retrieval tool which was de51gned to fit over the severed
cable and stem and thus prevent the ceble from jamming against the wall

‘during retrieval was successful in pulling up'the‘csble;’letCh,'snd latch

pin. The capsule apparentlylsnapped_its-cable and dropped to the bcttom
of the sampler cage when the latch ceme to & sudden stop at the latch stop.
The dotted outline of & capsule in Fig. 18.k4 shows how & capsule, once de-

‘tached,from‘its c&ble, could slip out of the sampler cage and then be

trapped by the mist shield in the pump bowl.

- Bince the capsule top is made of niCkel plated mild steei, a magnet
in combination with & go-gage (Fig. 18.3) was lowered into the pumphbcwi
to verify that the sampling tube was unobstructed end to pick"fip the cap-
sflle; Capsule recovery was unsuccessful and'since no adverse chemicsal or
mechanical effects were envisioned should the cepsule be allowed to remain-
in the mist shield, further retrieralrattempts were Ebendoned

While retrleval efforts were in progress, & replacement 1-C assembly
vhich was on hand was equipped with e latch made of 430 stainless steel
and with a sleeve to prevent future tangling of the cable in the gears of

the drive unit. However, subsequeht examination of the 0ld drive unit in
 

 

 

307

ORNL-DWG 67-1788

Y/g-in CABLE -~ | i ~'/4=in. CABLE

i/32-in. CABLE
tYa-in. DIAM

Vo ~in. FLEX
CABLE SHEATH

 

 

 

 

 

 

CABLE BEING NOOSE READY

PUSHED OUT FOR FISHING
NOOSE TOOL ’ DISLODGING TOOL

  
 
   
  

_ ine ~in. CABLE
/ 'fls ~in. CABLE

  

Uz

 
  
   

 

   
 

 

 

4
’ tin. 1D
| /  COPPER
’ ' a
|1/, in. SCHED 40 PIPE %
[4
| 3,in. DIAM x 3%2in. [
. MAGNET—- ;
/ :
M ) ;
5 TE ~1.375 in. DIAM

- BRASS

T

" LATCH RETRIEVAL . ... . GO- GAGE AND

TOOL CAPSULE RETRIEVAL
= o TOOL

o Fig. 18 3 _Tools Developed for Retrieval of Latch and Capsule from
s Sampler—Enricher

 

 
 

 

 

 

308

ORNL-DWG 67-10766

LATCH ASSEMBLY CABLE SHEARED OFF

 

LATCH
STOP.

 

FINAL CLOSURE
WELD

 

T A e

 

T L L TN T T TLs LT oYy

 

 

VA A A A A S A A A LS N R B, T TR T T T T R T g

 

 

een p
il
|

 

T Fodndd T

 

 

 

NORMAL OPERATING
SALT LEVEL

   

CAGE FOR
SAMPLE CAPSULE

Lt T T TN s SsS ST TSNNSO,

 

—

A

ACTEITTCYS

 

 

 

SALT INTAKE SLOT

Fig. 18.4 Location of Sampler Latch and Capsules in Fuel Pump Bowl

 

 

 

 

 

 

 

 

 

 

APPROXIMATELY AT THIS LOCATION

LATCH ASSEMBLY NORMAL POSITION

TOP OF FUEL PUMP

=
P

,v.‘.':

—

BAFFLE

POSSIBLE MOVEMENT

OF CAPSULE SAMPLE CAPSULE

   

SAMPLE CAPSULE
CABLE

e

; SECTION A-A

SAMPLE CAPSULE

LOST SAMPLE CAPSULE

3/ in. MAX
—‘J: 1 0 ' 2 3 4
,(;&‘ INCHES

 
 

 

 

309

a hot cell revealed that the cable waslnot tangled in the gears as had
been assumed; the failure to withdraw or.insert was due to a sharp kink
in the!cable-whicfi beceme lodgedrin the 17/32-in. diameter channel of the
- latch positioner. | .

When normal operations and sampling were,resumed, s commercially avail-
sble device,was.testedrfihich was capable of sensing the latch (now made of
magnetic material) inside the sampler tube. A design was prepared to mount
this proximity switch about 10 in. below the maintenance valve to detect
the latch as it is lowered during normal sampling; however, before it could
be installed, the latch agaln lodged in the sampler tube without detection
in March 1968, The usual 17 ft 5 in. of cable was reeled off and was tan-
gled in Area 1C. As it was rewound, the motor stalled with 13 ft 5 in. off
the reel. When repeated withdrawal attempts failed, the reactor was drained
and the capsule access portrwas_opened._ The capsule was visible through the
rlcwerecornerwof the opening and the latch could be seen in the back of the
chamber w1th many loops and coils, some of which appeared to extend down
1nto,the sampler tube. Rather than risk cuttlng the cable the isolation
valves were left open; however, when an attempt was made to 1lift the cap-

- sule out.of the chamber, the manipulator brushed some of the coils causing
them to spring out through the port and pull the capsule wire from the ma-
nipulatorsfingers. The capsule dropped into the sampler tube and the latch
tipped over, thus releasing the key; the capsule and key_thefi disappeared
down the sampler tube in what appeered»to be a bottom-up position.

After. several unsuccessfulrattempts were made to retrieve the capsule
by lowering magnets of varlous s1zes into the pump bowl, the reactor was
filled with flush salt. A 50—g capsule (5 in. long with an opening 4-1/8
in.efrom the bottom of thejeapsule)_wes lowered 1nto_the.pump.butmcame up
 empty; salt dropletsrclinging”te.the outside indicatedrthet'the capsule had
f'beefi.enly half sfibmefged Later a 10-g cepsule did collect a sample.

In the meantime a full-scale plastlc and metal mock—up of the sampler

'tube mist shield, and sampler cage was constructed and used to select the

. best tools and magnets which could be lowered via the sampler in another

attempt to retrieve the capsule,sh The simplest and most effective tool
proved to be 1/2 and 3/4-in. diameter Alnico-5 magnets. When the flush

salt was drained and retrieval efforts resumed, sounds. from a contact

 
 

310

microbhone on the fuel pump indicated that the 1/2-in. dia. magnet lifted
an object & few inches and then dropped it. After many sttempts, the mag-
net came up with an obJect'which 1ater proved to be the corroded top of the
old capsule. Further efforté-wiéh_either magnet were unsuccessful and the
second capsule was abandonéd. | ‘ '

During this shutdown, the proximity switch was installed sbout 4 in.
below the maintenance valve and the-sampling frocedure was modified to
stipulate that if the switch is not actuated when the capsule position indi-
cator shows that sufficient cable has beqn-réeled off to reach the proxé
imity switch, the drive will be stopped. ‘' Also a variac was added to the
drive motor circuit SO'that'the'operatihg voltage‘éould be éhanged’to 80
~ volts to lessen the possibility of seriously kinking the_éable; -
| - During the next stertup, with flush salt in the reactor, salt was
" trapped in a 10-g capsule but none trapped in a 50-g capsule which requires
an immersion depth of 4-1/8 in. to trap & sample,-'However;'é 30-g capéfile
requiring only 2-1/2 in., of immersion did trap a sample. These results
agree with the measurements taken during capsule retrieval which indicated
that the abandoned capsule was 1-3/4 in. above the bottom of the cage. -
However, with fuel in the resctor, the actual level in the pump bowl was
somewhat~highér than that indicated because of the high void fraction of
'the liquid in the pump bowl.r Consequently no problems wereé encountered
taking 50-g fuel samples throughout the remainder of operations eiéept'
during the latter part of 1969 when the fuel pump speed was lowered to
980 rpm. At this speed the fuel salt void fraction is approximately equiv-
alent'to_that*fdr flush salt at normal pump speed. The failure to trap &
50-g sample under these conditions indiéated that the positiofi-of the aban-
doned capsules had not changed after approximately one year of operation.

In October 1969, & serious ceble tangling problem was averted by the
use of the proximity switch. When the switch failed to actuate after 3 ft
10 in° 6f cable was peid out during an otherwise normsl capsule insert, the
cable was retrieved until the position indicator indicated that all of the
cable was back on the takeup reel and the capsule was fully withdrawn into
isolation chamber 1C. The operational and maintenance valves were then
closed and the access port was opened. The cable was indeed fully re-

trieved ag indicated and the capsule was fufily visible but was lodged
 

 

311

diagonally between the-ledgeé.of the doorway to the chamber. Apparently
the capsule had lodged at the sampler tube entrance or at one of the isola-
tion valves and as the ceble was paid out,‘it somehow encircled the capsule
because the capsule was lifted and relodged at a higher position (during
cable retrieval) than it was at the start of the sampling'procédure. The
capsule was then lowered to its normalAposition_and all subsequent cable
operations proceeded without interruptions.

Wiring Fault,——-In.November of 1967 as a sample was being withdrawn,
the 0,3-amp fuse in the dri#e'motor failed. Subsequent tests showed that
the insert mode was normel-at 0.2 amp but during withdrawal, the current
‘increased to 1.0 or 1.5 amp, indicating leakage.to ground. During further
tests, three_circuits opened, disabling the insert and withdraw circuits
and the upper limit switch. After the erection of & tent and other meas-
ures to prevent the spread of-contamination; a 3-in, diam hole was sawed
in the cover plate of Area 3A directly above the cable penetration into
‘the inner box (Area 1C). The failure was found where the wires were bent
back against the side of the plug on the lower end of the cable between
the inner (1C) and outer (3A) containment boxes. When a temporary con- -
nectionrindicated that everything within Area 1C was operable, the sample

was retrieved and the isolation valves were closed. The damaged section
| was abandoned and a new cable was installed having a penetration through
8 b-in. pipe cap welded over the sawed hole in the top cover.

mWhile the pi?e'cap was being welded in place, a heavy current evi-

dently went to ground through an adjacent penetration, destroying the plug
. end receptacle. Repairs were made'by cutting out and replac1ng this pene-

tration. As & result of thisgexperlenCe, an_isolatlon transformer and fuse
'~was'added.to each of thétthrggtdrivefunit-cables which*allowsrdne ground
' without interference with obéfétion;[ After the repairs, all eircuits were
roperable except the upper limit switch which stops the drive motor when the
_latch reaches the latch stoP._ Since the motor and circuit can easily with-
stand blocked—rotor current “the upper limit switch was bypassed and there-
7 after the motor was turned off when the p031tlon indicator indicated that
the 1atch was fully w1thdrawn,u

Repalr of Cable Drive Gear — In December 1968 during an attempt to
remove & capsule from the latch, it was found that the cable could be

 
 

312

pulled off the reel vhile the drive motor was stationary. The tentative
diagnoéis was that one of the'pair,of drive gears was slipping on its

shaft. In order to galn access to the drive unit, it was necessary to re-
move the shield blocks over the reactor cell which, in turn, required that
the fuel be dréined. After a temporary containment enclosure and contami-
nation zone had been set up around the sampler, the conteinment box, 1C,
containing the drive unit was disconnected, lifted, and a 3-in. hole sawed
through the side adjacent to the gears. One gear was found to be loose be-
cause its two setscrews had come loose. The gear was repositioned and tack-
welded on its shaft and the other gear was fastened withrjam'screws on;tdp
of its setscrews. A patch was then welded on the box and the unit rein-
stalled. The work was done without excessive exposure of personnei despite
the high radiation levels (10 R/hr at 12 in. from the box) by use of shield-
ing and extended tools designed especially for the Job.

' Since other parts of the sampler were easily accessible during this
shutdown, the manipulator hand was repaired, the illuminator port and view-
ing window lens were replaced because of discoloration by radiation, and
an imperfect seal in the removal valve was replaced. -

Vacuum Pump Problems ——-The_sampier was equipped with two vacuum pumps.
Vacuum pump No. 1 was used to evacuate contaminated areas such as contain-
ment areas 1C and 3A and discharged into the auxiliary charcoal bed; vacuum
punp No. 2 was used to evacuate non-contaminated areas such as the manipu-
lator cover and the plenum between the manipulator boots and discharged in-
to the containment air system filters and to the stack.

During the July 1966 shutdown, the oil levels in the pumps were
checked and a small amount of oil waéfladded, _

In January 1969 there was an activity release to the stack (<0.08 mc)
which was attributed to a gas lesk around the shaft of vacuum pump No. 1.
The pump was replaced because the oil seal at the shaft could not be re-
.paired without extensive decontamination procedures. Four months later,
repeated vacuum pump motor outsages (due to overload) were apparently caused
-by low oil level. On one inspection, the pump weas practically empty of oil
with no evidence of external oil leasks and was consequently refilled to the

proper level. On the next inspection, the pump contained approximately .
313

2 qts more than the normalhinventory. Apparently the previously lost oil
drained back into the pump. High pressure (>13 psia) in Areas 1C was nor-
mally vented into the.auxiliary charcoal bed through a line bypassing the
pump. Apparently high pressure was vented through the vacuum pump instead,
thus carrying over a large portion of the oil into the holdup tank and off-
gas line above the pufip discharge. After the 0il level was readjusted,
'pump operation returned to ndrmal>and the last oil check was normal.

Heated Shipping Cagk —- At temperatures less than LOO°F, radiation
from fission products can produce free fluorine in frozen fuel salt, To
minimize fluorlne production and thus avoid the effects of fluorine on the
results of oxide and trivalent uranium analyses, a shipping cask was fabri-
cated which‘maintained-the.fuel sample at approximately S00°F from the time
-itrwas removed from the sampier until it was unloaded at the analytical
’iaboratory. The cask used molten babbitt both as shielding and as a heat
reservoir. Built-in electric heaters melted the babbitt before the sample
waS'loaded and the heat of fusion maintained the samplerat a relatively
constant temperature for about 9 hours during shipment and storage. For
the elevated temperature, O-rings made of Viton A were substituted for the
standard neoprene O-rings in the transport tube. In addition to the double
O-ring seal provided by the transport tube, the transport tube was capped
inside the cask during shipment.

All samples were unavoidably cooled to room temperature in the sampler
where they were sealed in the transport tube at atmospheric pressure. When
the transport tube was inserted into the heated cask, the pressure in the
- tube 1ncreased to approx1mately 10 psig because of the temperature inerease.
:‘On one occa51on when the cap was removed from the cask airborne activity
was released from the cask caV1ty into the ventilation system of the un-
1oad1ng statlon at the analytlcal laboratory. Subsequently the sealing
surfaces of the top part of the transport tube were found to have been
eroded by repeated decontamination procedures to such an. extent that the
double O-ring seals were ineffective. After examlnatlon of the sealing
surfaces of the remaining five transport tube tops and reevaluation of the
decontamination costs, it was;decided to discard the reusable_stainless

steel transport tubes and use "throw-away" transport tubes made of mild

 
 

 

314

steel for all subsequent samples. A valve was later added to the'cap of
the heated cask so that any future pressure buildup in the cask cavity
could be released to a charcoal filter before cap removal. -

 Other Capsules and Samples — The size of the capsules,fhat could be
accommodated by the samplér'wds 1imited to 0.75 inches in diameter and
6-1/4 inches in length. 'Approximately 1/4 inch greater diameter could be
accommodated by the sampler but not by the transport'tubee Four of the
various types of capsules which were lowered into the pump-bowl are shown
in Figures 18.5a through 18.5d. The capsule in Fig., 18.5a was used to trap
"either salt or gas. The capsule was first evacuated and sealed with a mea-
sured emount of salt similer to the fuel. On entry into the fuel pump, the
- sealing salt mgited and salt or gas (depending on the capsulé elevation in
the pump bowl just prior to salt ligquefaction) was drawn into the capsule.
| When the beryllium addition capsules were withdrawn, they were en-
crusted with solids as shown in Figure 18.5b., As a result, the capsules
were difficult to loed into the bottom part of the fransport tubes even
though the plastic liner was removed. Also the enriching capsules were
sometimes difficult to load because of their length (6-1/L4 in.). The ca-
bles of these lohg capsules were occasionally caught in the threads of the
mating parts of the transport container and were difficult to unload at the
-hot,ceil.~r ' | L

The.iadiation levels of the beryllium capsules and the empty enriching

capsules fiere more than four times higher than the steandard 10-g sample
when delifiered to the analytical laboratory. N .

_ ' Miscellaneous Sempler Problems — During the latter part of 1966, 8
cotter key had come out of the top of the lower hinge pin on the access
fiort door. The‘hinge pin had then worked out of the top of the:hi'ngeo
Using only'the manipulator, the pin'was repositioned in the'hinge, 8 new
cotter key was inserted in the pin and the key was spread to lock it in
place. During this repair, the lowei key was dislodged and was also re-
paired.- ' | - )

Operation of the pneumstic clamps on the access port door occasion-
ally resulted in gaseous fission products being vented through the operator
discharge lines. Therefore a small charcoal filter was added to the op;
erator discharge lines to prevent this activity from'being discharged to
315

ORNL-DWG 67-4784

|_.—STAINLESS STEEL
|  CABLE

  

VOLUME

|_u—Ya~In. 0D NICKEL
2060 ——__ o |

- NICKEL CAPILLARY.

LiyBe F'4\

 

 

 

Fig. 18.5a Freeze Valve éapsule

PHOTO 9L278

 

 

Fig. 8.5b Nickel Cage
Beryllium Rod Assembly
after Exposure to MSRE
Fuel Salt for 10.5 hr

 

b e i e bt ¢

  
  
 

MIST SHIELO~—]|

 

 

 

ORNL-DWG 87-4783

CABLE

/

1'/2—In. PIPE

  
  
 

| | | |
1 v
1
Z
n
A Ui
A
h A ~———Cu0 IN STAINLESS

STEEL CLOTH BASKET

 

 

 

}
| GAS
| INLET
- ‘! .
=z Ao e - :?_:_E-
~ TS eveL sar U (|
- T T ||l =
-d - .- -

 

ONE ¥ -in. HOLE
THREE Y5~ in. HOL
THREE Y%g-in. HOL
THREE Ysp-in. HO

 

 

]

 

T F T T T Tt T T

 

 

 

 

 

 

 

 

 

 

 

 

o L L L L — S — .

 

 

 

3 Be RODS Y4~in. 0D x 1

SECTION B-B é

 

'/.;-in. LONG

ORNL -DWG 69-13105
4 in. BETWEEN WINDOWS

  

S,
ES,
LES

WELDED TAB WITH Y4g-in. HOLE

 
  
 
   
    

SOLID NICKEL SPACER

 

Lot 3/ -in, OD x Vap~in, WALL
NICKEL TUBING

T

L WINDOWS ———— |

A

A

 

A

—%g-in, OD X 3-in, TALL
GRAPHITE CYLINDER

 

 

 

 

 

 

 

 

 

| o— WELD ™
A a— 3, -in.~0D Ni TUBE

 

SHOULDER FOR GRAPHITE
CYLINDER TO SIT ON—1

-7
Ll
N
M
N
N
N
'
N
N
N
N
N
M
N
N
N
N
N
»
N
N
N
N
N
N
N
W
h
i

i — Ya-in.~DIAM PERFORATIONS

U U U
b 9: :O;-»p A?OUT EVERY Y4 in,
[T og_l

w

 

 

 

 

|
‘L et
N P’

Fig. 18.5d Surface Tension Capsule

 
 

316

the atmosphere via the stack., Also the vent valves from the gas operators
were kept closed except during openings and closures of‘the access door.

The neoprene seal at the access door continued to deteriorate., 1In
April of 1967, the increased buffer gas leakage across the inner seal was
countered by lowering the buffer pressure requirement in the safety inter-
lock system from 50 to hO psia. In July 1967, the safety switch setting
was lowered to 35 psia and the helium flow rate to the buffer seal was also
increased so that a satisfactory buffer pressure could be maintained.
During the intensive sampling of the lest run, the access port door fre-
quently required repositioning with the manipulators before the second set
of clamps were actuated to lock the door and obtain an adequate seal.

The cause of thé detefioration of this seal has not been determined.
The radiation level of some‘of the samples taken was in excess of 1000 R/hr
at 3 inches and during boot replacements when the sampler was empty, the
radiation level inside Area 3A was approximately 100 R/hr. Although some
long-term deterioration may be attributed to radiation damage, more likely
the reduction in buffer pressure during the last two runs can be attributed
to the fact that the left center Knu-vise clamp was loose and thus ineffec-
tive in its closed position.

The flexible containment membranes (boots) on the manipulator contin-
ued to be a chronic source of trouble. Generally, when a leak developed,
the sampling cycle would be completed before repairs were made; however,
during the remainder of the sampling cycle, containment was abetted by

throttling the boot evacuation valve to assure that any leakage would be

‘into the plenum between the boots and through the vacuum pump to the stack

rather than to the work ared. After full power was achieved, replacement
of the boots became somewhat more involved because the manipulator assembly
was quite contaminated. During & boot replacement in early 1967, the radi-
ation level of the manipfilator arm was 300 R/hr at 3 in. It was later de-
contaminated and saved for possible future use.

Also while sampling during early full-power operations, the radiation
level at the operating area increased to approximately 30 mR/hr. The radi-
ation level was reduced by an order of magnitude by changing the sampling

procedure so that a purge of helium from Area 1C to the pump bowl was
 

 

 

317

maintained when the operational and maintenance valves were open. Also on -
retrieval into Area lC,rthersamples'were purged for 1-1/2 hr to the off-gas
- system before further removel operstions were continued.

- On two occasions during the routine transfer of capsules between
Areas 1C and 3A, a capsule was inadvertently dropped and rolled out of
- sight and out of manipuletor'range in Area 3A. These were later found and
recovered by the use of s mirror and a grappling fork which were lowered
through the removal valve., '

During the latter part of Run 19, routine sampling weas suspended for
approximately four days and the sampler was adapted;to accommodate a col-
limator and & germanium crystal detector atop the sampler where the carrier
cask is;normslly positibned.j_A plastic plug in the removal area permitted
the unobstructed view into Area 3A where samples retrieved from the fuel
pump were positioned. Gamma4rsy spectrometry data were then collected on
short-lived fuel fission products within 50 minutes of their removal from

the circulating salt stream.
l8.l.h‘ Discussion and'Conclusions

The main operational problems encountered with the sampler-enricher
resulted from 1) manipulstor boot failure, 2) lodging of the capsule or
_latch at the sampler tube entrance, operational valve, or maintenance valve
with subsequent cable tangling, and 3) buffer seal lesks at the operation
~and removal valves and at the capsule access door,
| The cause of some of the boot failures were not determined because
the leaks were small and the actlons which produced the ruptures wvere not
elways coincident with the dlscovery of the leak. Also the boots were not:
examined in deta11~because,of the-high radiation level of the boots and
| manipulator_assembly, Itpis;believed that the boot failures which could
- not.beiattributed to any"spegifie'action occurré@'during'the capsule load-

'ing operation where the manipulator is retracted and pushed downward and
| to the left in order to load the sample into the liner of the transport
"“tube bottom. At times (almosfirlnvariablyW1th the long capsules), the
sampler operator had to resort to pulling the manipulator with both hands
in order to retract it.fsrhenough to insert the capsule and latch pin into

the liner. Had the capsule loading station been mounted to the left of

 
 

318

the manipulator centerline, it could have been located closer to Area 1C
and thus require less manipulator travel. Another possible alternative
would be to fEb:icete the boots with fever convolutions; however, the con-
volution fiidth would have to be increased to maintain the same extended
lengfih, The retracted thickness of the boots would then be decreased per-
:mitting the manipulator to be pulled back farther than previofiSlyo

- Why and where the capsule or latch occasionally lodged in the sampler
tube during cepsule insertion has not been satisfactorily answered. How-
ever, the use of the proximity switch minimized the possibility of serious
| cable'tangling problems, Perhaps two additional detection devices mounted
&bove and below the operational valve would have determined the exact -
trouble spot. ,

The buffer seal leak at the access port door started to increase in
January 1967 and continued to increase slowly for the remainder of MSRE
operations. Perhaps some ‘of this dete;ioration can be attributed to radi-
ation damage to the neoprene geskets. The radiation lefiel in Area 3A was
100~200 R/hr during intensive sampling. Also, the accees door was salways
in the closed position except during capsule transfer ‘therefore the neo-
'prene was almost always compressed which over the years has probably caused
it to yield to a thinner protrusion. In addition, the stellite plates on
the access door appear to be gouged slightly by the stellite;tipped Knu-vise
clamps. Also the left center_clamp'wae_discovered to be loose in its closed
position during the last run. Therefore, it is believed that ordinary wear-
and-tear of these components was the primary cause of seal leakage. The
beating that the door took could possibly have been iesSened by a flow re-
strictor to the clamp actuating cylinders. '

The top seals of the buffered velves developed legks whereas the bot-
tom seals apparently suffered only slight deterioration which indicates
that particulate matter which occasionelly drop on the valves was the cause
of seal deterioration. Consequently the volumes above these valves'showed
a gradual incresase in pressure and required constant venting. The least
desireble of these leaks was that at the removal valve. On removal of a
sample, the transport tube and removel tool which are withdrawn from
Area 3A are locked in position asbove the removal valve and the removal”

valve is then closed. If an apprecisble delasy is encountered between the

N\
g P b i RS L AT prgge ™ W pLLats 7 et e id e L ket et Tl RS SR G TR L s R s e pmmmn e mmm nAsmmmm o

319

time that the valve is closedfand‘the time that the transport tube is with-

- drawn from the removsl seaififito'the'shipping cask, appreciable pressure

(due to seal leakage) can build;up,intthe removal area and blow contami-

nation out.of the removal area to the top of the sampler and shipping cask

vhen the transport tube clears the two O-rings of the removal seal. Swab-
bing the removal area with wet wipes on a monthly basis greatly reduced

the contamination problem encountered in this area. During intensive sam-

pling, this area should be cleaned more frequently

The specifications for the electrical insulation on the lead wires
near the penetrations at Areas lC‘and 3A should be rev1ewed. A failure
at the connector pins occurred in May 1966 resulting in open circuits in

the insert and withdraw modesiand in the upper 1limit 1light. During in-

- spection of the 1C assefibly which was removed in August 1967, the insula-

“-tion on the lead wire was found to be very brittle and most of it was un-

av01dably stripped durlng removal of the drlve unlt In November 1967 an
electrical fault occurred in the same circuits as those in May 1966 except

the failure occurred in theewires between the two containment boxes. It

is not clear whether these failures-can be attributed to radiation damage

to the'insuletion or possibly to excess temperature caused by resistance
heating in the lead wire aue;to blocked rotor current (0.3 amps) in com-
binstion with the heat supplied by the illuminator lamp, ~

To increase the precision of uranium isotopic analyses, provisions
should be made to clean Area'3A periodically, especially after enrichments
are made, During the: enriching procedure, -small quantltles of enriching

salt could easily adhere to the ‘manipulator and/or be Jarred 1oose from

_ the capsule and fall to the floor of Area 3A where all capsules are laid
’-temporarlly vhile swapplng capsules to and from the latch.and the transport
- tube, A very small amount - of enrlched salt adhering to a8 sample submitted

1for isotopic analy51s could dlstort the 1nterpretation of the resulting

- anelyses.

 
320

18.2 Coolant Sampler

In prificiple, sampling{df the coolant system was the same as that for
the fuel system. However, the induced activity in the;coplant salt was
very short-lived and there was no residuasl radioactivity in the'sample;
therefore, no shielding was1required afld only single containment was neces-
sary during transport. Direct manipulation of the sampie by one hand |
through a glove part s1mp1ified sampler construction énd operation.

18.2.1 Description of the Coolent Sampler

The coolant salt sampler, shown in Fig. 18.6 consisted essentielly
of a dry béx connected to the‘éoolant pump by a transfer tube. Two ball
valves isolated the sampler from the pump bowl which has & mist shield and
capsule guide similar to that in the fuel pump. The 1-1/2 in. Sched-l0
transfer tube was similer to that used for the fuel system exdept,that no
expansion Joint was required because the coolant pump position did not
change with system temperature as did the fuel pump. The sampler was pro- ;o
vided with lighting and viewing ports and the necessary capsule trensfers gfi;
inside the sampler were done with one hand through a glove port. Sliding
gas seals at the top of the sample carrier along with a vacuum pump and a
helium supply were used to reduce oxygen contaminetion to a minimum during
sampling and during sample transport to the analytical laboratory.
A check-1list type procedure along with a system of locks and keys for
valve manipulation was used to assure proper sequence of operations. A
key was used to unlock one valve, which could then be opened. When the
valve was opened, it also locked the first key in position and released a
second key. Removal of the second key (which must be removed to unlock
the neit_valve in sequence) assured that the first valve was locked in the

proper position.

v

18.2.2 Opereting Experience with the Coolant System Sampler

The coolant salt sampler operated reletively trouble-free throughout
MSRE operations. Only 81 coolant system samples were teken as opposed to
T4 cycles for the fuel system.
321

 

PHOTO 68262

 

 

 

 

 

 

 

 

 

 

 

 
 

322

Early in 1966 during one sampling operation, the latch failed to slide
freely through the transfer tube. No reason for the difficulty'cculd be
found. However, the outside - dlameter of the latch was reduced and no simi-
lar trouble was reported

Also in the first part of 1966 one pin in an 1nd1cator 11ght circuit
shorted out at the penetratlon durlng a sampllng cycle. The receptacle was
removed and & nev. one welded in place. | |

Durlng the same period, the leak rate from the lower seal of the re-
moval valve increased. After the valve was disassembled, cleaned, and re-
asseflbled; the seai was satlsfactory, however, after apprcx1matelyge1ght
additiona1~semples, phe seal leakage again increased and Wae corrected by
replacing the valve-seele. ~Also an extra washer was added to the valve to
permit more mechanical pressurerto be_applied to the seals. Thereafter the
coolant sampler,operated_trodble-free for the finalfthree years of MSRE

operation.
18.2.3 Conclusions and Recommendations

From an operational viewpoint, two improvements may be worth con-
51dering. A

(1) The glove port cover should be hinged to sw1ng out horlzontally.
On several occasions when the clamps were removed, the glove port cover.
slipped from the operator s hand and dropped freely to its stopped posi-
tion. In doing so, it could have pinched & bundle of nearby insulated
electrical wires. 7 ) .7

- (2) The sample carrier should be lengthened approximately two inches

to accommodate the longer capsules which were used during“the'last run.
 

 

- 323

19.  CONTROL RODS

M. Richardson

| -19.1 . Description

Three control rods were used in the MSRE core vessel, The rods were
not exposed directly to the fuel salt but were contained in three "blind

well" thimbles which were located in three of the four channels surrounding

~the vertical centerline,of_theigraphite core. The thimbles were straight

within the core veeeel; butdabove'the reactor access flange they were off-
set — using two 16-in. radius bends to provide space for the core sample

containment standpipe. A roller was located at each bend to reduce fric-

‘tion, The control rod drive mechanisms were vertically positioned around

the standpipe. The control rods were fdbricated of flexible metal hose

with hollow center cyllndrlcal poison elements (38 per rod) beaded over

the lower 60 in. Rod motion was ‘supplied by motor driven 1/k-in. roller

chain and sprocket drive units. Included in the reverse locking power
train is an overrunning clutch whlch permits powver transm1551on in the in-

sert direction only. A magnetlc clutch was used to release each rod from

- the power. train gearing and permltted free fall of the rod into the core

at A0.h g.

Continuous indication of each rod position was prov1ded by two synchro

"torque transmitter-receiver palrs, one "flne and one "coarse." - A potenti-

ometer provided p051tion ;ndlcatlon for safety interlocks in the reactor

£i11 circuits._

The upper and lower limits of rod travel were controlled by four (2
upper and 2 lower) mechanically actuated switches.

- The poison elements were cooled by cell air (95% Nz) from the . compon— ‘

ent coolant system.r This entered the upper end of the hollow control rod
- and "exhausted radlally into the thlmble from & nozzle at the lower end- of
~the rod. | | | |

.Changes 1n pressure drop as thls nozzle moved through a flow restrlc—,

tor built 1nto the bottom of the thlmble prov1ded a rod position reference

~ point (flducial 'zero)., Comparlson of this with the rod position as indi-

cated by the eynchro rod position instrumentation was used to monitor

stretching of the rods and other malfunctions.

 
 

324

The rod drive units were identical as,fiere the rods except for a
slight variation in lengths of the individusl rods due to the differences
of the thimble offsets. This allowed interchange 6f_components for repairs
or trouble-shooting. For identification of a rod assembly, it was neces-
sary to specify the thimble number (T-1, T-2, or T-3), the rod number (R-1,

‘R-2, R=3, or R-l) and the drive number (V-1, V-2, V-3, or V-k). The rod

assembly connected to the servo controller was considered to be the regu-
lating rod and the other two rods were the shim rods. The rod assemblies
could be used interchangeably as shims or regulating rods by shifting the

~out-of-cell rod drive disconnects. Any one rod was capable of teking the

reactor subcritical.2! |

The rods and drive assemblles were not designed for in-cell malnte-
nance, Repalrs on the drlve units were done after dlsassembling the rod
from the drive and removing the drive unit from the reasctor cell by remote
maintenance ‘methods.

The control rod shock absorbers naed the general principles of a rypi-
cal hydraulic shock absorber but differed in that the working "fluld" con-

- sisted of 3/32-in,—diameter steel balls. At the end of a scram, the bottom

" face of the shock ebsorber cylinder struck perfianentiy-mounted steel blocks

which were belted to the hofising° The shock ebsorber plunger, to which the
control rod was attached, continued to move downward and was decelerated

by the fbrees developed by the”bnffer springs and by the flow of the ateel
balls around a knob on the plunger° The atrokenlength of the shock ebsor-

bers were adjusted to 3.5 inches for each rod,

19.2 Initial Testing

The rods were installed in the reactor vessel in January of 1965. The
limit switches were adjusted to eoincide with the rod position indicators
on the main console and electrical continuity tests were performed on all
drives prior to assembly to the control rods. Satisfactory ambient tem~
perature performance waS“checked(with'the_resnlts given in Teble 19.1.

/
325

Table 19.1 inifiial-Tests of Control Rod Assemblies

; | | "Spare‘
Thimble No. | - - P-1 - T2 T-3 - Components
Rod No.  R-l  Re2 R-3 R-4
Drive No. - V=3 V-2 V-1 V-l
Motor current withdraw —-ampé 0.6 0.6 2 0.6 - 0.6
Motor current insert — amps 0.59  0.58 0.58 . 0.59
Shock stroke, inches = 2.85 3.6 | 3.7 3.0
Scram time, secs’ 0.835 0.752 0.775 _—
Air flow, rod — scfm 3.8 3.9 3.9 e
Rod travel speed, in,./sec 0.53 0.53- - 0.53" 0.52
Full rod travel, inches 50,91 50,95  50.95 50.9
Fiducial zero, inches 1.k 1.hb o 1,35 -

19.3- Periodic Testing

During reactor operation and especially after power operatlons com-
menced it was not possible to run extensive tests on the rods. However,
it was necessary to do sufficient testing to assure that the rods would
scram if needed. It was also important to know that the rod position indi-
cators were functlonlng proPerly as these were used in m&klng reactivity
zbalances. To accompllsh this and to aid in antlcipatlng necessary mainte-
~ nance; the following perlodlc tests were made, Co /

A. Rod Scrams ——-Scram.tests were | routlnely conducted to determlne
‘rod scram times before & fuel: flll snd before critical operatlon. Fach
_rod vas withdrawn to a helght of 50 in. The magnetic clutch current was
broken by trlpping the menual scram switch. The rod'would free-fall to the
fully-lnserted position. - The 50—1n. drop was electronically timed from the
moment of release to the 1ower llmlt switch. The maximum permissible scram
time was 1.3 sec/50 in. | | o

B, Rod Exercise —-Sincé rod scram tests could not be made during

nuclear operation, each rod was exercised daily to demonstrate that they

 
 

 

 

326 | I

would move on command and also to flex the metallic hose. The flexible

hoses tended to stiffen with time if allowed to remain in one positién for

prolonged periods. Observation by the reactor operator of the rod—positién

indicators during the exercise ascertained that the synchros were working
properly and that the rods operated freely in the thimbles.

C. Fiducial Zero — As described above, the fiducial zero vas a fixed
reference position in each of the control rod thimbles which related the
actual rod position to the rod position indicated by the position-indicating
potentiometers. Fiducial zero tests were routinely made before a fuel fill

and before critical operation.

 

19.4 Operating Experience

Except for difficulties with the rod assemblies in thimble T-1, opera-
tion was satisfactory from Both an operational an& maintenance standpoint.
There were no unscheduled shutdowns because of a control rod failure. The-
testing program was adequate to allow advence maintenance planning and work
was done in conjunction with maintenance in other areas at the time of a
scheduled shutdown. |

During over 3,000 rod scrams from 50 inches (see Table 19.2) and num-
erous others from different elevations, there was only one time when a rod
failed to scram. Except for this and one other period, the scram times
were all less than the 1.3 seconds specified in the safety limits and wére
‘usually less than 1 second. The number or amount of rod movement'is_not
known. Oncerpositioned, there was little movement of the shim rods. The
regulating rod, normally the assémbly in thimble T-1, moved little at steady
power until bubbles appeared in the reactor. From that point on the regu-
lating rod moved frequently under servo control tb compensate'for the action
of bubbles on the reactivity.~ Sevéral rod jog tests were made and it is
estimated that the regulating rod moved about 720 jogs per hour at 0.l in,

_ per jog for a total of 32,400 jogs. The total movement was about 1h,000

inches.
 

327

Teble 19,2 Estimated Number of 50-inch Rod Scrams

 

Thimble No.  Rod No.  Drive No,  No. of Scrams

 

1 Rl V-3 750

T-1 R-1 V-1 150
T-2 . R-2 V-2 930
-3 ~ R-3 V-1 800
-3 Rk V-1 k30

 

Total 3060

 

The estimated life of the control rod motors was 10,000 hours at a
reactor pover of 10 MW (Ref. ”h9) The condition of all drive motors, lu-
brication, W1ring, and gears remalned satisfactory throughout the entire
operation. These were 1ast 1nspected in July 1969 et whlch time the reac-
tor had accumulated 92,805 Mwhrs. . ,

The hlstory of the MSRE control rods and drlves are best summarized
in Figs. 19.1, 19.2, and 19.3., Control rods R-1 and R=2 remalned in ser-
vice the entire operating period in thimbles 1 and 2. cOhtioi rods R-3
and R-l were 1nterchanged in thlmble 3 as shown 'in Fig. 19. 3 fThe component
1 failures of the drives are. tabulated in Table 19.3. |
| Heat—up and pre-crltlcal salt circulation commenced 1n January 1965
and during the early perlod of this operation, dlfficulty was encountered
with the limit switches. The 1ndications were that the sw1tch operators
were sticking or galling randomly on all three rods, Temporary repairs
~were made during the March and Aprll 1965 shutdown when'investigatioh re-
 vealed that the sllding meflbers of the switch operators had galled causing
the faulty operetion, - ,

The drives were removed from the reactor cell in July 1965 for instal-
lation of modified limit switch operators on all the units including the

 
T-1 ®TqUFYL UT SOFIquassy poy 3o Azénsrn °o7AI8S  1°6T 374

.. Maintenance History

" FIDUGIAL ZERO (in.)

W .

'SCRAM TIME (sac)

 

Installed temperature switch
Replaced Irimit‘ switch operator

Replaced potentiometer

—] 0000

8 & §
T

- —— 0OLO

"v 0080

-1 0080

-1 000°}

MAY

JUNE

JULY

AUGUST
SEPTEMBER &

OCTOBER
NOVEMBER

DECEMBER

 

 

 

 

 

Replaced temperature switch
Replaced potentiometer

w

-

 

ATTEWASSY 1—U 'E-A

 

JANUARY - -
FEBRUARY
MARCH
APRIL
MAY
JUNE .
JULY
AUGUST

9961

 

SEPTEMBER

OCTOBER
NOVEMBER

DECEMBER

 

ol a

 

JANUARY
FEBRUARY

MARCH

 

APRIL

MAY

JUNE

JULY

AUGUST

[ SEPTEMBER
OCTOBER

L9681

 

NOVEMBER
DECEMBER

 

 

F June

JANUARY
FEBRUARY

MARCH
APRIL

MAY

 

 

 

 

Bo6I

JULY

AUGUST
SEPTEMBER

OCTOBER
NOVEMBER

DECEMBER

 

Fine synchro failed
Potentiometer noisy
Rod No. 1 used as shim rod

Replaced V—3 drive with V—4
Rod No. 1 regulating rod

 

 

 

 

JANUARY.
FEBRUARY

MARCH
APRIL

MAY
JUNE

JULY

€961

 

AUGUST

|_SEPTEMBER_
OCTOBER
NOVEMBER

 

DECEMBER

 

8ct

209—€L DMA—INUHO

 

 

 
Z-1 STQUTYL U SOJTqUIssy poy jo KI03ISTH @9TAleg Z°6T *91d

 

FIDUCIAL ZERO {in.) SCRAM TIME {sec)
o o

Maintenance History ' g § § . § g -§‘
el

 

MAY
JUNE _
JuLYy
AUGUST = =

X

 

. w—t 00L0

Installed temperature switch { l ' ] ' I
Replaced _Iimit switch pm ‘ .

 

 

 

 

 

OCTOBER
NOVEMBER
DECEMBER
1 JANUARY
FEBRUARY
MARCH
APRIL .
MAY
- JUNE =
JULY §
AUGUST. ‘
{ SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
| JANUARY _
FEBRUARY
MARCH
. APRIL -
< , _ e
Replaced fine synchro . ’ » L MAY
d ' b JUNE
g - : JULY
. , ' AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
, B B JANUARY
° o _ FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
JANUARY
| FEBRUARY
MARCH
APRIL
MAY
JUNE . -
JuLy e
AUGUST
. SEPTEMBER
ot , OCTOBER
, Y NOVEMBER
| - C . ) DECEMBER

 

l

 

 

 

 

" A18W3SSY Z-Y 'Z'A"""""'

s e

 

 

e

 

-

 

 

-

 

 

 

 

£961

 

 

Replaced fine synchro

 

 

8961

 

 

 

 

 

 

 

 

 

— s

 

Replaced potentiormeter

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6ct

SEPTEMBER @~

€09—€L OMO—INKO
Maintenance History

FIDUCIAL ZERO {in.)

~N W

SCRAM TIME (sec)
o

3 8

0060

 

Installed temperature switch
Replaced limit switch operator

Repaired roller in T—3
Repaired R—3 upper hose

8 8 & B
bhor ]
)

—1 00,0

MAY
JUNE

 

[JuLy

 

AUGUST -
SEPTEMBER §{

 

 

{ OCTOBER
| NovEMBER

 

 

DECEMBER

 

Bent limit switch operator

R—3 sticking in thimble —
Replaced R—3 with spare
rod R—4

 

v-uls—u——

ATENISSY Z-H * | —A —masieits

*>

 

JANUARY
FEBRUARY

 

 

jMARCH
APRIL

 

 

 

9981

 

 

 

OCTOBER
NOVEMBER
DECEMBER

 

 

 

Bent limit swi_tch operator

 

|

AT8W3SSY P-4 'L—A

=

[June
lsuLy

 

JANUARY
FEBRUARY

MARCH
APRIL

AY

 

 

 

 

 

I

196

 

AUGUST

 

| sErTEMBER

 

OCTOBER
NOVEMBER

 

 

IDECEMBER

 

€-1 PTQWIYL UT SOTTqUossy poy JO LI0ISTH 99TAlaS €' *S1d

50 inch scram time increase

 

| SANUVARY

| FEBRUARY
MARCH

 

 

 

APRIL
AY -
JUNE
JULY
AUGUST

 

=

 

961

 

 

| SEPTEMBER

 

OCTOBER
NQVEMBER
DECEMBER

 

 

 

Added 1.5 Ib weight to V—1
0.9 scram time

Repaired rod R—-3

No. 3 assembly used as reg rod

No. 3 failed to scram on demand

" Replaced R—4 with R-3

Replaced potentiometer

 

 

 

—Y -
»
e aal

 

 

g epa—t— 1
°
b A
o -

 

 

AT8N3SSY
£-HI—-A

S—

 

 

JANUARY
FEBRUARY

 

 

MARCH -

 

APRIL

AY
JUNE:
JULY

AUGUST

 

=

 

6961

¥09-€L OMA-TINYO

 

 

SEPTEMBER

OCTOBER

NOVEMBER
DECEMBER

 

 

 

0te

 

 

 

 
 

 

 

331

 Table 19.3 Drive Unit Component Failures

 

 

 

 

Drive Unit No. = V-2 V-2 V-3 Vol
. e , (Spare Drive) Total
- S8ervice Time as Reg., Rod 4 mo., = --- 51 mo. 5 mo, _
Service Time as Shim Rod 56 mo. 60 mo. L mo. w=m
Temperature switches T - 1
Limit switch operator 1 1 1 1 b
Limit switeh operator . ,
push rod : -2 - - - 2
1000-R potentiometer - - h - |
Fine position synchro L
transmitter 0 2 1 | - 3
Coarse position : |
synchro transmitter 1 ) - - , - 1

 

- spare (V-4). The new operators contained case-hardened bearing_surfaces,
there were no more difficulties of this type with the switch operators.
Examinationrof the control rods during this period revealed that the
upper wire sheathed upper hose of the No. 3 rod was badly worn at a point
28 in. below the flange. Exemination of the No. 3 thimble revealed that
the upper roller, on whichfthe.coatrol rod operates, did not rotate which
caused the severe wear te.fhe rod. The roller was-replaeed and & new hose
glnstalled on ‘the No. 3 control rod.._ | |
-~ During this period temperature SW1tches were added to the lower drive
~meehanlsm. These small bi-metallic switches were to alarm should the tem-
~ perature within the drive unlt cases exceed 200°F., ‘Without the dovnward
_:sweep of gas through the housing, convection of the heat rising from the
 thimble could cause & rapld temperature increase and resultant demage to
?the drive assembly. After theirod.drlves had been reinstalled and the re-
actor heated and filled with salt, the switches were fiested.by turning off

 
 

 

332

the gas flow. Since no alarms oécurred, another method of measuring the
drive assembly was devised which, although not precise, was adequate for
operation. This was done by shutting off the integral drive motor cooling
fan and measuring the resistance of the windings of the fan motor. “A
slightly greater than- 10% ‘inerease in re51stiv1ty equaled.V50°F change.:
w1nd1ng temperature,‘wfrom the measured resistance and us1ng the temperature
'coefflcient of res1stance of the wire, & curve was plotted relatlng the
measured resistance to the approximate motor temperature. This value was
roughly related to the drive housing temperature. | ' fh-

In early operation, the control-rod servo system had not functloned |
properly due to coasting of the regulating rod motor and shim locating mo-
tor. Brakes were installed in both of these (see 2h.5).
19.4.1 Rod Assembly in Thimble T-1 ’

Figure 19.1 gives the rod drop times and fiduc1al zero values as well
as the maintenance history for the rod assemblies located in th1mb1e T-1.
This was first made up of Rod R-1l and drive unit V-3 and was usedeas the
regulating rod, except for a four-month period, for the entire reactor op-
eration and consequently performed the bulk of the control rod service.
Other then replacement of faulty drive components, there was little diffi-
culty with this assembly.. - , ' '

. - The .1000-R: potentiometer position indicating potentibmeter'failed 3
times due to an open.resistance coil and 1 time dQue to & wvorn resistance
coil. This difficulty had been anticipated as it is a common failure for
this type.component in continuous service. L :

" One-temperatureksvitch failed due to an electricel ground at tfie wire
connections at the switch. The ground was created by bumping the sw1tch
during assembly of the drive into the case,

.. There was &n eapparent.shift of 1/2-in. in the fiducial zero ‘position
in November of 1965. ‘The.change was attributed to slippage of the'chain
on the sprocket but examination.and testinglfailed-td reproduce the slip.
The 1/2-in. deviation was:recovered involuntarily in July of 1967~whiehi
wouldfinddcate-that the exact cause of the shift is not known. The fidu-
_eiallzerq position plot in Fig. 19.1: shows the;gradualiincrease,,$1/2 inch,
 

 

 

 

333

in the length of theAcbntrolfrod'from'the effects of usage. The change
in length which occurred in September 1969, is the result of changing the
drive unit. | | _

‘The rod scram times remained at 0.8 *+ .05 seconds from installation
until the spring of 1969. In March of 1969, the scram time increased to
1.03 sec for a 50~in, dr0prand5also during this period, there #as a failure
 of the fine position synchro transmitter and the 1000-R pot. Analysis (see
19.5) of the rod acceleration and velocity during a scram fevealed that

. there was an area of drag in the lower 20 in. of the scram. From the sbove

 

evidence, the V=3 drive was féplaoed with the spare drive V-L. With drive
V-4 and rod R-1 in thimble T-1 the scram time returned to n0.8 sec until

~ shutdown. A detailed inspection and maintenance program is planned for 3
drive V-3 but is as yet incomplete. . o

19.4.2 Rod Assembly in Thimble T-2

- As shown in Fig.'l9.2,_the assemblyrin thimble T-2 was made up of V-2
drive, R-2 rod. This.assembly was used as a shim rod, except for short
test periods, for the entire MSRE operation. Other than the mechanical
:failures shown in Fig. 19.2,}there Werélno operating difficulties.

 The mechanical failureo included two fine poSition synchro failures
 and one potehtiometer;':Ekémifiation of one of the syhohrosl(the'radiation
level of the synchro was NlO mR/hr) revealed the failure to be the wiper
arm contact between the rotating shaft and the windings. '

' The fiducial zero history shown in Fig. 19.2 clearly shows the pro-
gressive stretchlng of the rod to be ml/2 in. from the tlme of 1nstallation
to the end of operatlons.'i,a o '
~ 19.4.3 Rod Assembly in Thimble T-3°
_ This assembly was. made up of V-1 drive and.R—S rod at the time of in-
.3stallation es shown in Fig. 19 3. As 1nd1cated by the scram history shown

in the figure, this assembly was ‘the most troublesome of the three MSRE
'icontrol rods. The drive units were removed during the precritical mainte-
:nance period in the summer of 1965 to install the modified limit switches
described elsewhere. The No,_3rcontrol rod was 1nspected at this time and
found to have a torn Sectioh'in'the_uppor hose, 28 in. below'the flange.
The damage to the rod was the result of the_failuré'of the upper roller
 

 

 

A e g e e

 

334

in the No. 3 thimble vhich had jammed and would not rotate. The roller
and rod were repaired and after reassembly, the rod scram time was <0.8
sec. B S A
_ During the low power period of operation,'it was found that after a
full seram the lower limit switch would not clear when.the rod was with-
drewn. _It was necessary to push the switch actuator by withdrawing the.
rod to the upper 1limit to clear the lower limit switch. It would function
nofmally'until the rod was again scrammed. The bottom end of the limit.

switeh push red.was found to be bent below the lower retaining bushing.

The rod had been bent by striking the lower housing flange as the result
of & weak push rod-springe Repairs consisted of straightening—the;push-rod
and installlng a stronger spring. o o , .
Later the control rod commenced to stick in the thimble at a point
¥1-1/2 in. above the fiducial zero position. -The rod could be  freed by
Jogging and #ould'hang in no other pesition. Directjexemination of the

thimble was not possible, but the exterior of the rod was examined and ap- -

peared to be in good condition. The difficulty appeared to be.that & sharp

projection such as a broken weld at the throat section of the fiducial zero

position existed on which the rod could snag. However, when the R-3 rod

was replaced with the spare rod R-L, it did not snag.

This rod operated within limits but the scram time increased slowly
fromr0,85 sec in May 1967, up to 1.26 sec in December 1968. Examination
of both the drive and rod revealed do apparent defects. Replacing the V-1
drive with the V-2 drive unit in the T-3 position for test purposes made
little change in the scram time. Therefore a l,S-lb #eightwwas ‘added to
the shock absorber assembly o the orlglnal V-1 drive unlt and 1t was re-
installed. This reduced the scram time to less than 0.9 secof ‘

The reactor was shut down in June 1969 by manuslly scramming all the
cohtrolrrods. The assembly in thimble T-3. (R—h and V-1) was at 35.0 in.

and did nof scram,59 The rod was inserted 0.2 in. and scrammed again, this

" time the rod dropped freely Repeated attempts"towmake the rod stick fol-

1owed thls incident bux vere unsuccessful Fig,w 19.h shows the results -of

»scram.tests at 2—1n° 1ncrements taken on June 2 befbre the rod was exam-

ined and clearly shows the ares of hlgh drag.
 

 

1.8

e

14

1.2

SCRAM TIME (sec)

06

04

- 02

-0

~ Fig. 19 ih'f_jlncr_émental Rod Drop Times

335

ORNL -DWG 69-89T6A

 

 

 

 

 

 

 

 

 

 

L o JUNE 2, 1969
8 | | e avsust 4,1969

 

 

 

 

 

 

 

 

 

0 10 -2 - 30 40 50
.~ STARTING POSITION (in.)

60

 
 

336

To correct this difficulty, Rod R-4 was removed and-replaced.by Rod
R~3. This was the rod which had earlier snagged in the thimble. Before
reinstalling it, the air exhaust tube (a poSsible‘causéof the snagging)
was filed smooth to eliminate all the sharp edges. The defective potenti-
ometer was also replaced on the drive {V-1l). The greatly improved perform-
ance of this assembly (R-l4 and V-1 in T-3) suggests that the trouble was
in the rod (R-4) rather then in the drive. The extreme radioactivity of
rod 3 (R-3) made it impractical to examine.

Figure 19.h4 also shows the results of drop tests conducted on August k4,
after all repairs had been accomplished for all three controi‘rods. This

was the condition of the assemblies during the final period of operation.

19.5 Improved Rod Scram Testing

It was obvious from the No. 3 rod failure to scram that the method of
testing each rod by’a single 50-in. scram was inadequate. The method
failed to expose any local areas of high drag.  The incremental method of
testing, scram time vs staftingnpoéftioh, es shown in Fig. 19.4, showed
no region of excessive drag for afiy of the rods. In an effort to provide
a quick testing procedure that wohld not reqfiire numerous rod drdps, yet
would reveal regions_of.abnorQal drag, a procedure developed for the EGCR
was renovated. ‘The:output'frdfi the 1000-ohm position potentiometer was
- smplified and transmitted by wire to the main ORNL area, where it was
passéd through a filter with a 5-Hz time constant to remove transmission
noise, digitized at 2000 samples/sec, and stored on magnetic tape. The
data was then sent to the IBM—360 computer for smoothing and analysis by
a program especially developed for fhis purpose. This procedure when acti-
vated during a rod drop from 50 in. gave about 1800 data points during the
drop. From this velocity afld acceleration curves were generated. Data
teken before repair of the sticking rod assembly (R-4 and V-1 in T-3) showed
a region of near zero acceleration between 26 and 40 in., in good agreement
with the incrementsel results shown in Fig. 19.4, After reinstallation, all
three essemblies were tested by both incremental drops and the single drop
computer analyzed method. Both methods showed reasonsbly constant acceler-
ations of 10 ft/sec? or mofe° A typical plot of acceleration vs position

is shown in Fig. 19.5.5%
 

 

POSITION - INCHES

Fig. 19.5 Rod Acceleration Curve

 

LEE

 

 
 

338

The signal filtering and data smoothing that are fequired.because of &u;?
the noise that is inherent in the position signal genérator and transmis-
sion system made it impossible to detect with éertainty a zero acceleration
zone 1éss than about six inches long if it is more than about six inches
below the scram starting point. | 7 "

Revised criterie wereAestabliShed-for the monthly testing of the-rods.
This consisted of scrdfiming the_rbds from 40 and 50 inches. These two
points were Just &bove the normal operating positions of the regulating
and shim rods. 'The:criteria were that the 50-inch scram time must'be less
than 1.00 sec, the 4O-in. scram time must Se less than 0.90 sec, and the
acceleration curves must show no area whiéh_indicates abnormal drag. Ir
an area of high drag was noted, the rod wouid be dropped from a point within
this area. This scram time must be no more than 20% longe; than normal for
that point. Prior to a fuel salt f£ill, the rods would be scrammed from 24
inches, which is their normal position for a fill. The scram times from
this poifit must be less than 0.7 seconds and the acceleration curve shbw
no abnormality.so

 

19.6 Mszintenance Experience

- Repairs perfdrmedron the drive units usually’cbnsisted of a simple
repiacement of worn components which were readily accessible after the unit
had been removed from the reactor ceil. In the event of major difficulties,
both a spare drive assembly and a spare control rod were available as re-
placement items. | ,

It was possible to perform direct out-of-cell maintenance to the drives
without excess exposure to personnel. The radiation level was about 600
mR/hr at the drive case (the_drive unit was removed from its case during
maintenance). The overall level of radiation at the drive mechanism was
gbout 200 mR/hr and, as én example, the level at the fine synchro trans-
mitter which was mounted on the gear case was about 30 mR/hr. Maintenance
on the control rod itself after criticality was very limited. Activity on
‘the upper hose was sbout 400 mR/hr at contact and the section containing

the poison elements was 20 R/hr at 18 in. after a two-year storage period.
 

 

 

339

Since the four drive assemblies were identical and the spare parts were
common to all the drives, the maintenance problems regarding replacement of

faulty components was not complicated. However, investigation as to the

- reasons for slow scram times, even when the areas of high drag were located

(see 19 5), was by trial and error method. Past experience had shown that
certaln areas such as the 7/16-1n.-d1am air tube, on which the control rod
slides, could become slightly gmlled or bent. These areas were examined,
adjusted, and the unit operated on a test stand from the main console. The
finai test was to reassemble the complete assembly in its normal operating
position in the reactor cell and proceed with the scram tests as described

elsevhere,

19.7 Discussion and Recommendations -

The overall operatien-of the rods and drives was quite good. There
was only one time when a rod feiled to scfam in the four years of use. At
no time were reactor operations terminated specifically for control rod
maintenance. _Construction'of_the assemblies was relatively simple since
very rapid response was not a requirement. The flexible controls rods were
adequate but the possibility of binding, breakage, etc., always existed.

It was possible to control the rod position closely, within 0.1 inch, from
the "fine" position indicator on the console but the exact physical loca-
tion of the poison elements was subject to weaknesses inherent to a semi-
confined flexible metal hose of this type. The only known position was at
the fiducial zero positlon in the thimble to which all. other p031tions were
related. '

It required about 12 heurs:tO'open the reactor cell access to approach

~ the control rod assemblies, the actual removal required about one hour,

" Location of the drives in a more readily accessible location is recommended.

The manner of attachlng the control rods to the drlves and the drive

_units to the thimbles was hindered by (a) poor lighting at the bolt loca-

“tion, (b) lack of space to insert more rigid tools to perform the bolting

which was done from a distance of about 18 feet, and (c) poor visibility

from the tool operator position at the maintenance shield. A modified

 
340

external, rather than internal, bolted flange attachment between the rod
housing and thimble which is clearly visible is recommended.
The "fine" position cynchro transmitters and potentiometers, such' as

were used in these units should be of & more durable type for extended

service,
 

341

20, FREEZE VALVES

) M. Richardson
,20;1‘ Introduction

The fiow of salt in the MSRE drain, fill, and proceés systems was con-
trolled by freezing afidrthawing short plugs of salt in flattened sections
of 1-1/2-inch pipe, called "freeze valves." This method of control was
adopted because of a lack of a proven relisble mechanical valve. ' Although
méchafiical.type valves would have had the advantage of faster action and.
the ability to modulate flow, the "freeze valve" concept had a good record
of satisfactory performance, the freezing and thdwing times were satisfac-
torily Short;'and the "off-on" type control did not impose any particular
hardship. . . -
| Désign and develoPment df the MSRE freeze valves began in 1960.%0 The
basic design was'estdblished.and_hgd been tésted by - the time the MSRE con-
struction had'begun.‘ Howevef, cbfisiderable effort was expended at thé re-
actor site before the valves were ready for routine operation. Successful
freeze valve operatibn is tqtallyidependent on the mannér in which the heat
and/br coolant gas is applied;and controlled. The\effectfibfflthe heaters,
type,of service, local environment, etc., made it necessary to "tune" each
of the valves for its spécific location. This effort continued until

criticality.
20.2 Description df'the;Design'of the Freeze Valves

The valves were located in the system as shown in Fig. 20.1. All

twelve valves were.similaf;andVWere made of'l—l/2—in.'Sched~h0 INOR-8 pipe.

The valve body, oriflattened séction, was cold-formed to make a 2-in.-long

. flat, &2-in.,wide, with a l/é-in.'internal thickness or flow area. A cool-
~ 1ing gas shroud was welded around the valve flat to direct the gas for

~ freezing the salt. The annulus formed between the valve body and shroud
 was 1/2-in. The 3/l-in. gas inlet and exhaust lines were welded to the

shroud. See Fig. 20.2.
A pair of thermocouples located at the center of the valve inside thé

air shroud, and a pair upstream and downstream Just outside the air shroud

-

 
 
     

LINE 204 LINE 206

87
FV=204 FV-206

342

TO COOLANT SALT DRAIN TANK

TO AND FROM COOLANT SYSTEM

COOLANT SYSTEM FILL AND DRAIN LINES

‘FV-104

 

FFT

 

~#———— LINE103 ———&—

FV-105

 

ORNL-DWG 73-605

REACTOR

  
 

LINE 519

 

 

ARRTERRRNRNNNNNN

 

FV-103

 

:

 

 

FV—106 LINE 110

 

 

 

 

Y
TO CHEMICAL PROCESS

FV-110, FV=-111, FV-112

 

 

FD—-1

DIFFERENTIAL
V-CONTFK'.)LLEI"R

FUEL SYSTEM FILL DRAIN AND TRANSFER LINES

Fig. 20.1 Location of Freeze Valves
 
    
    

   
 
  

 

: '\rk‘- AN Mg. '
LN ¥ !
\\ ALL THERMOCOUPLES
3 NOT SHOWN,

 

 
 
  

THERMAL
~"INSULATION .

  
 
 
 

COOLING GAS
/ SHROUD -

       
   

   

1
1
}

-,
-
-

 THERMOCOUPLES

- fl

e e e e
s :

r

!

 
 
 
 

7.

  
  

/
/

M

  
 
  

 

CERAMIC

  
 
   
 

| ) ' |sTAflcE
ELECTRIC RES
coot'x\‘.iflrms MEATING ELEMENTS
COOLING GAS
OUTLET

-~ THERMOCOU

Fig. 20.2 Typical Freeze Valve

1

ORNL-DWG, 64-6899 -

E£4€

 

 
 

 

 

344

were connected to the controi modules and to the cooling air‘modfilating &fij
controller. A single thermocouple spaced 5 in. from the valve centerlifie,
upstream and downstream on the top side of the 1-1/2-in. pipe, supplied
information for setting the manual heater controls. The heaters on either
side of the-valve were individually controlled. |

The thermocouples at the valve body were connected to an Electra Sys-
tems modular control network (31x modules per- valve) which operated auto-
maplcally at preset temperstures to maintain each valve within its speci- -
fied condition. The control network of the critical valves, descfibed
be;ow; included an automatic cooling air flow controller which maintained
the valve center at a constant temperature. The ultimate contfol, however,

remained with the modules.
20.3j'Descri§tiofi’of the Operation of the Freeze Valves

Valves were usually operated in one of three steady-state conditions

or modes. . ,
(1) Deep—frozen — the heaters were adjusted to maintain the valve at (i)
i | . 400—500°F without cooling air. R
(2) Thawed . — The heaters were adjusted to maintain the valve at.
n1200°F without cooling air.
(3) Frozen — The heaters remained as in the thawed condition but

. the cooling gas flow was sutomatically adjusted to
| hold the frozen valve in condition for & rapid thaw.

Operation of the valve required only a hand switch (or control action)
to freeze or thaw the salt. When the valve was switched to "thaw" from the
"frozen" condition, the air was cut off.and the heat supplied by the shoul-
der headers would melt the salt in the valve body. Similarly, when the
vaelve was switched to the "frozen" from the "thawed" condition, the air was
turned on and the salt in the_valve body would freeze.

Each valve had three coolant gas flow conditions: off, full on (blast),
or reduced flow.Khold).V[The flow condition was determined by the tempera-
ture control modules. Aifihough the exact module setpoints varied, for the
purpose of explanation, the following conditions are used: (See 20.4.4 for
FV-103 control.) o . Qfij
 

 

 

345

Module No. Setting = | ~ Function

FV-1A1 ~ 850°F+, 800°F+ Shoulder temperature range

FV-1A2 ~ TOO°F+ ) Low shoulder temperature

FV-2A1 500°F '~ Center temperature -- low alarm
FV-2A2 ~ 1300°F+. Center temperature —- high alarm
- FV-3A1 | 850°F+, 800°F+ ~ Shoulder temperature range

FV-3A2 TOO°F+ Low shoulder temperature

Assume that the wvalve was thawed at 1200°F. When it was switched to freeze,
the blast air (&EM scfm) came on. When both valve shoulder temperatures
decreased to 800°F, the "plast" air was cut off and the "hold" air auto-
matically came on and held the valve shoulder temperatures between 800°F
and 850°F, If the shoulder temperatures continued downward for any reeson,
all air was cut off at TOO°F by modules 1A2 or 3A2, If hold air flow was
‘insufficient and either of the shoulder temperatures exceeded 850°F, the
blast air automatically switched on by the action of the 1Al or 3A1 module
and reduced the temperature to 800°F.
~ On those valves used most frequently, FV-103, 105, 106, 204, and 206,
a temperature controlled dlfferential air flow adjustment was prov1ded for
the hold air in addition to the overrldlng on-off module control. These
control;ers maintained the valve center at a constant temperature.
The normal valve'condition during reactor operation with fuel salt was
as follows: (See Fig. 20. l )
(l) FV-103 was frozen but in the event of an emergency drain, it was
requlred to. thaw in less than 15 minutes. ,
| (2) FV-105 and FV—106 were thawed and were required to remain thawed
77 even in the event of a power outage.’ In the event of an emergency fuel
.draln, the salt flowed to both fuel drain tanks through FV-105 and FV-106,
| A normal draln required free21ng one of ‘these valves prlor to the drain to
.dlrect the salt to the selected recelver tank
(3) Fv-104, and FV's 107 to 112 were deep—frozen.
(4) Fv-20L and FV-206 were frozen during coolant salt c1rculat10n.’
These valves were requlred to thaw rapidly to prevent the coolant salt from

freezing in the radiator tubes.

 
 

 

346

~During flush salt'operation, FV-103 was frozen end FV-th was thawed.
A11 other valves were deep-frozen. '

FV's 107 to 112 were used only during shutdown perlods for salt trans-
fers and additions. The thaw and freeze time for these valves was-not im-
portant' When not in use, they were frozen or deep—frozen dependlng on the

operatlon in- -progress.,

20.&_ Operating Experience

After the inltlal difficulties described below, operation of the
,freeze valves was quite satisfactory. Table 20.1 is a tdbulatlop of the

freeze and thaw cycles for each of the valves.s

Teble 20.1 Freeze Valve Freeze-Thaw Cycles

 

Freeze Velve No. o No. of Cycles
103 o 21
10k | - L6
105 . | 97

106 T8
107 - : - 36
108 - | L5

- 109 o 53
110 1k
112 o - 3
204k : o 5T
206 - )

The valves which were requlred to melt rapldly, less than 15 mlnutes,
were timed whenever the system.was dralned These valves were FV-103, FV-
20h and FV-206. Periodically they were tested under complete power fail-
ure condltions to assure that the valves would melt w1thin the maximum time
alloved. The normal thew time without power for FV-103 was 9—11 min, for
FV-204 and FV-206 the normsl thew time was 12—1k minutes.
347

During early operation, FV-10k, FV-105, and FV-106, which were re-
quired to remain thawed (>850°F) for ~30 minutes, were also tested under
- the same power failure conditions. At an initial temperature of 1150°F,
the shortest period of time required to reach the salt freezing point was
367minutes, with'nOJsalt"flowing through the pipe.

_ It required approximately 8 hours to bring a valve out of the deep-
frozen condition (LOO—500°F) into the frozen condition. Heat was applied
selectiveiy during the heatup-to insure that the salt was melted progres-
sively from either end towards the center of the freeze valve. This-was
to avoid having molten salt trapped between two frozen plugs and possible
danger of pipe rupture. The siphon pots were interlocked so that the pot--
temperature was requlred to be greater than 950°F before the assoc1ated
valve could be thawed.

20.4,1 Testing of a Frozen Valve

After a valve had been frozen, it Was—pressure tested to assure that
a solid, leaktight frozen plug had been obtained.' Early testing revealed
that due to the-fuel drain line geometry, it was possible to have insuf-
ficient salt remaining in the freeZe.valve sfter a drain to make up a solid
plug. ‘ o

, Investlgatlon, using a. glass pipe model demonstrated that after the
draln line had been blown down through the head of salt in the drain tank,
~ the salt remained in the siphon pot and would not flow back into the valve
body to fill the void. After blowing down line 103, a 4 or 5 psi AP exist-
ed between the drain tank: and reactor when attempts were: made to equalize
rthls AP, the salt was forced out of the pot and past the freeze valve back
into the 103 line._rf S R | | |

- The method which was- used ‘most successfully was to reduce the reactor
system pressure by ml/2 psig: which usually allowed sufficient flqw-back tor
- £i11 the valve and make a good seal A similar. procedure worked satisfac-
torlly for the transfer freeze valves., '

20. 4.2 Control Modules ' j |
During the initial checkout of freeze valve control c1rcu1ts and gub-
'sequent operatlons and tests, problems were encountered 1n the operatlon

of the modules which make up the Electra Systems Corporation alarm

 
 

 

—}

348

monitor system. The setpoints drifted off the pre-set temperatfirettrip
value, double trip points occurred, and failures to respond to alarm'sig-
nals were encountered. | ' |
Failures and malfunctions ef-the modules were traced to.inferior qual-
ityfiComponents and corrosion or oxidation at the printed circuit board con-
tacts. These faults were mostly eliminated by replacement of faulty com-
ponents, gold-plating all module contacts, and minor circuit changes. The
setpoints for the T2 modules were checked at least once per year there-:

after, In general, the modules remained within 15% of the preset values

- after theuipitial difficulties were :resolved.
- 20.4.3° "Heat Control to Freeze Valves

~ As installed, the heaters on either side of the freeze sections were
on & common control. Also the heat supplied to the lower section of the
siphdnipots adjacent to the freeze'valve heaters was marginal due to the
manner of installation.. Therefore, a balanced temperature gradient on ei-
ther side of the freeze plug was not obtained. Since the setpoints for:

the shoulder temperature control modules were adjusted for a balanced heat

,distrlbution, it was difficult to maintein the valves within the module

limits. Separate heater controls were added which permitted separate con-

.trol of the heat to each side of the freeze plug: The heater. wattage of
.the shoulder heaters was increased from 15 W/in2 to 30 W/in?. These added

- features relieved most of the gbove difficulties.

~ -The proximity of FV-105 to FV-106 (35 inches apart) closely related
the temperature effects of one valve to the other beyond the limits of good

.eontrol, This could only be corrected by a change in the piping. Since

this was not practical, the problem existed throughout the reactor opera-

_tien. This was especially true during a salt fill from FD-1 or FD-2 since
~ the hot salt in line 103 was 19-3/k in. from FV-106 or 15-1/L in. from -
_FV-105. ' :

20.k, h"Comments'on Operation of Individual Freeze Valves -

FV-103 ——-Slnce this was the main fuel salt drain freeze.valve, it
was required.to thaw in less than 15 minutes. A coolant gas flow rate of
75 scfm was .available for freezing. Heat was supplied to the valve from
the aflblent temperature within the reactor furnace and therefore no valve

heaters were required. Operation of this valve was complicated by & dual
349

set of operating conditions: (1) Immediately following a reactor fill, the
line-103 on the downstream side of the valve remained full of salt for a

period of several hoursj (2) Line-103 was then emptied of salt leaving no

-salt on the:drain tank side.’ Figure 20.3 shows. the temperature distribug

_tlon for both condltlons.

In order to meet the fast - thaw requirement and still meet both above

conditions, the module:control_p01nts were set to function as listed below:

FV-103-1A1 -- Blast air on at_lOlO°F+ no hystereszs — (shoulder

temperature. )

FV-103-1A2 —- Low. temperature air cut off — U65°F+ — (shoulder

temperature)
FV-103-2A2 - High alarm 600°F+ "2 of 3" alarm matrix (center
' temperature) ,
FV-103-3A1 -- Blast air on at T65°F4 - T19°F+, 50°F hysterisis,
(shoulder temperature)
FVf103-3A2 -~ Low temperature air cut off — 515°F¢, (shoulder

temperature).

The "hold air" dlfferentlal air flow controller was normally set to

~hold the center temperature (TE-FV-103-2A1) at 375 to LOOCF.,

The temperature adjustments required for the proper valve operation

'resulted in only one shoulder control module being effective in each con-

dition of the valve shown in the figure. During the initial freeze cycle,
the TS-FV-103-3Al1 was the principal control and after the 11ne-103 was
drained of salt, the TS—FV—103-1A1 was-the key temperature control since
there was no salt below TS—FV—103-3A1. During a thaw cycle, the 1A1 tem-

.....

on that’ side of the valve. The 3Al temperature responded very rapidly

_fs1nce the source of heat was through the frozen plug and- the amblent tem-

perature. Selectlon of these temperature alarm p01nts was carefully made

.. after many tests.

Once the valve conditlons ‘were establlshed FV-103 operated well ex-

:'cept for one unscheduled drain which occurred on April 15, 1969. This was
~caused by an. Upward 50°F drift of the setp01nt of the FV-103-3A1 module
from T50°F to 810°F plus an administrative change of the FV-103-1Al set-

point from 1010°F to 1050°F,

 
 

. ORNL-DWG 73-806

Fv103)\ (FV1C

2A1 J\ 2A2 )Fv103\/Fvi

A\ 1A1 M\ 1A2
FV103Y FV103 T - ‘
3A1 N\ 3a2 A\ g , - ’
1 | - . .~ REACTOR ;

 

 

 

 

 

 

 

 

 

 

TEMPERATURE (°F)

 

1200

g 8 8 §

g

 

 

 

 

 

 

  
 

O.NORMAL FREEZE, NO SALT IN LINE .
103 ON DRAIN TANK SIDE — 9—11 min MELT

® VALVE F.ROZEN, SALT BOTH SIDES
OF FREEZE PLUG .

 

 

B-1 A1B B 28 1B j | R27 — 30
TE NUMBERS. | | | "

© Flg. 20.3 "Tefiperature Distribution at FV-103

0st

 
351

The upward drift had occurred prior to the previous freezing operation
and a short plug (<2 in.),of salt was formed in the valve. The - controls
and freeze test indicatedea good freeze., At the time of the incident, the
salt temperature was not at_1210°F for which the controls were originally
adjusted, but at 1110°F with the cooling gas flow controlling the valve
center temperature at 420°F,

Figure 20,4 is a graphicflhistory of the incident and it can be seen
that, with the FV-lOS-lAl-adjfisted to 1050°F, there was little if any warn-
ing to the operator prior to the melt. The valve normally melted when
TE~FV-103-1A1 was sbout 1030°F—10k0°F when the fuel salt was at 121Q9F at
the reactor outlet. The lAlefiodule'administrative'change to 1050°F-was
based on an untested theory that the valve should be adjusted on the re-
actor inlet salt'temperature'rather than the outlet. As long as the freeze
valve had a normal 2-in, freeze plug, the cooling air could be adjusted to
maintain the valve within the normal operating range. In this 1nstance,
Vdue'to short plug, the_valvermelted.as-a result of changing the cooling
air flow which initiated a heating cyoiéflbeyond the range of the cooling
air to control. The air flow #as'routinely.adjusted_50?F opwefdewhen the
ffiel salt was at a redueed tempefefuie butethe combination of raising the
lAl alarm point, the short plug resultlng from the 3Al upward drift before
the previous freeze, and raising the cooling setpoint resulted in the fuel
draln.”

FV-th — This valve operated without dlfflculty since normally it
.was deep-frozen or thawed., It was suff1c1ent1y spaced away from.the line-
- 103 so as to be unaffected by a fuel drein. See Fig. 20.1.
 'FV's 105 and 106 — FV-105 and FV-106 were difficult to control during
e'alfill'or drain operation because of the proximity of the valves to each

‘other and to the 103 line which was common to both freeze valves. See
Fig. 20.1.  Heat conduction from the hot salt passing through line-103
strongly affected that valve which was to remsin frozen. In an emergency
drain situatlon, both" valves were thawed and the fuel would drain to both
tanks. A scheduled drain requlred that one of these valves be frozen to
| direct the salt to the selected‘drain-tank only., In several instances
during a scheduled reactor drain ihe frozen valve thawed whieh resulted

 
 

TEMPERATURE (OF)

352

ORNL~-DWG 73-607

  

SALT IN REACTOR VESSEL AT 1?10°F

 

 

 

 

 

 

  
  

 

 

1100]—
TE FV 103—1A1 -
1000}=
800]—
800f—
J00l— : | © SIMULTANEOUS ALARM AND
THAW—-SALT FLOWING THROUGH
_ 1/2in. EMERGENCY DRAIN TUBE
600k ¢y - |
103-3A1 - VALVE CENTER TEMPERATURE RISING, COOLANT GAS
? FLOW INCREASE NOT FAST ENOUGH TO CONTAIN HEAT—UP
L i , Lo ER Sttt -
103-2A1 RESET CONTROL POINT

 

TIME

Fig. 20.4 Teinperatures During Unplanned Thaw of FV-103
 

 

353

in an added salt transfer oPeration to put all the salt into'the proper
tank., The condition was most troublesome during a reactor fill operation,
ia time—consumlng functlon at best, when the frozen valve would go into
‘alarm and control interlocks would stop the fill or the freeze-valve thaw
with the system partlally fllled and drain the salt to both storage tanks,
| By proceeding very carefully during the initial stages of the fill this
condition could, and usually_was, overcome by stopping the fill with the
103 line filled to above theftee'and allOwing the frozen valve to approach
~equilibrium. However, the last failure of this sort occurred as late as
April 11, 1969. 7 _ - | '

A leak in the primary'syStem.became evident after the final-system
drain of December 12, 1969 which appears to be in the viecinity of FV-105.
This is discussed in Section 5

FV's 107 through 112 ——-There were no difflcultles or unscheduled :
_thaws with these transfer valves.

FV's 204 end 206 ~— The coolant valves have the same 1nherent fault
as FV-105 and FV-106 of being too close to each other. Since the valves
were always operated as & pair, either frozen or thawed this did not cause
any difficulty. There were no operating difficulties with these valves
after criticality. N

' 20;5' Recommendations

1. Freeze valves of thlS type should be so located‘by distance or
:separate lines :so as. to avoid any temperature 1nteraction..p",
, 2. They should be BO- positioned that the adJacent piping is p051—
":tively full of salt, The valves can. be installed in & variety of ways,
='vertically, inclined etc., to insure there are no v01d spaces in the piping.
3. - Those valves which. are not eritical should be simplified by elimi~

.'F;rnating the automatic modular control feature.- These valves would be less

 

'“expen81ve, less time-consuming to adJust and easier: to control by simple
 manual air control._’l'f' LR ST

) 4, The. automatic differential cooling gas controller should.be con-
sidered for all critical valves, This feature greatly reduced the amount

 
 

 

 

 

354

of time required-to.maintainlthervalveS'within'limitsAand_reduced the
post-freeze thermal  cycling. | , - ,

5. The shakedown period for adjustment of the valve temperature con-
trols at the reactor was time-consuming for a variety of reasons. Initially
the control modules were found to. drift off setpoint and hed double trip
points. These difficulties were essentially, but not completely, elimi--
nated by local redesign. There were six control mbdules_per-valve for a
totai of seventy-two individual modules to be maintained. Each module con-
tains two indicating lemps, "Alarm" and "normalq vhich are wired in series
with the control-circuit,} There were frequent failures of these bulbs
which in turn affected -the module fimction° ‘Many of.the operators had dif-
flculty in understanding the functlons of the modules and were at a loss
as to what corrective action was to be taken when an alarm sounded. Al-
though the Electra modular system succeeded in controlllng the valves, a
simpler system is recommended, .

6. Individual control should be prOV1dedfor each of the heaters which
affect the valve temperatures. An excess of hesat is;needed to.overcome heat
losses in the valv?e;area° PlaCement'df‘the heaters shouldvbe'given careful
consideration to eliminate cold spots in adjacent piping. |

T. An,adequate,supply of -cooling gas should be availsble to each
valve. A supply of L0—50 scfm for 1-1/2-in. pipe "blast air" cooling
would be sufficient for good control. »

8. The modular control setppints were established for normal 1210°F
reactor operating temperature at steady state. The critical "fast thaw"
valve control setpoints were adjusted for-this temperature as vas the -con-
trol point for the automatic cooling_gasflcontrol. Deviation -from ‘the nor-
mal temperature for which the control points were adjusted (within 75°F)
required sdjustment of the cooling air flow only. This reised or lowered
. the valve center temperature to maintain the shoulder temperature control
modules within limits. During transient conditions, such as heatup of the
system, the valves were severely cycled because of a large'tempersture
gradient between two valve shoulders_onrthe seme valve, One shoulder would
be sbove the module alarm point and that control would turn on the blast
eir. The other shoulder would be below the low temperature alarm point -
“and cut.all the ceoling air off. The module controls would allow the

O
355

‘velve to cycle back and forth across this range until the low tempe?ature
shoulder would heat enough (QTOOpF) to permit the differentisl controller
to control the air flow. In fifaétice'all_air was manually turned off
during heatup until the valve shoulder temperatures were above TOO°F. The
valve then would remain in the ffozen condition but heat much more rapidly
- and- the thermal cycling would be greatly reduced.

 
 

 

 

356

21. FREEZE FLANGES
R. H, Guymon

21.1 ‘Description

Mechanicel-type - joints were provided in the S—in, selt piping inside
the reactor cell to permit major components to be removed for maintenance.
Figure 21.1 is a sectional viéw.of-thé so-called "fréezeiflange" type of
joint used. The design was such that the "O" ring was in a relatively cool
location. A frozen salt seal protected the ring joint séating surfaces
from contact with salt which could cause corrosion when the salt wes ex-
posed to moisture. Each freeze flange had six thermocouples; four were
located on & U4-1/2-in. radius, 90° apart, one on a T-in. radius and one

on a 10.2-in." redius.

21.2 Operstion

It was never necessary to.remove a major component and therefore one
important function of the fieezé'flangés was never tested at the MSRE,
Earlier tests on a prototypé flangé indicated that making and breaking
these could be done remotely without appreciable difficulty.

The leak rates of the five freeze flanges dfiring eafly operation are
given in Teble 21.1. Similar leak rates were observed throughout operation
except that one of the freeze flanges (FF-201) 1ocated:near the point where
the coolant salt left the heat exchanger lesked more than fihe allowable
leak rate;of 1 x 10~3 cm3/sec when'salt-was not ‘in the piping and the cool-
ant system had been cooled down., This leak was measured at about 0.2 :'L
cm3/sec on August 2, 1966. The flange alweys became acceptably sealed when
the system was heated. <. o

Although there was some variation between individual flanges and some
variation with time, the thermocouples located 4-1/2 in. from the centerline
of the pipe were usually between 750 and 950°F. Those on T-in. radii‘were
sbout 550 to 650°F and those at 10.2-in. were sbout 450 to 550°F. |

The thermocycle history is shown in Table 21.2. When it was decided
to extend the operation of the MSRE iOnger.than the original plans, the
 

— - e . Sw— - = - - - - -

357

ORNL-LR-DWG €3248R2

 
 

FLANG
CLAMP GAP WIDTH

 
    
 
  
   

BUFFER
CONNECTION
(SHOWN ROTATED)

' MODIFIED R-68
RING GASKET

FROZEN #%e-in. R
SALT SEAL

      
 

" '5-in. SCHED-40 PIPE

 
  

 

 

 

 

 

 

 
 

 

 

 

 

358

Table 21.1 Observed Leak Rates of Buffer Gas from Fréeze Flanges

 

LéakqgeéRa%ef(sdd emd/gec)
" Y a

 

SystemHot -~
o .- - Drained | System Cold System Cold
Freeze - Initial Circuleting After After After
Flenge = Heatup Selt .- Runl ~Run.1 Run 3
: K i ,

1 x10-3  x 10-3 x 1073 x 10~3 x 1073
100 2,0 . 0.5T 0.7 1.5 2.0
101 1.3 - 0.l 0.25 2,26 1.2
102 . 055 . 0‘3 0.30 . .2033 - 100
200 1.0 0.21 - 0.41 1.48 © 0.3
201 0.6 0.22 - 0.21 - 1.23 - 1.8

 

freeze flange test lodfi whibh_had cycled & prototype flange for 103 cycles
vwas restarted. After cycle'268, examination»using dye-penetrant revealed
e crack in the bore in the‘viéinity,of the weld attaching the alignment
stub to the face of the flange. The crack appeared about the same after

the final thermal cycle, number 5L0.

21.3 Conclusions.

. o 1 . i ‘
The freeze flenges performed satisfactorily throughout the reactor

operstion. The higher lesk rate on FF-201.yhen‘cold caused some concern

but did not require repairs.
 

359

Table 21.2. Summary,of.Unséheduled Scrams at MSRE with Reactor Critical®

 

 

 

| Operating Hours V'Number of Unscheduled Rod Scrams
' . Fuel in ., , ., Human Power b
| Ygar Qua::er_ Core . Critical Total Error Failures I&C Opher |

1966 1 6712 62 4 2 0 11
| 2 - 1203 1070 13 2 3 6 2
3 554 - 413 2 0 2 0 0

4 1266 1221 3. 1 1 1 0

1967 1 1861 1852 2 1 0 1 0
2 1254 1186 = 2 11 0 0

3 1318 1292 1 0 1 0o 0

4 2159 2u4. 2 0 1 1 0
1968 1 2048 2045 O 0 0 0 0
2 o 0 0 0 0 0 0

3 88 0 0 0 0 0o 0

4 1000 735 1 1 0 0 0

1969 1 1850 1800 « 2 0 0 0 0
2 1385 1375 3 0 1 0 2

3 1076 1054 . 2 0 0 0o 2

4 1203 1176 0 0 0 0 g

Total = 19027 17425 37 8 10 10 9

 

%rhere 1s no record of ahy'unscheduled scrams during 1965, when fuel
~was in the core for 1062 hr and ‘the reactor was critical (at 1 kW or less)
for 230 hr.

bestly equipment faults.. For example, five of the last six scrams

~ due to "other" causes occurred when the speed of the variable-frequency

" generator- being used temporarily to drive the fuel pump sagged below a
: prescribed limit. Co

 
 

 

 

360

22, CONTAINMENT
P. H. Harley

22,1 Description and Criteria

" Containment at fhe MSRE was required to be edequate to prevent the
escape of multicurieramounts of radibacfivity or dangerous amounts of other
materials to the atmosphere. A minimum of two barriers was provided.

During operatlon the prlmary barrier vas the piping and equlpment
which contained or was connected to the fuel salt system and off-gas sys-
tem. Block valves were installed in all the helium supply (cover gas)
lines to prevent backup of act1v1ty. Primary containment extended out the
7 off-gas lines and through the charcoal beds. The entire primary system
was of all-welded constructlon with leak-detected flanged Joints except at
a few less vulnerable locations where sutoclave fittings were used., Es-
sentially zero leakage was permissible from the primary system°

The secondary containment currounded. the primary containment. It con-
sisted malnly of the reactor and drain-tank cells and appendages to them,
There are NTOO penetratlons for service lines into the cells., Flgure 22,1
is a simplified diagram showing the various type penetrations and methods
of sealing or blocking these. - .

Under conditions of the MCA, the secondary containment would be limit-
-ed to & pressure of 39 psig by the vapor condensing (see 22.4), The cell
temperature would rise to 260°F. Under these conditions the maximum al-
lowable 1eekage rate as specified in the MSRE Safety Analysis Report®S was
1% per day of the secondary containment volume. ) ;

- During malntenance, the reactor was shut down and the amount of radio-
activity availeble for the chances of its release were COnsiderably'reducedv
When the primary system was opened, the cells became the pfimary'eontainment

with the high-bay area acting as secondary containment.
22.2 Methods Used to Assure Adequate Containment and Results

Most of the time the primary and secondary containment was operated
well below its design capabilities. Various means were used to assure that

containment would be adequate -under the worst conditions. -
e e i e et e i R e A

       

«c C

 

 
 
    

ORNL DWG. 672688

. E,‘" s s P et
- .. e - o ‘ . + S . oA
SIEMBZANE SEALTTT . < . : ’ o

 

 

 

 

 

 

 

 

P Yol STBHIELD  BLOCKS .
. . . e — | ] . .
- . | . R I\ " . “/ .‘ - T - l. ‘
@ | ’ ag coNTaoy [T | T Sae suery
' : o ‘ N : SAEETY VA. ,
: AR COMPRESSION ‘

 

   
 

   

QU\ cw
o\sco\.:uthé

\\IA"E.& WATER
TO STACK

= 4, areR S o e
o S qob‘rqenr.—c 2% . : : ‘ . . - -
. s ; : L ) DAFETY v, .

X
' @FE’TY VA.

    
 

EITTINGS

 

 

 

  
  
 

 

 

 

 

 

 

 

 

 

 

 

 

    

 

C 'E:Lf:CTgI\_A
‘ NS e WELILT — podew,  \ T Cam. & ‘
L -.E\..-...P' -3.9‘5..\(2 -: . . u Ak 4 . : ‘
- L - : | TETAL -TO -cmm\ ' ' TSEaLinG COMPONG. SeAL
S %Ar-r.-"v Vi, WEDS . : =eaL o ‘ Ry BEAn
\.»"‘:"EALEO CONTAINMENT [ . . \ . 1 PRESSURLED WY " "
. WoLOSURE ‘ B ' oo TeoLoE copptl, o
FLORRRE (METAL TO GLASS 2 Opota, =3
: . : ' - revy va: S = VA,
. - TUERMOCOUPLES SARETY ,  SATETY
WRLOS : ‘ oL EVALOATION 6 Soace
‘ L | _
r )
ruEL !
o1 PUTE C ' conpoufh GASLET
Pu:\:u‘-’ ) , Sums n t
( £
TASETY VAL ‘ VELTS - ]
o BNTICATED . | TAINTENANCE o S
o e R -Tp ENCLOSLEE- . B el IR = )
i ‘ ‘

80" BUTTELELY VA
VUSRBES, SEALS

 

 

 

 

T VAPCD,  CONDENSING
- SysTEm

 

 

Fig. 22 1 Schematic of MSRE Secondary Containment Showing Typical Penetration
. Seals and Closures '

T19¢

 

 
 

 

362

The primary system was pressure-tested annually and the cell air ac-
tivity was continuously monitored as an indication that no leak had devel-
oped during operation. The secondary system leask rate was checked annually
at a positive pressure and was continuously monitored at normal cell pres-
sure (—2 psig) during operation. In addition to this, all primary and
secondary system block valves were tested ahnually to‘dssure acceptable
lesk rates. .

22,2.1 Primary System

'Radiographs; dye checks, and other inspections weré carefully reviewed
during’cbnstructibn of the primary system. All flanges had metal "0" rings

.which‘were remotely 1eak4detected‘by pressurizing the ring groove to 100
psig with helium, (See_Section'lh«Of'thiS'report.) Although the system
was opened 33 timés.for fuel additions, graphite sampling, and maintenance
as indicated in Table 22.1, there was no leskage of any flange above the
allowable 10-3 cc/min:while the reactor was operating.

In 1965 before starting nuclear operatidn, all of the block or check
valves were teéted in place or removed and bench-tested. This was part of ‘a;
the containment (primary and secondary) check list, Section LE of the MSRE
'Qperating Procedures.zg The specified maximum leak rate on the primary
system valves was 1 to 2 ce/min at 20 psig. A number of valves in the |
helium cover-gas lines had excessive leskage. This was due to damaged "O"
.rings and/or they had foreign particles in them, usually metal chips from
machining. After cleaning and installing new “O" rings they had no detect-
ablé-leakage. After reinstallation, the lines were pressurizéd-and the
fittings-or welds were checked with a helium leak detector.

These_tests'were repeated annually. The results are given in Table
22.2 along with results from tests of the secondary block valves which are
described later. _

In addition to the gbove, an annual strength test was run on the pri-
mary system, This was normally done by pressurizing the fuel system, in-
cluding drain tanks, to 60 psig with flush Salt-circulating_at normal oper-
ating temperature (1200°F). 'No abnormalities were noted during any of
these tests. |
 

 

363

- Teble 22.1 Opening of MSRE Primary Containment Flanges

 

- No. of

 

 

 

 

 

Location ! Times - Primary Resasons
- Reactor graphite.sampler. -6 Sémples, core inspection
 Off-gas line at fuel pump T Line festriptién
Off—éas-line-in vent house L ‘Valve,zfilter, particle trap
Fuel pump vent line | 2. .fPlugged'valve, modify leak detector
Overflow tank vent 2 Plugged valve, access to a lower
' - line : :
Drain tank No. 2 vent "1 Fuel eddition ,
Drain tank No. 2 access -6 Fuel addition, samples, inspection
Drain tank No. 1 access 1 Inspection
DT to FP equalizer 1 ‘Plugged capillary
'FP upper off-gas line 1 Test oil cateh tank
FP rbtary'elemgnt‘ 1 Remote practice, inspection
FP level reference 1  Restricted after overfill
Total | 33 |
Tablei22.2“ Tests of Block Valves
" Alloweble - No. of " Number Exceeding
L ~Leak Rate =~ Valves Allowable Leak Rate
 Type.Service At 20 psig ~ Tested 1966 1967 1968 1969
. Heltun  lescc/mn. 69 2 3 3 )
 Water  5-10 sce/min 23 1 s w -
 AMrorCell Alr 35 ce/min - 125 5 5 k-

 

 
 

 

364

There were no leaks in the primary system until after the final drain.
At this time a leek was indicated by an increase in cell-air activity.
Subsequent investigation showed that this was in the vicinity of one of.
the drain freeze valves (FV-105). See Section 5 and Reference 25.
22.,2.2 Secondary Containment. |

 The criteria for the secondary containment was not as figid\as for

the primary containment. Small leeks could be tolerated but it wasrstill
necessary to assure that containment was adequate for the worst condltlonso
The methods used are described below. |

Only one Strength test was made on the reactor and dréin-tank cells,
This was in 1962, soon after construction of the cells was‘éomplétedo The
~tops of the cells were closed temporarily by steel membranes and the pene-
tration sleeves were temporarily blanked off. The cells were then hydro-
statically tested separately at 48 psig (measured at the top of the cells),
A leak rate acceptance test wes also made in 1962, The lesk rate was ac-
ceptable, however, none of the penetrations had been installed by this '
time.s6

Priqf to powér operation,;the containment block and check valves were
tested in place or removed snd bench-tested as described in the containment
startup check list (4E of the Operating Procedures).?? The specified maxi-
mum leek rate at the test pressure of 20 psig varied according to the lo-
cation of the valve but uSually was set at 1 to 2 scc/min for helium cover
gas (primary contaifiment),_S to 10 sce/min for other gases and 3 to 5
cé/min for liquids. A considerable number of 1éaks were found and repaired.

Leaks which occurred in the water system were mostly in the check
valves. . These were the result of foreign particles, apparently washed out
-of. the system.and trapped'between the fixed and moving parts of the valves.
Two hard-seated valves An the water system had to be replaced with soft-
seated valves even though the seats were lapped in an effort to get them
to seal. One was a swing check valve in a water 11ne between the surge
tank and the condensate tank. The other was & sprlng-loaded hand—operated
vent valve on top of the surge tank ,

The instrument-air block valves with few exéeftions were found to be

satisfactory. There were nfimerous leaking tube fitfings in the air lines.
 

365

,Several quick disonnects on air lines inside the reactor cell and drain-
tank cell were found to be leaking when checked with leak-detector solu-
tion. These did not:constitute a leak in secondary containment since each

'g'line has & block valve outside'the cell; but a leak here would affect the

cell lesk rate indication when air pressure was on the line to operate the =
valve, | | |

The butterfly valves in the 30-inch line used for ventilating the cells
~ during meintenance nperationsifiere checked by pressurizing between them.
rThe leakage measured by a flowmeter was excessive, and the valves had to
‘be removed from the system to:determine the cause. There was considerable |
‘dirt on the rubber seats, and one was cut. These seats were cleaned and
repaired. It was found that the motor drive units would slip on their
mounting plates by a small amount, thus causing a slight error in the indi-
cated position of the valve. Dowels were installed in the mounting plates
to prevent this. Small-leaks:fiefe:also found around the pins'which fasten
the butterfly tO’the operating shaft - These leaks were repaired with epoxy
resin.

The line from the‘thermaieshield rupture discs to the vapor-condensing
systems had numerous threaded jJoints that leaked badly when pressurized
with nitrogen; 'Each'Joint'Was-bfoken,'the threads were coated with epoxy
-resin, and the joint was remade;'.All Joints were leaktight when rechecked
with leak-detector solution. o ) ’

The cell pressure vas then 1ncreased in 1ncrements and extensive soap-
checking and 1eak-hunt1ng were done._ Alternate top blocks were 1nstalled
‘and the cells were pressurlzed to 1l ps1g to leak~-check all membrane welds
- with leak—detector solution._ No leaks were found in the welds.. With the
cells at 1 p31g, ell penetrations pipe 301nts tube fittlngs, and mineral-
j'iinsula.ted (MI) electrical cable sealsrsubjected to this pressure were
[checked'W1th leak-detector solution. Numerous leaks were found in tube
fittings and MI cable seals, Manzflof the leaks were stopped by simply
tightening the threaded. parts of the seal However, the lesk rate vas
still about h500 ft3/day, indicating some major lesk that had not been
found.

 
366

All shield blocks were 1nstalled, the cells were pressurlzed to 5 psig, \fi;
end leek-hunting. continued. Three large leaks were locsted. One was
through the sleeve that surrounds the fuel off-gas line from the reactor
cell to a ventilated pit in the vent house- the sleeve was: ‘supposed to have |
been welded to the line in the reactor-cell but this had_beenfoverlooked
Therefore, the sleeve vas closed atlthe,vent house, where it was accessible.
Another-leak wvas in an instrument air line to‘an'in-cell valve (HCV-523),
and‘the third was from the vapor-condensing system to the drain-tank steam
domes and out to the north electric service area through a.line that was
temporarily cpen. _All_penetrations,.tube fittings, and MI .cable seals were
again checked with leek-detector solution. Many MI cable seals which_had'

. not leeked at 1 psig were found to be 1eaking, and some .of those which-had
been tightened and sealed at 1 ‘psig now leaked Agein, many.of the-seals
were tightened. As & result, several of the gland nuts split. and soldering
was required. All large leaks vere sealed or greatly reduced, and many of

. the small leaks were stopped, | :

Leak—huntlng and repairs were continued at 10, 20, and 30 psig.j The _
MI-cable seals as & group accounted for & large percentage of the remaining _‘fiJ
lecsks. To stop small-leaks vhich may not have been located, all MI-ceble
seals were coatedjuith epoxy resin at the seals outside of the cells. This
. was done with the cell pressure-at —l.5 psig so thet the epoxy would be-
drawn into the seals. Teflon tape, used extensively on MI-cable seals,
~ other threaded pipe, end tube fittings, did not perform satisfactorily in
providlng a gas seal. _ | ( t l
| Durlng the 30—p31g test large lesaks were located in both component-
coolant-pump-dome flanges using leak-detector solution. (One_of them was
audible.) The inlet and;outlet:valves to these domes were thenclcsed,and_
uthe domes were opened for work on the gaskets. The,broad,_flat,ffiiton—
rubber gaskets were found to be undamaged., The leakage'waS"attributed to
inadequate loading pressure on the ‘broad gaskets, and after they were nar-
rowed, no further leeks were cbserved., L L G

Leak-rate dats was then taken at 30 20, 10 .and -2 psig with the o
component-coolant-pump domes velved back into the system. These are plot-
ted in Flg, 22.2 along with curves which relate the ellowable leaksage at
various pressures to the allowable leskage at: 39 psig (MCA pressure) for £=J

P A
< x‘ ‘‘ ¥ -’f.‘ ._w.“._‘ “, TRl
 

 

 

367

ORNL-DWG €66-4753

 

500

 

400

 

300

 

200

 

100

 

CONTAINMENT LEAKAGE RATE (scfd)

 

 

 

ASSUMED LINEAR
ORIFICE FLOW
TURBULENT FLOW

 

-100

 

 

 

 

e ® O O®O

 

 

 

200

20 o o

. CONTAINMENT PRESSURE (psig)

LAMINAR FLOW 7

EXPERIMENTAL DATA

RELATIONSHIP

20 ~ 30 - 40

- 'F'ig'.”22;'2",'_?_8'é¢6i1&é;ty .' Containm'ent Leak "Ra'iiés'

 
 

 

368

verious flow regimes. Also shown is the highly conservative linear re=-
lationship which has no physical basis.

The allcwable leek rates based on the orifice flow curve (the most
conservetive realistic curve) elong with the measured leak rates during
these first tests in 1965 are given in Table 22.3,

Teble 22.3 Cell Leek Rates

 

 

 

Test: Alloweble . = 7' 0Bserved Lesk Rate
Pressure Leak Rate : , “(sefd) '
~ (psig) (sc-fd)- 1965 1966 1967 1968 1969
- ‘
30 360 130 |
. .
20 290 125 - 35 58 150
* . ..
10 o195 Lo 65 |
¥% ¥ :
~p -5 —20 to =50 =20 to =65 <10 to -3 -15 to —20 —23 to =30

 

%
The instrument air block valves were closed during these tests.

% o

The leak rates at —2 psig are those calculated for the periods fol-
lowing the pressure tests. The instrument air block valves were open during
these periods, |

Subsequent ennual checks of the secondary conteinment consisted of:
(1) testlng and repalrlng ell block valves, (2). checKing the lesk rate at
:some elevated pressure (usually 20 psig), and (3) checking the lesk rate
"while operating with the cells at —2 psig. - The number of valves found
1e&king durlng_these tests are shown in Table 22,2, The measured 1e§k rates
ere given in Table 22.3. o
. During oPeratlon of the reactor, the cells were maintained at -2 psig
an@ tpe leak rate was monitored somewhat continuously. On three occasions,
the reactor was sfiut down due to indicated high cell leak ratés. Further
 

369

investigation showed that leafiage would not have been’ excessive during an
accident on any of these occasions. These are described below.

In the latter part of May 1966, & cell lesk rate of 100 scf/day at
—2 psig was calculated. The reactor was shut down and subsequent investi-
gatlon, including a 20-psig pressure test, disclosed a hlgh leakage rate
through the thermocouple sheaths from the pressurized headers into the cells.
Since no significant leakage was detected from tHe thermocouple headers
outside the reactor cell, a flowmeter was installed in the.nitrogen supply
line. This flow wasfincluded in the lesk-rate calculations. Prior to this
time the headers had been.maintained between 5 and 50 psig;by periodic pres-
snrization. To decrease the pnrge to a minimum, the procedures were changed
to maintain the pressure at 5 psig. A containment block valve was installed
in the supply -line. | ,

In November 1966, the indicated call leak rate increased to 300 scfd
‘at =2 psig. . Therefore, the reactor was drained on November 20 Leaks were
found in two sair supply lines and one vent line used for in-cell alr—operated
valves, One of the supply llnes was capped 51nce 1t supplied & valve which
was only operated perlodlcally. A rotameter was installed in the other
supply line and the vent line was connected to the cell

After measuring a cell leak rate of 45 scf/day durlng a 10-psig pres-
sure test,,Run lorwas started. Early in thy run the new rotameter in the
air line indicated that the ln4cell leak had increased' Rotameters were
1also 1nstalled on three more alr 11nes vhich were found to be leaking in
qhe cell., Later, the known leakage increased to 3500 scfd. Although the
lcalculated cell leak rate appeared to remain at about 50 scfd Run 10 was
termlnated in January 1967 because the p0551ble error in the measured purge
_rates in the leaklng instrument a1r lines exceeded the permlssible leak
.rate.

During the shutdown, the air—llne leaks were traced to qulck dlscon-

_f nects in which neoprene seals had become embrittled. Out of 18 disconnects

 

'W1th elastcmer seals; elght dlsconnects, all near the center of the reactor
cell ‘were 1eaking. Seventeen of these were replaced W1th special adaptors.
sealed at one end by an. alunlnum gasket and at the other by a standard

metal-tubing compression flttlng. (One disconnect was not replaced be~
‘cause it was on a line that is always at cell pressure.) No similar

 
 

370

difficulties were encountered. After sealing the cell, the lesk rate
wa's 50 Scfdo - . o ’ o N

22,3 Discussion of Cell Lesk Rate Determinations -

There were at least three ways of determining the cell lesk rate.
- These were:- (1) a material balance plus compensation for pressure and tem-
perature using changes of the differential pressure'betfieen the cell atmos-
phere and an in-cell reference volume; (2) a materisl balance. plus compen-
sation for absolute temperature-changes in the cells, and'(3) an-oxygen
balance, 7 o

The first method was used for calculating all official leak rates at
the MSRE, A discussion of each method follows., |

Method 1 -- Materiel Balance Plus Compensation Based on the Differential
Between the Cell Pressure and the Reference Volume Pressure

In order to. determlne the cell leak rate in =& reasonable time, very
accurate indication of cell pressure changes was necessary. The instrument (iiJ
used to determine the pressure changes was essentially a "U" tube manometer
‘(Plgo 22.3) called a "hook gage. Pointed shafts attached to'micrometers
| protruded through "o" rlng seals in the bottom of the chamber. . Readings -
taken by adjusting the point of the shaft until it was at the 1iqu1drsur-
face and then reading the mlcrometer. Thus changes in water 1evel corre-
sponding to cell,pressure changes could be accurately measured The or1g1—
nal gage had a range of 2 inches of water. A later model'was dbtaaned in
whlch the reference chamber could be moved a measured amount whlch gave a
range of 12 inches of water.,

The changes in cell pressure were measured relative to that of an in-
cell reference volume. The reference volume was distributed throughout the
- reactor cell and drain-tank cell in an attempt -to compensate for cell tem-
,Aperature changes° It was found that very small leaks in the 11nes to the
) reference volume could give large errors in the leak-rate data. Weldlng
.these lines ellmlnated this difflcult.y°

Approxlmately 5% of the contained volume was located outside of the

reactor and drain-tank cells and_had no temperature compensat1on. Thls f }
 

 

 

371

ORNL DWG. 67-2689

 

TRANSPARENT
CHAMBER -(

 

 

 

X LIQUID
| LEVEL —7

 

 

 

 

 

 

 

 

TEMPERATLRE . - | WOOK  GAGE-
COMPENSATING ' ,
¢ REFERENCE
VOLUME'j

 

 

  

 

 

CONTAINED  SAFETY ™LOCK

VOLLME- L VALVES
(nc.ovc, Lgs0) -

 

 

Fig. 22,3 Schematic of Hook Gage and Piping for Determining Pressure Change
- of Contained Volume Relative to the Reference Volume

 
 

372

included the 30-inch-diameter reactor cell ventilation duct (1line 930) in
the coolant drain tank cell and the component-coolaent-pump domes in.the
special equipment room. These areas were sealed off as well as possible
to minimize temperature changes. At pofier there was still e_oonsiderable
emount of air leskage pest these from the main blowers which caused day-
night temperature variations up to 20°F during winter months., The effect
of these temperature fluctuations was magnified by the presence of water
vapor 1n the cell air. (Since June 1967, there was & small continuous
vater leak (<1 gal/day) into the reactor cell. ) This water evaporated in
the cells and condensed in the cooler sections of the containment (1ine 930
and the component-coolant—pump gas cooler) This weas periodioally drained
from the system. In colder weather more water would condense, |

Reactor power also had an effect on the indicated cell leak rate due
to cell'temperature:changes. When the .reactor was teken from zero to full
pover, gamma_heeting in the thermal shield and-otfier'equipment in the cell
caused the average cell temperature to increase 4 to 8°F and indicated a
high cell leak rate for approximately 2k hours.v-These changes were not -
adequately compensated for by the pressure referenee rolume.

After the cells were closed following each'in—oell.maintenance period,
the containment was purged with dry N, to remove ox&gen, so-the cell gas
started out dry As the water evaporated, the initial cell leak rate was
usually high (75-130 scf/day hes been calculated) but gradually'decreased
to an equilibrium value in 5 to 7 deys at which time condensate started to
form and was drained from the system. During-this initialtperiod until
condensate appeared, we relied on results of leak-testing at positive pres-
sure before the ¢ell was evacuated and purged. Due to the scatter in the
data and the relatively small change in pressure or temperature that repre-
sents-e*large“leak'rate,* considerable'time was'required to obtain data
from which relisble lesk rate could be determined, To minimize the tem-
perature and other effects; initial and final ideta was usually taken at

i ) . i

 

. . : _

A change in cell pressure of 0.3 incges of water (0.0l psi) per day
or & change in the average temperature of 0. 0h°F will change the indicated
leak rate by about 10 .scfd.
 

373

the same.time of the day-and at the same operating conditions. Calcula-
tions of intervals of several ‘days were'more-consistent than shorter
Vperiods; Figure 22, h'isiasfilotgef the data used«tordetermine the leskage
rate at 30 psig in 1965. - | '

. Typical flow rates for the materlal balences were: 9 scfd to the sump
bubblers, 13 scfd purge for the thermocouple header, and Orto 1000 sefd
evacuation flow., Since the evacuation flow was normally the largest and
was changed_more often, its sighal was sent to the computer which inte-

grated it and each shift typed out a value for the total volume evacuated.

Method 2 -~ Using the Absolute Cell Pressures and Temperatures

This method was not used because the absolute cell pressure indication
was not as accurate as the indication of differential pressure by the hook
gage and compensatlng for temperature changes using absolute temperatures*
did not give as consistent results as those obtained using the reference

- volume.,

' Method 3 -- Using an Oxygen Balance

The - MSRE containment was purged with N, to keep the O3 concentration
<5% primarily toleliminete the danger of an explosion if an oil lesk should
develop in the fuel-pump lubricating system. This made keeping an oxygen
balance on the cell appear as an attractive method of measuring the cell
leak rate when the cell was at a negative pressure.

The precision to vhich we- could read the Ojp analyzer was only *0,1%
‘which is equivalent to m60 ft3 of air in the cell. With a leak rate nor-
‘mally %20 ft3/day, it became apparent that the only accurate calculation -
would be. over very long tlme periods. In comparison to Method 1 (cell lesk
rate using the cell pressure change and a.flow balance), the oxygen balance
leak rate was consistently,lowrby v15 sef/d&y. In fact, when the leak rate
'-caléulated by Method‘l-was'?15.sef/day5 the 05 data indicated a negative
lesksage. | | |

 

% ,
7 Ten ambient thermocouples were located in the reactor cell, 6 in the
drain-tank cell, and 1 in the special equipment room.

 
 

374

<

ORNL- DWG 65-13149

 

 

 

 

 

 

 

 

 

 

2.2
- s ,.;\ - |
,§ oLs g s § e @ g’ \ /.\\;/‘"K;
g DRAN TANK CELL AVERAGE \\b // A /h“5—~
g i o, / \ T : o
¥ /] \J/

21.0 '

89.2
e i 4

. ™ - \ :

fi \-'—-o/ St gy, |
E 88.4 REACTOR CELL AVERAGE
&

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SPECIAL EQUIPMENT ROOM

TEMPERATURE (°F)

 PRESSURE CHANGE (in, H,0}

 

? 14 © B 20" 2 24 o2 04 06 08 0 ?2 4 6

 

s —ee ({27 =65 —tnj g - 1-28-65 “——'———. s - - -!_

Fig. 22.4 Cohtainment‘Conditidns During Leak Rate Test étVSO psig
 

375

. Errors involved in the pressure method of calculating the leak rate
would probabiy be associated'fiith the rotameters; Eutlthis would imply a
25% error in rotameter calibration. All of the rotameters involved were
disconnected and bench-callbrated at least twice. They were found to be
accurate to well within 5%.

The containment cell was nominally at —2 psig, but Just;downstream.of
the component-cooling pumps, the pressure was +6 psig. It wes thought that
if gas were leaking out here and into the cell at a proportionately higher
rate, we might account for the anomaly. - Simultaneous solution of the leak-
rate equations for this situation showed that although a solution was mathe-
matically possible, the in-leak would have to be at 3% 0, and the out-leak

at’ 20% 0y whlch is not practieal.

These results caused us to thlnk that somethlng in the cell was chemi-

cally consuming some of the oxygen. Removal from the cell air of %2.5

'_ scf/day of pure 0, would account for the leak-rate discrepancy.

One.suggested‘mechanism-for chemically removing'the O, was by oil de-

composition. If some of the oil from the component-cooling pumps were con-

~tacting hot pipe (say at FV-103 or in control rod thimbles) it would prob-

- ably at least partially decompose according to the approximate relation:

/

Cell air samples were takern to determlne the COz content and they showed

'"NO 1% CO». This was at least three times higher than the CO, content in

air, but was less than one would expect to see if all the 02 were being
consumed'by 0il decomp051t10n.,' ' '

The containment cell and support structure was composed -primarily of

"carbon steel so another prime suspected oxygen depletlon mechanlsm.was rust.
“'If all the m1331ng o7} were being consumed to form Fey03, it would take
gbout 120 g/day of Fe. Slnce there 1s Ebout 5000 £t2 of carbon steel sur-

face area in the cell, this would correspond to a corr031on rate of only

0. 5 mils/year (neglectlng the support structure and plplng of component

coolant system).

 
 

 

 

376

This analysis led to the conclusion that enough oxygen was being re- |
moved chemically by oil decomposition and rusting to produce the discrepancy
that existed between the different methods of cell lesk-rate calculation.

22,4 Vapor Condensing System

An accident can.be'conceiied_ip which molten salt and water could
simultaneously leak into the reactor or drain-tank cells. The guantity of
. steam produeed could be such that the pressure in the cells-wbuld increase
-above design pressure. A vspor-condensing sYstem‘was previded to prevent
the steam pressure from rising sbove'the'39—psig design pressure'and to

retain the non-condensable gsses. A 12-in. line connected the cells to the

vapor-condensing system, This line contained two rupture dlSCS in parallel

(a’3-in. disc w1th a bursting pressure of 15 psig and & 10-in. disc Wlth a
burstlng pressure of 20 p51g). The line from the rupture discs went to the
bottom of a 1800-ft3 vertical tank which contained 1200 ft3 of water to
condense the steam., The non-condensable gases went from the top of thls
“tank to & '3900-ft3 retention tank. 1
" The tanks were pressure-tested‘by7the'vendof"to 45 psig. After being
connected to the reactor cell, the system was leak-tested at 30 psig during
initial testing of the reactor and drain-tank cells in 1965.
During the annual cell leak tests at pressure, the vapor-condensing
system was also pressurized. This protected the rupture discs and provided
an integrity test of the vapor-c ondenszng systemn, o
The water used in the vertlcal tank contalned pota351um n1tr1te—)
potessium“bosate‘as a corrosion inhlbltor. Based on annual samples, one
- inhibitos addition was made. There was no increasein_the.iron cpncenfira—
tion which indicated little or no corrosion. . _ o
Inltlally L level prdbes were provided to assure proper water 1evel
,However, after one of the 1ow level probes falled a bubbler—type level
1nstrument was installed for measurlng the level perlodlcally,,'
Slnce there has been no unplanned increases in the pressure of the '

reactor and dra;n-tank cells, the vapor-condensing system hss_pop been usedo
 

 

377

22.5 Recommendations_

Leak~checking the numerous valves required disconnecting many lines,
partlcularly tube fittings and autoclave connections., Disconnecting nu-
merous fittings greatly increases the probability that one or more will o
lesk when connections are remade. For this reason, an effort should be
made in the-design_of the piping to minimize the number of disconnects
necessary for leak-cheCking. For example, nitrogen lines could be con-
nected as shown in- Flg. 22.5 to facllltate checklng both the safety valve
and the line 1n31de the cell.

All block valves should be of high quality and soft-seated. Accessi;
bility to the items to be checked and the points used in checklng,them
should be considered. Flexibility should be_provided:in the lines which
must be disconnected. - : '

Lesks from in-cell sir lines cause errors in 1eak-rate-calculations.
When disconnects are necessary; the effect of flux on materials on con-
struction should be considered. | ‘

MI-cable seals of the type used at the MSRE (brass and stainless
steel) should be modified or another-type used.'debstituting ferrules of
teflon, graphite—impregnatedrasbestos, or other relatively soft material
for the brass ferrules may be sufficient to prevent 1eakage through them,
However, radiation damage and sheath temperature must be con51dered Stan—
dard plpe-threaded connectlons for stalnless steel to stainless steel
joints should be avoided where poss1ble.

Teflon tape should not be used on threaded connectlons for a gas seal

-nor on small lines where fragments cen enter the line and clog it or pre-

vent check valves from seatlng properly. All pipe, tub1ng, and valves
should.be cleaned internally prlor to installation.
MI-sheathed thermocouples are preferable to the type used. However,

__1f cost or other reasons (1. e. seals at the ends) preclude using them, the
initial design should provide forfipressurlzing the terminal headers and
 measuring the leakage.  Soldered Joints in the header system are recom-

- mended where practical,

Gas analyzers used to continuously monitor the cell atmosphere should

be located in a warm area to prevent moisture from condensing in them,

 
by

i
O\

Wall

' //T////// *

7

 

 

i

X

ORNL DWG. 67-2690

CLOwW meTen,

 

 

 

 

 

 

 

SEALED L/ U — / T\ !
- ®mApSS va, —z | | | l - I I : ‘ l
2 [ e e e e
CONTROL
| VAL.VE" =
C e
B J

 

{iwston, mn, SuspLy WEADER

e

SARETY
PLOCK VA,

    

 

 

N, evi.

Fig. 22.5 Schematic of Instrument Airlines with Leak Checking System.

8LE
 

379

Fluids used in manometérs or similar gaées which are continuously
connected to the contained volume such as the hook gage should have a low
vapor pressure to avoid a signifiéant loss during operation. The piping
to the gage should be pitched to prevent accumulation of_water and should
‘not be connected to the bottom of temperatuie compensating or reference
volumes. | ' | |

Tempersture compensating volumes should repmesent the entire contain-
ment., The 30-in. cell vent line and the component-cooling system enclosure

(5% of celi_vnlume) had no reference volume. The temperature reference
volume should be distributed throughout the cells. The MSRE used 6-in.
vertical pipes on a weightedrvblume basis in the reactor and drain-tank
“cells. Smaller'diameter piping with the same total volume might have given
- a more representative temperéture compensation., _

' The reference volume'Systém external to the cells must have a minimum
volume and should be located in an area'held at as constant a temperature
as fiossible. Sufficient cell temperature thermocouples should be located
- in all sections of the cont&inment“to enable calculation of average tem-
perature and temperature changes as.accurately as poséible.

If a sensitive cell préésure.measuring device similar to the hook gage
is used in future reactor containment, it should be designed to meet sec-
ondary containment requirements so that block valves are not required.
However, if block valves are needed to isolate the instrument, a separate
circuit from the other block valves should be used. Pressure readings are
required above the preséfirés,axuwhiCh the other block valves are closed.
Safety jumpers could be usedftb'keep this eircuit opén'duiing cell~pressure

'testing.

 
 

380.

23. BIOLOGICAL SHIELDING AND RADIATION LEVELS

T. L. Hudson

The crlterlon for the MBRE biological shield design was that the dose
rate would not exceed 2.5 mrem/hr during normal operation at any p01nt on
thie shield exterior that is located in an unlimited access area. - Since
the MSRE had to fit wifhin an existing reactor contaigment_cellrand build-
ing, the shield design allowed for addition of shielding as needed to re-.
duce radiation.level at localized_hot spots. The shielding calculatiohs
are reported in Reference 7. _ ;

The reactor vessel was completely surrounded by a water-cooled steel-
‘and-water-filled thermal shield. The thermal shield and fuel circulating
loop were located in the reactor cell. The top of the reactor cell had two
layers of concrete blocks. An annular.space filled with magnetite sand and
water provided the shielding for the sides and bottom. See Fig. 23.1.

When the reactor was not in operation, the fuel was draijned to one or
both drain tanks which were 1ocated in ‘the draln-tank cell. * Magnetite con-
crete walls faced all accessible areas and the top consisted of two layers

of concrete blocks.

23.1 Radiation Sfirveys — Approach to Power

ExtensiVe'health-physies surveys of the reaetor'aree'were performed
during the initial approach to power. These surveys were made to bring to
| light shield inadequacies. With the exception of the areas dlscussed in
the fblloW1ng paragraphs, the shielding was found to be adequate.

23.1.1 Coolant Drain Tank Cell |

At 1 kW the scattéring of fast neutron and gammsa rays from the reactor
cell into the coolant drain tank cell through the 30-in. reactor cell ven-
tilation line 930 caused radiation readings of 8 mrem/hr gamma and 50
mrem/hr fast neutron near the exit of the line. At 25 kW very high readings
were found again at line 930 (70 mrem/hr gamma, 600 mrem/hr fast neutrons
and 30 mrem/hr thermal neutrons). A‘wall-of_16 in. of concrete blocks and
6 in. of borated polyethylene was builad adjacent to the 930 line in the
coolant drain tank cell. In addition, tfle reactor off-gas line 522, in the
 

 

 

 

 

-
HOLD-DOWN

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

~ SECTION "xx"

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

w
® a4
m — | I m :
o 7 . ~
: 7, \\.\h.......l..lol.n..lufi.flfhfl \\
m \ < — .//).0 , .l/ o
oLw&wV\ JJV@ z
|||.|||I..I|u‘\| — ] | O
\ e\\\ . / 14. ,//%u /.r ..HrL....
[ %
off/ \ / o — ¢e \ wo
===l % &
=——5cs N
ofy 9 t
i MO Y = =
& s Lt N E 3
A Bl e 1z
. oN . \\o _ m_
/ : ny//l _ 3N 7 \.\9 : w ,
'\ = ~? 77 1 e
v LS ; y C o=t
/ ° ny( | {J \_J \\o\e \ MMW _ g
,/ ° 4/ \\X A . m
. © /0' ! -\\\l\@\\ue o
Illl.llrr. _ nl\\\\\\\ :

 

 

 

 

ANNULUS-

Fig. 23.1 Shield Block Arrangement at Top of R_eé,c_tor Cell

 
 

 

382

coolant drain tank cell was found to be giving a high background to the
area. At the end of the 25-kW run, the reading was 100 mR/hr at 1 in. from
the 3-in. thick lead shield around the 522 line. Later at higher power
levels it was determined that part of the rediation was from the drain-tank
vent line 561. Therefore this line was shielded with 3 in..of lead. Even

though shielding was added inside the coolant drain tank cell, the radiation

at the door (500 mR/hr gamma,llso mrem/hr fast neutrons, 75 mrem/hr thermal
neutron) and halfway up the acceSSfremp;(22 mR/hr gamma, 3 mrem/hr fast
neutrons, 32 mrem/hr thermal neutron (was hlgh during full-power operation.
This aree ‘was clearly marked'with radiatlon zone signs at the entrance to
the ramp and the door at the bottom of the ramp was locked durlng nuclear
- operation to prevent entry into the cell,
23.1.2 North Electric Service Area o

When the reactor power was raised to 1 MW in April 1966;'the radiation
level in the NorthElectric Service Area (NESA) was found to be high: 20
_mR/hr on the belcony‘and 8000 fiR/hr at the west weii Investigation showed
that there was radioactlve gas in the lines through which helium is added
to the drainitanks. Two check valves in each line prevented the gas from
getting beyond the secondaryucoptainmentrenclosure, but the enclosure, of
1/2-in. steel, provided.little-gemfia shielding. The pressure in the fuel
system at that time was controlled by the newly installed pressure control
valve with rather coarse,trim, and the pressure fluctuated around the con-
trol point (normally § psig) by & ¢2% '»These pressure fiuctuations caused
fission product gases to diffuse more rapidly into the drain tanks and back
through the 1/h-in. 11nes through the shield into the NESA. The radiation
level was lessened by installing a temporary means of supplying an inter-
mittent purge to the gas-addition lines to sweep the fission product gases
back into the drain tenks. During the June shutdown, a permenent purge
system was installed to sfipply a continuous helium purge of TO cc/min to
each of the three gaes-addition lines.. This was proved successful by sub;ft
sequent full-power operastion in which the general background in the NESA
was <1 mR/hr. | )

30 mR/hr gamma was found on the southwest corner ebout 4 ft above the floor.
 

383

A 2-ft by h—ftEby 1-in. thickriead sheet was attached to the wall over the
~ hot spot and s, radiation zone was established.

Vent House -—-During the initial approach to full power, stacked con-
crete blocks were added to,the~floor area of the vent house, over the char-
coal beds and between the_ventfhouse and the reactor building to keep dose
rates low. Very narrow beamsficoning from cracks were shielded with lead
bricks., Even- so, the background radiation level in the vent house was
',NT mR/hr at full power. The vent house was established as & radiation
zone area. . o | f | o o

Water Room ——-Induced act1v1ty in the treated water rose to an unex-
pectedly high level during Run 4. The actlvity proved to be 12, A hr *2x
 produced in the corrosion 1nhibitor. A survey of possible replacements ,
for potassium led to the ch01ce of-lithiumg-highly.enriched in the 713 isoQ-
tope to minimize tritium production. See Water System (Section 12) of this
report for additional- detalls.~~ |

Top of Reactor Cellif—-Itgwas’expected that additional shielding would
be required directly above the'reactor, where there are cracks (¢1/2 in.)
between the'shield'blocks;fffiuring a”fullépower'run:inlJui§'1966,'two very
narrow beams-of'y_with;neutronjradiation'were found between cell blocks for
the first time.-'These“read'np,to‘lo rR/hr gamma and 60 mrem/hr fast neu-

tron. They were properly marked.

23,2 f Radiation Levels During Operation

 

Durlng operation at full power, the rediation levels in all opera-
tional areas were acceptable.“ Some narrow beams were noted from time to
- time. These were malnly in the vent house. Due to. induced activaty in the

i treated water, areas near 1arge equipment such as the surge tank and heat

3fexchanger were treated as radiation zones. Periodic rad1ation surveys have

funot disclosed any appreciable changes in radlation levels.

Some typical radiation readings inside the shieldlng are glven in
ff'Table 23,1, Changes in radiation levels follow1ng a shutdown from power are
shown in Figs. 23.2, 23.3, and 23.5L, ' '

 
 

 

 

Tgble 23.1 Radiation Dose Rates in Various Aress

384

During and Following Full-Power Operation

 

Gamme. Dose Rate (R/hr)

 

 

.

b Reactor'
' - _ Reactor Drained and

LOCATION At T MW 10 kW Drained Flushed

Reactor Cell 7 x10% 5.4 x103 2.4 x 103_ 2 x 103

Drain Tenk Cell 4.2 x 103 4.2 x 103 2.6 x 10 |
Coolant Cell 100
| Fuel-Sampler-Enricherd 1000
500

 

A Off-gas Samplérd

% trer 5-hours operation at 10 kW which followed sustained operation

at full power.

bImmediately following the\ébove 5-hr operation at 10 kW,
CPwo deys after the above fuel drain.
dInside the sampler shielding during sampling.
_ RADIATION (R/hr gamma)

385

ORNL-DWG 73-608

 

REDUCED POWER FROM
105 |— 75MWTO 10KW ,
| O EAST WALL REACTOR CELL — RM 6000—1
5 '® OVER REACTOR VESSEL — RM 6000—2

. 104

108
o ’.--

kDRAIN FUEL SALT TO FD-2

T .0 1 27 -3 45 8 -y | 8 9 10 11. 213
S .: DAYS DECAY ' LT

- Fig. 23.2 Gamma Radiation Level in Reactor Cell

 
 

 

386

ORNL-DWG 73-609

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

105
5
FUEL SALT DRAINED TO FD-2
2
‘©
E
E
g
£
. ]
S ~ L
< ~h
o ~y
< [ ~.
o
fi\-
oy,
ha
5
2
Yy
103

0 1t 2 3 456 7 89 101112131415 1617 18 19 2021 222324 25

DAYS DECAY

Fig. 23.3 Gamma Radiation Level in Drain Tank Cell (Rm 6000-6 between
FD-1 and FD-2) v
 

 

 

387

ORNL-DWG 73-610

EDUCED POWER FROM
309 .5 MW TO 10 KW

2

RADIATION‘ (R/hr 3 ft from coolant pump)
3,

-DRAIN FUEL SALT

100

9 ngUE TO OPERATORS TRAINING SESSIONS
THE POWER VARIED BETWEEN 10 KW AND
100KW |

 

10~1 , — _ ; :
| o 1+ 2 3 4 5 6 7 8 & 10 1M 12 13
| DAYS DECAY

Fig., 23.4 Ganuna-Radiation Level in Coolant Cell (Rm-6010)

 
 

388

23.3 Conclusions

~ The biological shielding wasradéquate_as désigned éxcept for the few
locelized areas discussed previously. Periodic radiation surveys have not

indicated any shift or deterioration in any of the shielding.
389

24,  INSTRUMENTATION.

 dR. H.. Guymon

24,1 Introduction

It is not within the-sccpe of this report to ccverrin detail the per-

- formances of all instrumentation;WEAn attempt has been made to review it

from an operational viewpoint end/report significant items. Information

on instrumentation which'nesfbeen given in previous sections on the perfor-

~mance of the entire plent'cr'individual systems and components will not be

, repeated here. The performénce of the on-site computer is covered in

‘Reference 58,

2h 2 Descrlptlon

Most of the 'MSRE 1nstrumentetion wes of conventional type found in

;other reactors or chemlcal plants. -Empha51s was placed on assuring ade-

quate containment and reduc1ng radiation damage. All c1rcu1try was de-

signed to fail safe. A very brief descrlptlon follows.

_2h 2.1 Nuclear Instruments -

The primary elements of all of the nuclear instruments were 1ocated

~in a 36-in.-dia. water-filled thimble.which extended from the high bay
~ through the reactor cell to the vicinity of the reactor vessel. (See
Fig. 24.1.) e | o

Three uncompensated ion chambers and their &SSOCl&ted fast-trlp com—

parators provided safety instrumentatlon for scremming the control rods.

".(High reactor outlet tempereture also caused e rod scram.) These .were used

~ in a two~out-of-three configuration.s_:'

Two f1ss1on ch&flbers were provided These chembers had ‘automatic po-

:—731t10n1ng dev1ces.r Thelr count rate and “the effect of their position were

'm_ffactored into a 31gnal whlch was recorded on a 10—decade 1oger1thm1c power

"',fdrecorder.: They were used to prov1de rod inhibit, rod reverse, and other

mfiijcontrol interlocks.,

 

 
 

390

ORNL-OWG 64-62%

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

  

 

  
  
 
   
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I 241t 2in. B
-
' & FLANGE FACE
NEUTRON INSTRUMENT
JUNCTION BOX :
HIGH BAY AREA 4«* 50’
- : FLANGE
WATER-SAND ANNULUS et
- T,
; FINISHED FLOOR v B e
; EL. 8521t Oin. \ 1 EL.851# 8in.
: WATER
LEVEL
W 849 ft Bin,
i @Q'
T ‘o
> WATER LEVEL INSTRUMENT
= CHAMBER GUIDE
=+ TUBE {TYPICAL)
RTfiflcTOR R - ——¢ END OF 48-in. 00 SLEEVE
ERMAL ~ EL. 841N 13 in.
SHIELD N %
48-in. 00 OUTER SLEEVE
ASSEMBLY (WITH EXPANSION JOINT)
~~EXP JOINT -
_ 4 CONTAINMENT
PENETRATION EL. 834t 82 in.
REACTOR
VESSEL - C EL.830ft 3in.
\
AN
\msuunou

 

 

 

REACTOR CONTAINMENT

Fig., 24.1 Elevation View of Nuclear Instrument Penetration

 
 

391

- Two compensated ion chambers were used for servo control of the regu-
,latlng rod and for some 1nterlocks. The power was recorded on a linear
recorder. Seven decades were covered.by means of manual ‘range selector
,sw1tches.

- One high sensitivity BF3 chamber was provided for power 1nd1cat10n
during filling of the reactor. |
24.2.2 Process Radiation Instruments.

p Two kinds of detectors, ion chambers and Geiger Mueller tubes, were
used to indicate the level of: act1V1ty in various process streams and to
provide necessary interlocks.

24,2.3 Health Physics Monitoring

| - Gemma radiation wasvmonitored‘by'seven monitrons'located throughout
the building. The air contamlnation was monitored for beta—gamma emitting
-partlcles by seven constant a1r monitors. The building evacuation system
operated when two or. more monitrons or two or more constant- air monitors
from a speclflc group of. 1nstruments detected a high level of radiation or
air contamination.

2k,2.4 Stack Activ:Lty Monitorlng

The containment stack air was checked for beta gamma particulates by
passing a side stream through a fllter paper andrmonrtorrngrthe activity
.. with & GM tube. After'passing through the filter paper, the gaseous sample
passed through a charcoal trap which was monitored by another GM tube to

| detect 1od1ne. A second side stream passed through another filter paper.
'_*ThlS was monltored for alpha partlculates using a thellium-actuated zinc
sulfide screen detector. o '

"2h 2, 5 Temperature Detection o , -
. One thousand seventy—one chromel-alumel thermocouples (some of which

were installed spares) were prov1ded for temperature 1nd1catlon. A1l of

' the salt thermocouples except seven were located on the outside of the 1ines

_p,or equlpment. The seven 1n—thermocouple wells vere located as follows:

- one in the reactor neck one and two spares in the radlator inlet line;

'-3§and “one and two spares in the radlator outlet line, - The more important

'ftemperatures were disPlayed on. S1ngle or multipoint recorders or indicators,
~ Readout of others was accomplished by meens of a scanner system'whlch al-

lowed the signals from up to 100 thermocouples to be sent to a rotating

 
 

 

 

 

392

‘mercury switch and subsequently to an oscilloscope for display. A switch

allowed selection of any one of five groups of 100 signals.

 

24,2.6 Pressure Indicators |
Most of thé remote indicating pressure and dp instruments were pneu-
matic or electric force elements. When containment was required; the vents
were referenced to atmospheric pressure through”roiling disphram seals.
Strain gage and Bourdon pressure gages were used extensively.  Small changes
in reactor cell pressure were determined using a “hook gage". This was es-
sentially a'water'manomEter w@th & micrometer for accurately reading changes
in water level. ‘ | '
24.,2,7 Level and Weight Indicators |
Bubbler type level instrumentsrwere,used‘for measuring the salt levels
in the fuel pump, coolant pump, and overflow tank. A fioat—type level in-
strument was also installed in the coolent pump. The drain tanks were sus-
pended by pneumatic weigh cells to determine the amount of salt ‘that they
contained. In addition to this, two.resistant-type,single;point level
probes were provided in each drain tank.. -
Sight glasses, floats, bubblers, dp cells, etc., were used in the

auxiliary systems.

 

24.2.8 Flow Méasfirément;f

No flowrinstrument wes provided in the fuel salt loop, however & ven-
turi meter was installed in the coolant salt loop. This was a standard
venturi with a NaK-filled dp cell. Flows in auxiliary systems were meas-

‘fi:ed by‘orifices, capillaries pitot tubes, rotameters and matrix type flow

| ;eiéments. |
2h;2.9 Miscellaneous

A semi-continuous mass spectrometer was used to monitor the coolant
~air stack for beryllium. Other instrumentation included the following:
pulse-type speed elements, ammeters, voltmeters, and wattmeters for the
salt pump motors; potentiometer and synchro position indicators for the
- control rods, fission chambers, and radiator doors; oxygen and moisture
analyzers for the helium cover-gas system; and an oxygen analyzer for the

cell atmosphere,
 

 

 

 

393

24,3 Initial’Checkout and Startup Tests

A comprehensive functional checkout of the control, safety, and alarm
instrument was made by Instrument and Controls personnel prior to operation
fof each system, Although some design and wiring errors were found, these
-were-of minor nature_and were easily corrected, In general the quality of
installation was excellent.; The following were included in this checkout:
the setp01nts of all switches were adjusted to the proper values; contin-
uity and resistance checks of all thermocouples were made “the location of
.each thermocouple was determined by heating the thermocouple and measuring
the voltage at the patch panel' and the continuity of all circuits was
, checked and all recorders vere put into operation. Most of this was done
as construction was completed or as the instruments were put into service.

Prior to the first circulation of flush and coolant salts, e complete
',operational check was made of all instruments. scheduled to be used, This
was done follOW1ng the- instrument startup check list.22 This involved:

(l) & complete checkout of all c1rcuits. This was done by changing the
variable or inserting a false signal at the primary.element and assuring
that all circuitry functioned at the proper setp01nts- (2) a complete check
of the patch panel to assure that all thermocouples were connected to the
proper readout instrument and that the operational records were up to date,
(3) a complete test of all standby equipment to assure that it would start
if needed and would function properly; (4) a complete inspection to assure
7_~¥that no switches were inactivated or Jumpers installed; and (5) tests to

"assure that critical equipment fUnctioned properly, such as rod drop times.
o - This instrument startup check list was repeated approximately every
year.- Early- tests revealed errors in wiring, setpoints which had drifted
and other instruments which did not function properly. Later tests 1ndi—
pcated setpoints which needed to be reset and occasionally a malfunctioning
ljinstrument.

- Most of the troubles which were discovered.by doing the instrumentation

",pnstartup check lists were corrected immediately. Therefore_records are in-

’iadequate for statistical analysis."__.-“

 
 

 

394

24l Periodic Testing

 

Due to the importance of some instruments or ciréuits; they were also
fiested periodicelly during operation. As indicated by the sections which
'fbllow,.these tests did not reveal many serious troubles which needed ¢or—
rection. They did provide assurance that the circuits should function if
heeded. In determining the amount and frequency of testing, careful con-
sideration.should be given to the above'weighed against the harm done to
equipment due to repetitive testing and the possibility of interruptibfi of
operation° Rod scrams, load scrams, and reactor'drains occurred at the
MSRE due primarily to testing of equipment or instrumentation. '

The periodic tests made at the MSRE and the results of these testé
~are given below. | |
"2h.4,1 Nuclear Instruments

A complete check of all nuclear instruments was made each month during
nuclear operation. These were done by Instrument and Contfols personnél
using Section 8A of the Operating Procedures.22 All tests were satisféc-
tory except those listed in Table 2h.1. | -

- 2k.4.2 Process Radiation Monitors

The process radiation monitors were tested each week during operation
by inserting = éource near thé primary detector and noting that each con-

trol interlock functioned properly and that all'annunciations_occurred.
All tests were satisfactory_except those listed in Table 24.2. Some early
difficulty was encountered in that the hole in lead shielding'for inéerting
the source was not properly placed.
oh,4.3 Personnel Radistion and Stack Activity Monitoring

Periodic tests of the health‘physics monitors and staék aétivity in-
struments were originally done by MSRE personnel per Operating Procedure
8C.22 fThis was later taken over by other ORNL groups. The briginal séhe—
dule for the health physics monitors was to make a source check‘of-each

" instrument weekly, a matrix check monthly, and a complete evacuation test
semiannually. As confidence in the instruments was established, the fre-

quency of these tests was reduced to monthly, quarterly, and semiannually.
 

 

395

 

 

Table 24.1 Results,cf Periodic Tests of Nuclear Instruments
Date Repairs_Necessary
3/22/66 Replacéddccndenser in the power suppiy of linear power
- channel*Nc.*ls -
7/19/66 Replaced perlod balance module of nuclear safety
channel NQ. 1.
16/16/677' Changed out high voltage supply of nuclear safety
channel No. 3.
T/1k/67 - Repaired scalerrof wide-fangeiCbunting channel No. 1.
8/12/68 Replaced reley of nuclear safety channel No. 3.
1/7/69 | Replaced fast trlp comparator of‘w1de-range counting
' channel No. 2.
2/3/69 Replaced chaflber of wide-range counting channel No. 1.
9/11/69 Replaced scaler of'W1de-range countlng channel No. 2.
10/21/69 Replaced operational amplifler of wide—range counting
channel No, 1.
10/20/69 Replecedéeceler'of wide-range counting channel No. 1.

 

 
 

 

 

 

396

Tableazh.2 Results of Periodic Tests
of Radiation Monitors

 

Troubles Detected

 

Date

4/13/66 Alerm did not occur, RM-596.

5/10/66 Instrument would not calibrate, RM-55T.

1/9/68 Two indicator lights burned out, RM—565,,

1/26/68 Indicator light burned out, RM-675.

7/21/68 Indicator light burned out, RM-82T.

3/3/69 Indicator light burned out, RM-565.
Indicator light burned out, RM-565.

10/10/69

 
 

 

397

The stack monitors were originally tested with a source each week.
Due to their excellent performance record, this was later reduced to
monthly tests. - - ‘ fl

Troubles encountered were repaired immediately.
2k, h.h Safety Circuits

Periodic checks were made of all.circuits and instruments designated
by the Instrument'ano'Cofitrols=group as being safety. These included rod
scram-cirouits,,fuei pump and_overflow.tank'pressures.andzleVels, helium
supply pressures, emergency fuel drain'cireuits, reactor cell pressures,
coolant pump speeds and flows, radiator temperatures and,sampler-enricher
interlocks. These tests involved simulating a failure apd checking as much
of the circuitry as possible without interrupting operations. At.first
these were performedrweeklyijthen ell except the rod scram checks were
changed tora‘monthly basis.flAll-inatruments functioned Satisfaotorily ex-
cept those indicated in;Tablefléh.B.‘

24,5 Performance ‘of the Nuclesr Safety Instrumentation

 

These instruments proved to be very relisble. There were periods when
an sbnormal number of spurious trips occurred. Many'of‘these‘weré believed
to have origInated in faulty, vibration-sensitive relays-in commercial elec-
tronic switches which provided the high temperature trip- 51gnals.r Another
possible source was the chatterlngrof the relays which- change the sensi-
tivity of the flux amplifierelin;thetsafety,cirCuits;'"Correction_of_the e
chattering and- eliminatiofi'6f*noiee'producing components“elsewhere'in the.™
system reduced -the frequency of‘these to & very tolerable: level.:

* The fast trip comparators were found to be inoperative if a sufflci-
ently large signsal was applled to the input. A diode was added to the fast
trip comparator modules to elimlnate this difficulty.

Operation of-the relay matrix in the nuclear system generated con-
siderable noise, maklng it dlfflcult to reset the safety—system channels.
The resistor-dlode combination for damping the voltage 1nduoed by relay
operation’ wes replaced by & Zener diode and a diode comblnatlon ‘that was

more satisfactory.
 

 

 

398

Teble 2k.3 Results of Periodic Tests of Safety Circuits

 

 

Date Troubles Detected
5/2L /65 Noted that PR-522 and PI-522 did not agree.
- 6/9/65 _Two safety channels tripped when testing one chennel, scram
setpoint was at 120% instead of 150% and low current test
_ did not function properly.
| 1/15/66 Fuel pump pressure switch setp01nt needed resetting.
h/10/66 Reactor cell pressure switch setpoint needed resetting.
5/20/66 -Two sampler-enrlcher pressure switches setp01nts needed
resetting.
6/12/66 Sampler-enricher pressure switch needed resetting¢
6/15/66 Rod motion was Jerky.
9/19/66 Two sampler;enricher-access door latches did not function
properly.
11/2/66 A safety chanuel would not reset.
12/16/66 Reactor cell firessure switch setpoint needed resetting.
11/27/67 ‘Burned-out ‘indicator 1ight in safety,ehannel. |
.5/13/69 Defective solenoid coil on reactor cell block  valve header.
9/8/69 . Reactor outlet safety interlock would notclear.
11/24/69 - Safipler-euricher pressure switch setpoint needed resettiné.

 
399

In the summer of 1967, the 1-kW L8-V-dc to 120-V-ac inverter, which
supplies ac power to one of the three safety channels, failed during
sw1tch1ng of the 48-V de supply. It was repaired by replacement of two
power transistors. During this same period the output of the ion chamber
in safety channel 2 decreased drastically during a nonoperating period, and
the chamber was replaced prior to resumption of operations. The trouble
proved to be a. failure in a glass. seal that allowed water to enter the mag-
nesia insulation in the cable, which is an 1ntegral part of the chamber.

A period safety amplifier failed when llghtning struck the power line to
the reactor site, and a replacement amplifier failed as it was being in-
stalled, The field-effect transistor in this type oféamplifier was sus-
ceptible to damage by transient voltages, and it was found.that under some
~conditions, damaging transients could be produced when the amplifier was
removed - from or inserted 1nto the system. A protective c1rcuit was de-
signed, tested and 1nstalled The module replacement procedure was modi-
fied to reduce the poss1bi11ty of damage 1ncurred on 1nstallation of the
module., Two relays in the safety relay matrices failed both with open.
coil oircuits. A chattering contact on the fuelupump motor current relay
caused safety channel 2 to trip several times before the problem was over-
come by paralleling two contacts on the same relay. A defective SW1tch on
the core outlet temperature also caused several channel trips and one re-
actor scram before the trouble was identified and the SW1tch was replaced.
A w1r1ng ‘error ‘in a safety circuit was  discovered and corrected | Interlocks
had recently ‘been added in the'"load scram. channels to drop the load when

the control rods scram._ A wiring design error resulted in these interlocks

- ”being bypassed by a safety Jumper. Although the c1rcuits were W1red this

'-way for a time before. being discovered the scram 1nterlocks were always
operative ‘during power operation, 51nce the reactor cannot go into the
'"Operate, mode when any . safety Jumper is inserted. '

" Late in 1965, the power’ supply to the model q-2623 relay safety ele-
~ ments was changed from 115-V—ac to. 32—V—dc. ThlS was done to eliminate
.ac pickup on the other modules through which the relay COll current was
routed, The 115-V—ac relays were not changed at this time. These func-
tioned satisfactorily until the-summer of 1967 when two of the 15 relays
failed. By the'end of l967;mseven,had failed, all with open circuits or_

 
 

 

 

400

safe conditions. In April 1968, all 15 were replaced with relays designed
for 32-V-dc operation. Soon after installation, 3 of these failed due to
contact welding (unsafe condition). This was apparently due to early |
failure of defective relays. There were no more failures during.extensive
tests made at this time or during subsequent operation. Starting Septem-
ber 30,.1968, daily tests were conducted on the entire rod scrams relay
matrix to detect single failures. (The matrix had been tested weekly be-
fore that.) Noise,suppressors were installed across some of;the non-
safety contacts of these relays to alleviate the noise which had sometimes
caused difficulty during in-service tests.

24,6 Performance of the Wide-Range Counting Chennels

Initial'criticality tests disclosed that the neutron flux attenuation
in the instru:hent penetration _.did' not follow an -i’dea..l exponential curire.
The deviation was too large to be sdequately comnensated for by the verni-
stats in the wide;range counting'channels. This is illustrated byrcurve
‘A of Fig. 2L.2 which shows fission counter response (normalized count rate)
vs withdrawal from the lower end of the penetration in guide tube 6. It
 was reasoneble to conclude from.this, end from similar curves, that the
excess neutrons respongible for the distortedrpart of the curve were en-
tering the penetration aslong its length. The count rates in. other guide
tubes nearer the upper half of the penetration were even more distorted
than curve A. Since a flux field with attenuation per Curve A precluded
successful operation of the wide-range counting 1nstrumentation, shields
of sheet cadmium were inserted in guide tubes 6 and 9 to shield the flSSlon
chanbers from stray- neutrons. Curve B of Fig. 2h 2, normalized count rate
' vs distance, shows the improvement for guide tube 6. o |

~ Reactor period: infbrmetion from the wide—range counting channels 1n—
hibited rod withdrawal and caused rod reverses.} Reactor period signals
from counting channels operating at low input levels are characterized.by
" slow response. This inherent delay produced a problem'w1th servo—controlled
rod withdrawal during start. In & servo-controlled start, the demand sig-
'nal‘caused the regulating rod to withdraw until the period—controlled."W1th—
drew inhibit" interlockzoperated. If the period continued toidecresse, the

c
 

 

 

 

NORMALIZED COUNT RATE

-10°
MSRE
GUIDE TUBE NO.9 -
- COUNT RATES IN GUIDE TUBE ' ' { A
-~ NO.6 BEFORE AND AFTER ‘ _

0! ' 'ADDING CADMIUM SHIELDING ‘ CONCRETE FILL g &’ ' : '
= ‘ N\ & SECTION THROUGH
i~ REACTOR SHIELD— | < NUCLEAR INSTRUMENT

_ - \ ' ' . PENETRATION

w—a | \ \\ l ' ) ’.I"{

. F— ' ' 57 -

- \ . \ \ GUIDE TUBE

w3kl X - - \ ’ A * NO.6 r

ELIN] A \ | ‘ &
- \ SN - &0 REACTOR -
SR e - N

N : _ \ : : g CONTAINMENT

ol N Ll 'MIDPLANE | 5in ‘ 4 CELL:
=T ‘SR OFCORE | § ‘ ,
= \ ' ‘ = — ) FISSION COUNTER

1073 |=— CURVE “B" COUNT \ \ _ \ , l ]““m i A .

© = RATE WITH CADMIUM \ ‘ o "’ i 7

|2 SHIELD ‘ ] } w .
S - ¢ N

. ' . REACTOR

oL - — CURVE "A" COUNT RATE— VESSEL

, = " WITHOUT CADMIUM
= SHIELD
) - CADMIUM SHIELDING

1077 L o : SUBASSEMBLY -~ TYPICAL

) 20 40 60 0. IN' GUIDE TUBES THIS LENGTH 100%
x, DISTANCE WITHDRAWN (in.) 6 AND 9 _ ( CADMIUM WRAP
- e _ 10f-0in. ' 1011-Qin. ‘ SECTION A-A
r__}'i o | : LOCATION OF
_ - : : GUIOE TUBES
, = =3 L3 O v{l (:)‘ |

 

    

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

OUTER SHELL REMOVED

FOR CLARITY—SHADED WEDGES
ARE CADMIUM —UNSHADED WEDGES
ARE ALUMINUM

Fig. 24.2 Guide Tube Shield in the MSRE Instrument Penetration

  
   

 

ORNL-OWG 66-2741 .

    
 

 

 

 

 

 

o~
)
ot
 

 

 

402

"reverse" interlock acted to insert the rodslin_direot opposition to the
servo demand. These "withdraw inhibit" and "reverse"'trip'points were
originally established at periods of +20 and +10 sec resPectively. The
‘delayed low-level response of the "inhibit" ioterlock allofied sufficient
incremental rod withdrawal to Produce a lb;sec period-and thus cause a re-
verse. The ‘situation was aggravated by coastlng of the shim-locating motor
in the regulating rod limit switch assembly. To correct this, ‘the "withdraw
inhibit" and "reverse" period trlp poxnts were changed to +25 and +5 sec,
respectively, an electro-mechanical clutch-brake was inserted in the shim-
locatlng motor-drive train and dynamic breking circuitry was installed for
the regulating rod drive motor, |
Throughout operetion, dlfflculty was experienced with m01sture pene-

tration into the fission chambers. The cause wes diagnosed as excessive
strain and flexing of the tygon)tubing sheath on the electrical cables.
'I'he avera.ge lifetime was about 6 months prior to July 1968. At this time
the type of tygon tubing used to cover the ceble and the method of sealing
the chamber connections were .changed. This seemed to improve the moisture
re31stance. : :

~ In early 1967, a failure occurred due to a short in the cable to the
: preamplifler. The drive tube unit was modified to provide for controlled

ceble bends and there were no reoccurrences .of this type of difficulty.

24,7 Performance of the_Linear Power Channels

The linear power channels and rod servo instrumentetion perfOrmed very
well throughout the operation. When the fuel was changed to'233U, the rod-
control servo was modified to allow an increase in the servo dead band to .
compensate for the more rapid flux response of the reactor.

The compensated ion chembers use & small electric motor to change com-
pensation. Early in 1966 one of these motors had to be replaced.

A water leak occurred in one of the compenssated ion chambers in the
summer of 1969. When it was examined, humerous leaks were found along the
seam weld in the 321 stainless steel bellows sheathing the cable, and the
aluminum can at the outer support'ring was"honeycoflbedioy eorrosion. (The
water in the instrument shaft contained lithium nitrite buffered W1th boric
acid for inhibition of corrosion.)
 

 

403

2h.8 BFz Nuclear Instrumentation

Because of very-unfavbrablergeometry,'the“strongest practicablepneu-
tron~source would not~prodmce72'counts/secu£rom the fission'counters in
the wide-range counting chahnels until the core vessel was:approximately
half full of fuel salt; neitherrwould itfproduce 2 counts/sec with flush
salt in the core at any level- .This'was the minimum count rate required
to obtain the permissive confldence 1nterlock which allows filling the
core vessel and w1thdraw1ng the rods. Therefore a counting channel using

a sensitive BF3 counter was added to estsblish "confidence" when the core

~ vessel was less than halfefilled with fuel salt. This was installed early
- in 1966. The chamber had to be replaced in the summer of 1967 due to mois-

ture leakage'into the cable. .No other troubles were encountered.
24,9 Nuclear Instrument Penetration

‘The:nuclear power produced at the‘MSRE was_determihed by an overall
system'heat balance. All;nuclear power inStruments were calibrated to agree
with this primary standard. Durlng extended runs at higher powers, the
nuclear instruments 1nd1cated 15 to 20% higher than the heat balance. Th1s
was found to be due to a rise in ‘water temperature in the nuclear instru—
ment penetratlon,which apparently changed the attenuatlon characteristics

of the water. In June 1966, & heat exchanger system was placed in operation

- to cool the water which reduced the temperature change from zero to full
| power to A18°F -from 72°F and reduced the difference in the two power meas-

_'urements to about 57

- 24,10 Pé&férman¢é°6f*£fié Process Radiation Monitors

. The process radiatlon monltors proved very rellable.p,No-radiation

._elements had to be’ replaced durlng the entire operation. Occasional re-
;:pairs were necessary on .the electronics. _One rod end load scram resulted
from a false signal from- one of these monltors (RE-528).

 
 

 

 

404

24,11 Performance of the Personnel Radiation = -
._Mbnitoring,andrBuilding Evacuation System

During early testing it wes found that there were some areas where

the building evacuatlon horns could not be heard Two addltional horns

and four addit10nal beacon alarm 1ights were installed Other than oc-
ca31onal minor repairs, the system has functioned satisfactorlly.

2h.A12 . Performance of the: Stack Monitoring System

The . stack mbnitoring sysfem.was very relisble.- The manuallrange

_-SW1tching made it difficult to 1nterpret date from the recorder charts ‘and

the: Rustrak recorders were very inconvenient to use..

2h 13 Performance of Thermocouples-end the- .
Temperature Readout ~gnd~Control- Instrumentation

A total of 1071 thermocouples were installed at the MSRE. Of these,
866 were on salt systems (351 on the. circulating loops and 515 on the drain
tanks,ydrain-lines, end freeze vaelves). Of the 1071, only 12 have failed

in five_years.of-serviee; 'Five others were damaged during construction and

maintenance. A breskdown of the failures is given in.Table 2L.l,

Table 24.U4 Failures Among the 1071 MSRELTflermOCOuples

 

.. Nature of Failure 2 ¢ .. Number

 

Damaged during construction
Damaged,duriag maintenance'
Lead ofiefi-during operation
Abnormally low reading (detached?)

Unknown reason

Ww w o N w

T ' Total | AT

 
 

 

405

Only three thermocouple wells were provided in the c1rculat1ng salt
systems " one each in the coolant radiator inlet and outlet pipes ‘and one
in the reactor neck. The remalning thermocouples vere attached to the pipe
or vessel walls. The thermocouples on the radiator tubes were insulated
‘to protect them from the effects of the high-velocity air that flows over
'”them during power operation; the others were not insulated and thus were
subject to error because of exposure to heater shine and to thermal con-
vection flow of the cell atmosphere within the heater insulatlon. In
March 1965, with the fuel and coolant systems circulating salt at isothermal
condltlons, a complete set of readings was taken from all the thermocouples
that should reed the temperature- of the circulating salt, A similar set of
data was taken in June 1967 at the start of Run 12, The results of the two
sets of.measuremehts are shown in Table 24.5. Comparison of the standard
~deviations for the radiatorrtherm000uples‘with those 'for the other thermo-
couples shows the effect of_inSulation on reducing the scatter,,;Comparison
of the sets of data taken over two years apart shows very little\change,
certainly no greater scatter. -Figure 2h.3 shows that the statistical dis-
tributioniof the deviations on individual thermocouples from the mean also

changed little in the two years.

-

Table 24k.5 Comparison of Readings of Thermocouples of
Salt Piping and Vessels Taken with the Salt Isothermal

h™

 

8 Indlcated Temperature (°F)
Thermocouple -

 

.. Location '.fl'March 1965 © June 1967
"rRadiator tubes - 'll62;6 % 6.7f o _1208.5 # 3 3
Other - - 1102.1 % 13.0 - 1206.7:+-12.3

AL - 1102,3 % 10,6 . 120T.4 £ 9.8

 

 
 

 

406

ORNL-DWG 67-41786

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 ¢ , — fg' o Te—® o -
o5 R
‘ O
O
90 o
g
&
<
= 80 5
g o
o .
. a 0 °
Lt
I
U ® DATA TAKEN -
g 60 ~ MARCH 1965 5
- o DATA TAKEN °
a | JUNE 1967
w 50 '
- ' o
W
S | [
o 40
< _
L
m §
S 30
e
i | o
E 20 —8
o : 0
&
. .. . ’o
10 3
. . Q
"
- o485
Ole—L o ov%

 

 

 

 

 

 

 

-60 -40 -20 O 20 40 €0
DEVIATION FROM AVERAGE (°F)

Fig. 24.3 Comparison of MSRE Thermocouple Data from March 1965 and June 1967
 

 

407

The scatter in the various thermocouple readings was reduced to an
acceptable level by using biases to correct each reading to %he overall
average measured while both fuel and coolant systems were circulating salt
at isothermal conditions;“rThese biases were entered into the computer and
were automatically applied to the thermocouple readings. The biases were
revised at the beginning of essh run and were checked when isothermal con-
ditions existed during the runs. Generally the biased thermocouple readings
were reliable, However in a few cases, there were shifts which caused cal-
culational errors. -

During early operation the thermocouple scanner gave considerable dif-
ficulty due to 60~cycle noise pickup, poor stability, and drifting of the
salvaged oscilloscopes. Refinements were needed in design to allow better
identification of scanner'poifits and provide a means by which the operator
could calibrate the instruments. After these were corrected, the system

operated very satlsfactorlly. The rotating mercury switches lasted much
7'longer than'the expected 1000-hour mean llfe. One switch failure occurred
when the nitrogen purge gas was 1nadvertently stopped. | |

Single point Electra Systems alarm switch modules were used for con-
trol of freeze valves and for other alarm and control_actlons, These gave
considerable trouble during early operation due to drifting or dual set-
points and general maloperafidn. A number of mo&ifications were made to
correct these. Printed circuit—board contacts were gold-plated to reduce
contact resistanee, the trim pots used for hysteresis adjustment were re-
placed with fixed resisfiofs; and resistor values in modules hafiing ambig-

- uous (dual setp01nts were . changed to restore the proper blas levels. These
changes, together w1th stabllizatlon by aging, of crltlcal resistors in
,the SW1tch modules and more :1gorous perlodlc testlng procedures<1mproved
the performance. AAcheck sh6fie§ that out of 109 switch setpoints, 83% had
shifted less than 20°F Q#er sfsix-month'period. Multiple setpoints still
occurred and various other*faiIUres were encountered. During'1968 and
1969, records were kept on . the fallures of 1nd1cator llghts on these mo-
dules, There were about 50 fallures per year. ‘This was 1mportant because

& burned-out llght bulb could cause alarmror control actlon.

 
 

. S e

 

 

N 408

24 .14 - Performance of PressurewDetectors_{

Most . of the pressure instruments at the MSRE performed very well, Dif-
ficulty was- encountered with the differential pressure cell used to obtain
the pressure drop in the helium flow through the charcoal beds (PAT-556).

The span end zero settings shifted badly although the pressure capsbility

had never been exceeded. Three dp cells failed in this service. Two of
Aé}these were removed, tested, and inspected without determining the cause of
" the trouble._ The last’ replacement functioned satisfactorily at first but
 then gave similar difficulties. It is still instelled:

_2&,15' Performancenof.Level Indicators

The: bubbler-type level instruments used in the fuel pump overflow tank
and coolent pumps performed well. More details are given in. Sections 5 8

and 6.5. Some difficulty was encountered in controlling the purge flow un-

\til the throttling velves vere replaced,-

A high and. 16w level resistance type level probe was provided on. each (i}
drain tank. During early operation the excitation and signal cable leads
on both probes of the fuel flush tank failed. These failures were caused
by - excessive temperature which caused oxidation and embrittlement of the
copper-clad mineral—insulated-copper—w1re cables. These cables were de—
signed on the assumption that they would.be routed in- eir. above the tank
insulation and that their operating temperature would-not exceed 200°F-

however, in the actual installation, the cables were covered with 1nsula—'

tion and the temperature at the point of attachment to the probe was prob-

ably 1n excess of 800°F Repairs were accomplished.by replacing the copper-

- ‘cled, mineral-insulated copper wire excitation and - signel cables and por-

“tions of the probe head assembly with a stainless-sheathed, ceramic-beaded

nickel-uire cable assembly. In the summer of-1968, one of the probes in
fuel drain tank No. 1 failed. The failure was found to be an open lead
w1re inside the cell The probe-wasrrestored to service by a cross con-
nectionvouts1de the cell to the equivalent lead of the other probe,' |
The’ initial 1nstrumentation provided to assure proper Water level in

the ‘vapor-condensing tank consisted of Ik resistance type prdbes spaced ‘ U
 

et bt s brdnnon gl @ mre agigg ere

 

 

409

h in. apart near the desired water level. In September 1966, while doing

------

the instrumentation startup check 1list, one of the two high-level sw1tches

was found to be defective.j Loss_of this switch caused the loss of one‘chan-"

‘nel of information needed to'establish'that the water 1evel in the tank was

correct ' Since a second switch failure might require that the reactor be

shut down until the sw1tches could be repaired, and since the removal of

Vthe SW1tches from the tank is a difficult operation, a bubbler-type level

: measuring system, which 1ncluded a containment block valve and associated

safety circuits in the purge supply, was designed and was installed in the

~reactor~-cell vapor SUppre381on tank. This installation utilized 8 dip tube

vhich was included in the original design in. anticipation'of such need.
This new level system also enabled the operator to check the water level

in the tanks as a routine procedure.

ol 16 - Performance of the Drain Tank Weighing Systems

The same type pneumaticiueighing devices were used on the fuel.drain

.tanks and the coolant drain tank' ‘The coolant drain- tankrweight indicators
' proved to be stable, show1ng no long—term drlfts or effects of external
'variables.- With 5756 1b of salt in the tank at about l200°F the extreme

spread of hO 1ndicated weights over a period of a week was + 22 1b. This
was only 0. h% and was quite satisfactory. o
' The 1nd1cated weights of ‘the three tanks in the fuel system exhibited

rather large unexplalned changes.- In some cases these amounted to 200 to

,300 pounds. The mechanism causing this was not definitexy established but
probably was due to changes in forces on the syspended tanks as tempera-

_tures of attached piping and the tank furnaces changed. Reactor.cell pres-

sure seemed to also affect the readings. The weighing systems were useful
in observing transfers of salt and for filling and draining the reactor.
In addition to this calibration drift difficulty was experienced with

" the multiposition pneumatic selector switches. Manometer readout was ac-
”:complished by selecting a particular weigh cell channel‘uith pneumatic'se-

tr_lector velves. The valves were composed of & stacked array. of indiV1dual

valves operated by cams on the operating handle shaft. Leaks 'in these

valves gave false weight indications. A redesign of the switching device

solved this problem.

 
 

410

Prior to power operation, one of the weigh cells failed and was re-
placed. " The failure was determined- to be due to p1tt1ng of the baffle and
" nozzle in the cell, This pitting was apparently caused by amalgamation of

| mercury with the plating on the baffle and nozzle, How the mercury got in-
,‘to the weigh cells has not been determined however, it was believed to
have come from the manometers and to have been precipitated on the baffle
| by expansion cooling of- the air leav1ng the n_ozzle° Apprec1able quantities
of merciry were also found in the.tare;preSSUre regulators on the control
panel; however, no mercurypwas'found'in thedinterconnecting tubing_or in

other portions of the system,

24,17 Performance of the Coolant Salt Flowmeters |

Several days after the start of coolant salt circulstion, the output
of one of the two salt- flowmeter channels started drifting down scale. The
output of the other chpnnel remained steady. The trouble was-isolated to
a zero shift and possibly a span‘ahift in the NaK-filled differential pres-
sure tranSmitter in the drifting channel Since the exact cause could not
be determined e spare dp cell was installed. Both channels functioned
satisfactorily throughout the remainder of the reactor operations,; Tests
on the defective unit were inconclusive- however,_ it was determinedithat
the shifts were temperaturewinduced zZero shifts possibly caused by incom-
plete f£illing or a leek in the - silicone oil portion of the‘instrum.entq

An operational inconvenience per51sted “throughout’ operations° When-
ever the radiastor air flow was increased, air leaking through the 1nsula~
tion around the salt legs to the dp cells caused the temperature to decrease
1rap1dly. This necessitated adJustment of the heaters.

| Due to the discrepancy between the reactor power level 1nd1cated by
fuel burnup, ‘heat balance ete., it is planned to recalibrate the dp cells
dnring the next fiscal year.
 

 

 

 

411

2,18 Performance of Relays

The difficulties encountered with the control red relays:are described
in Section 24.5. Experience with other relays is given below.

After ebout 2 years of operation, the 48-V-dc-operated relays showed

~considerable heat damage to their bakelite frames. The manufacturer,

General Electric, advisedvthat.overheating of this particular model was

a common problem if the relays were continuously energized. Early in 1967,
twenty of the 139 relays were replaced with a later improved model. Within
a few months some of these also showed signs of deterioration. Therefore
in June 1967, all 139'relayS‘wefe field—mbdified'byrreplacihg the built=in

resistors with externally_mounted_resistors.-‘No trouble was experienced

after this modlflcation.

In September 1969, the load-scram c1reu1t » Which drops the radiator
doors and stops the blower, tripped several times. Investlgatlon showed
that some of the relasy contacts had developed unusually high resistance
due to oxide films, Because of the way the contacts were paralleled in
the matrix, the film was not Burned off each time the contact closed, as
in a fiormei application. These,contaetsrwere cleaned and no further dif-

ficulty was encountered.

24,19 Training Simulation

 

Two "on~site" reactor kinetics simulators were developed for the pur-

 pose of training the MSRE operstors in nuclear startup and power operation.

The startup simulator -used the control rod position signals as inputs, and
provided outputs of log COunt'fateg period, log power, and linear power,
The resctor's period interlocks, flux control system, and linear flux range

selector were also operationel ~ In addition-to this, the power level simu-

:lator used the radiator door p051t10n and cooling air pressure drop signals

as .inputs and provided readout of key system temperatures.. Both simulators

were set up on general purppse,_pprt&ble EAT TR-10 analog computers. Much

of the actual MSRE har&ware;feuch-as control rods were used rather than

simulated. Thus the operatofs manipulated the actual reactor controls and

became used to the instrumentation and controls system. This proved to be

a very relisble training tool.

 
 

 

412

2k.20 Miscellaneous

The original mass spectrometer used to monitor for beryllium in the
coolant stack was replaced with an improved instrument near the start of
power operation. Onlyroccasional repeirs or preventative maintenance was
required since then. | | | : | |

Fdlded-charts were used on some recorders.-  Thése did not function

properly due to the low chart speeds being used.

24,21 Conclusions andaRecommendations 

_.'Considering the quantity and complexity of the instrumentation, a
mihimal amount -of difficulty was encountered. Since the MSRE was an ex-
perimen€al réactor; it was in many areas over-instrumented. This was proba-
bly due to not -knowing what information might be needed and not taking
enough credit for the gbility. of the operators. This led to unnecessary
difficulties in normal operstions or in running special experiments. At
the same time, there were areas where additionsl informetion would have

been beneficizl.

-
 

 

 

 

. 1'0

3.

4.

5,
6.

7.

9.

10.

11,

12.
13,
14.
15.

16.

s
- and Foam in the MSRE 0RNL—TM+3027 (June 1970)

: 719'

20,

21,

413
 REFERENCES

M. W. Rosenthal, P, R, Kasten, and R, B, Briggs, "Molten-Salt Reac-

~ tors —-History, Status, and Potential, " Nucl. Appl Tbch., 8, 107

(1970).

| MSR Program Semiannu. Progr. Rep., Jan. 31, 1964, ORNL-3626.

MSR Program Semianmu. Progr. Rep., July 31, 1967, ORNL-3708.

' MSR Program Semiannu. Progr. Rep., Feb. 28, 1965, ORNL-3812.

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

MSR Program Semiannu. Progr.’ Rep., Feb. 28, 1966, ORNL-3936.

MSR Program Semiannu.Progr. Rep., Aug. 31, 1966, ORNL-4037.

. MSR Program Semiannu. Progr. Rep., Feb. 28, 1967, ORNL-4119.
- MSR Program Semiannu. Progr. Rep., Aug. 31, 1967, ORNL-4191.

" MSR ProgramVSCmiannufcRrogr,rR@p., Feb. 29, 1968, ORNL-4254,

MSR Program Semtannu Progr.,R@p.,rAug. 31, 1968, RNL—4344.
MSR Program Semzannu;Progf; Rep., Feb. 28, 1969, ORNL-4396."

MER Program Semtannu.'Pfégr.flep.,_Aug. 31, 1969, ORNL-4449.

MSR Program Semzannu.-Prbgf. Rep., Feb. 28, 1970 0RNL-4548.

‘R, C. Robertson, MSRE Design and Operations Report Part .-I — De-

scription of Reactor Design, 0RNL-TMF728 (Jan. 1965)

R. H. Guymon, P N Haubenreich and J.,R Engel MSRE Design and

__Operations Report Part XI-— Test Program, RNL—TME 11 (Nov. 1966)

TR. B, Lindauer, Processing of the MSRE Flush and Fuel Salts, ORNL—
- TM-2578° (Aug.,1969).:~,;;g,, |

Jo R. Engel P.-N..Haubenreich and A. Houtzeel Spray, Mist Bubbles

P. N. Haubenreich and M Richardson, Plans for Post—Operation Exami-
' nation of the MSRE ORNL-TMFZ974 (April 1970).

}C. H, Gabbard Reactor Power Measurement and Heat-Transfer Performance
in the MSRE, 0RNL—TM~3002 (May 1970) -

J. R. Engel MSRE Design and ‘Operations Report Part XI—A-— Test
Program for 2*2U Operation, ORNL-TM-2304 (Sept. 1968)

 

 
 

 

 

22,
23.
24.

25.

26 *

270

28.
29.
30.
31.
32.
33.
34,
35,
36.

37.

 

414

-

References (continued)
R H. Guymon, MSRE Design and Operations Report Part VIII, Operating
Procedures, Vols, I and II, ORNL-TM-908 (Dec. 1965).
R. E. Thoma, Chemical Aspects of MSRE Operntions, ORNL-4658 (Dec. 1971).
R. H. Guymon and P. N. Haubenreich, MSRE Design and Operations Report,
Part VI, Operating Safety Limits for the Molten-Salt. Reactor Experiment,
0RNL—TM—733 (3rd revision), (July 25, 1969).
R. H. Guymon, R. C. Steffy, Jr., and C. H. Gabbard, Preliminary‘Evalu-
ation of the Leak in the MSRE Primary System Which Occurred during the
Final Shutdown, ORNL-CF-70-4-24, (April 22, 1970).

P. G. Smith, Development of Fuel- and Coolant~Salt Centrifugal Pumps
for the MSRE, ORNL-TM-2987 (Oct. 1970).

C. H. Gabbard, Inspection of MSRE Fuel Circulating Pump after the

‘Zero-Power Experiments (Run 3), ORNL-CF-66-8-5 (Aug. 1966).

R. B. Briggs, Measurement of Strains in Heat Exchanger Nozzle and
Piping of MSRE, ORNL-CF-66-2-66, (Feb. 1966).

C. H. Gabbard, Thermal-Stress and Strain-Fatigue Analyses of the MSRE
Fuel and Coolant Pump Tanks, ORNL-TM~78 (Oct. 1962).

C. H. Gabbard, R. J. Kédl, and H. B. Piper, Heat Transfer Performance
of the MSRE Heat Exchanger and Radiator, ORNL~CF-67-3-38 (Mar. 1967).

C. H. Gabbard, Reactor Power Measurement and Heat Transfer Performance
in the Molten Salt. Reactor Experiment, ORNL-TM-3002 (May 1970).

H. E. McCoy, An Evaluation of the Molten Salt Reactor Experiment

Hastelloy N Surveillance Specimens — First Group ORNL-TM 1997 (Nov. 1967).

H. E. McCoy, An Evaluation of the Molten Salt Reactor Experiment Hastel-
loy N Surveillance Specimens — Second Group ORNL-TM-2359 (Feb. 1969).

H. E. McCoy, An Evaluation of the Molten Salt Reactor Experiment Hastel-
loy N Surveillance Specimens — Third Group, ORNL-TM-2647 (Jan.‘1970).

H. E. McCoy, An Evaluation of the Molten Salt Reactor Experiment Hastel-
loy N Surveillance Specimens — Fourth Group, ORNL-TM-3036 (Mar. 1971).

H. E. McCoy and B. MbNabB, Intergranular Cracking of iNOR—SCin the MSRE,
ORNL-4829, (Nov. 1972). _ .

C. H. Gabbard, Design and Construction of Core Irradiation—Specimen

" Array for MSRE Runs 19 and 20, ORNL-TM-2743 (Dec. 1969).

C

O
 

 

 

415

'References (continued)

-38. R. B. Briggs, Effects of Irradiation of Service Life of MSRE, ORNL-
CF-66-5-16 (May 1966) : . ,

39. R. B. Briggs, Assessment of Service Life of MSRE ORNL—CF—69 8-3,
(Aug. 1969). |

40, B. E. Prince et al., Zero-Power Physics Experiments on the Molten-Salt
Experiment, 0RNL-4233 (Feb 1968) '

41. J, R. Engel and B. E. Prince, Zero~Power Experiments with 2°°U in the
MSRE, ORNL-TM-3963 (Dec. 1972) _

42. C. H.-Gabbard and C.;K,xMcGlothlan,-A Review of Experience with the
MSRE Main Blowers, 0RNL—CF—67—4—1, (Ahg. 1967).

43, C. H. Gabbard, Bearing Failures of Main Blowers MB-1 and MB-2, ORNL~-
: CF—67—4—1 Addendumrl (Mar. 1968).

44, R, L. Baxter and D. L. Bernhard "Vibration Tolerances for Industry,"
paper presented at ASME Plant Engineering and Maintenance Conference,
Detroit, Michigan, Paper No. 67-PEM-14 (Apr. 10-12, 1967).

45. E. C. Parrish and R, W. Schneider, Tests of High Efficiency Filters
: -and Filter Installation at ORNL,; ORNL-3442, (May 1963).

46. H. R. Payne, The”Mechanical.Design of the MSRE Control Rods and Re-
actor Access Nozzle, MSR-61-158 (June 1961).

47. J. L. Crowley, Temperature and Salt Levels in the MSRE Reactor Access
Nozzle — A Possible Relation to MSRE Power Blips, MSR-69-13 (Feb. 1969).

48, P, N, Haubenreich, R. Blumberg, and M. Richardson, "Maintenance of the
- Molten-Salt Reactor Experiment'" paper presented at Winter Meeting, ANS,
Washington D. C. (Nov. 1970) . _

49, C. H. Gabbard, Radiation Dose Rates to the MSRE Fuel—Pump Motor and
Control Rod Drives, MSRF65“48

50. R. C. Robertson, MSRE Design and Operations Report, 0RNL-TM—728

51. R, H. Guymon, MSRE Design and Operations Report, Part VII Operating
Procedures, ORNL"TMFQOB Vol. II (Dec. 1965) .

52, R. B. Gallaher, MSRE Sampler—Enricher System Proposal ORNL-CF-61-5- 120,
- (May 1961).

53, MSRE Staff, MSRE Sampler-—Enricher, An Account of Recent Difficulties
' and Remedial Action, MSR~67«73 (Sept. 1967).

 
54-

55.".

56.
57.
58.
59.

60.

61.

62.

63.

 

416

References (cpntinued)

P. N. Haubenreich and MSRE Staff, An Account of Difficulties with the
Sampler-Enricher that Led to a Second Capsule Being Left in the Pump
Bowl, MSR-68-125 (Sept. 1968).

S. E. Beall et al., MSRE Design and Operations Report Part V, Reactor

‘Safety Analysis Report, ORNL-TM-732 (Aug. 1964).

R. H. Guymon, Tests of the MSRE Reactor and Drain Tank Cells, MSRr
62—95 (Nov. 1962). |

P. N. Haubenreich et al., MSRE Design and Operations Report Part III,
Nuclear Analysis, ORNL-TM-730 (Feb 1964) .

G. H. Burger, J. R. Engel, and C. Do Martin, Computer Manual for MSRE
Operators, ORNL-CF=~67-1-28" (Jan. 1967) .

P. N. Haubenreich, Safety Considerations in Resumption of MSRE
Operation, 0RNL—CF—69—8—10 (Aug. 1969).

'P. N. Haubenreich, ReSponses to RORC Recommendations for MSRE, MSR-

B. F. Hitch et al., Tests of Various Particle Filters for REmoval of &;fi
0il Mists and Hydrocarbon Vapor, ORNL-TM-1623 (Sept. 1966).

W.VC.-Ulrich, Purge of MSRE_Off—gas Line 522, MSRr66—8; (Mar. 1966).

R. H, Guymon and J. R. Engel, Introduction of Oxygen in Fuel-Drain
Tank, MSRr67—1 (Feb. 1967).
 

 

 84-85.

1.
2.
3.

5.

.6.

7.

9.

10.
11.
12.
13.
14-18,
19.
20.
21.
22.
23.
24,
25.
26.

57+58.

. 50-60.

61.
62,

63.

64.

54-56.

417

 

ORNL-TM~3039
Internal Distribution
S. E. Beall o 27. R. N. Lyon
E. S. Bettis o -~ 28. R, E. MacPherson
R. Blumberg o - 29. H. E. McCoy
R. B. Briggs : : 30. H. C. McCurdy.
W. B. Cottrell - | 31. A, J. Miller
J. L. Crowley S ' 32. R. L. Moore
F. L. Culler ' 33. L. C. Oakes
S. J. Ditto- : o 34. A. M. Perry
J. E. Engel 35. M. Richardson
D. E. Ferguson o - 36-37., M. W. Rosenthal
A. P. Fraas ' 38. Dunlap Scott
C. H. Gabbard o 39. M. R. Sheldon
W. R. Grimes 7 40. M. J. Skinner
R. H. Guymon oo 41, I. Spiewak
P. H. Harley o 42. D. A. Sundberg
P. N. Haubenreich _ 43. J. R, Tallackson
H. W. Hoffman o 44, R. E. Thoma
T. L. Hudson , 45. D. B. Trauger
P. R. Kasten : 46. G. D. Whitman
A. I. Krakoviak 47-48. Central Research Library
Kermit Laughon, AEC—OSR' 49, Y~12 Document Reference Section
M. I Lundin : 50-52. Laboratory Records Department

53. Laboratory Records (RC)

. External Distribution

Director,.Divisionrof Reactor Licensing, USAEC, Washington D.C. 20545

" Director, Division of Reactor Standards, USAEC, Washington, D.C. 20545

N. Haberman, USAEC, Washington, D. C. 20545
D. F. Cope, AEC-ORO

‘M. Shaw, USAEC, Washington, D.C.. 20545

65.

66-82.

- 83.

David Elias,_USAEc,-Washington, D.C. 20545

A, Houtzeel, TNO, 176 Second Ave., Waltham, Mass. 02154

R. C. Steffy, TVA 303 -Power Building Chattanooga, Tenn. 37401
Manager, Technical Information Center, AEC (For ACRS MEmbers)
Research and Technical: Support Division, AEC, ORO :

Technical Information Center, AEC .