ORNL-TM-1856.txt

 

) fi_;d r

    

OAK RIDGE NATIONAL LABORATORY
operated by

UNION CARBIDE CORPORATION
NUCLEAR DIVISION e
for the

U.S. ATOMIC ENERGY COMMISSION

ORNL- TM- 1856

AP
COPY NO. - A Aiud

DATE - May 22, 1967

CESTI PRICES

INSTRUMENTATION AND CONTROLS DEVELOPMENT FOR H.C. $.3.0% ; MN G5
MOLTEN-SALT BREEDER REACTORS

J. R. Tallackson, R. L. Moore, 5. J. Ditto

ABSTRACT

Instrumentation in use in the MSRE provides a good
basis for development of the instrumentation for large
molten-salt breeder reactors. The development would in-
volve primarily the testing and improvement of existing
instrument components and systems. New or much improved
devices are required for measuring flows and pressures
of molten salts in the fuel and blanket circulating systems.
No problems are foreseen that should delay the design or
construction of a breeder reactor experiment.

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

DESERTION OB B

7
e gt i et

e erea s o £ e 3w

gt e

 

 

 

 

~A. Makes any warronty or representation, expressed or implied, with respect to the accuracy,

 

,LEGAL NOTICE

This roport was preporsd os an account of Government spomond work, Neither the Unlhd States,
nor the Commission, nor any person ucflng on behalf of the Commission: : :

completoness, or usefulness of the ‘information contained ‘In this report, or that the use of
any information, apparatus, method, or process duclosed in this report may not infringe
privately owned rights; or

B. Assumes any liabilities with uspcc' to the use of, or for damages rosulting from tho use a! e

any information, apparatus, method, or process disclosed in this report.

T At 81 1 W it

 

‘As used In the abave, “person acting on behalf of the Commission” includes any omploye- or - :

controctor of the Commission, or 'mployeo of such contractar, to the axtent that such employes - - ‘

or contractor -of the Commission, or smployes of such coniroctor prepares, disseminates, or - : i'

provides access to, any information pursuant 1o his employment or ‘eontract with the Cornmuslon, - .

or his omploymcm with such contractor, o o i ' i . ) &‘\ ;
. : : : I

 

 

 

 

 
=,

-} @’ %]

e = e

TABLE OF CONTENTS

Abstra.ct v-oono-oci-'oq-co>o:9oé_ooao‘|‘oonooo-ooitoco_ocooo.oonto .

IntrOduCtiQn l."..ill"l....’.'..l..'...."".l.‘OIOQ,DQO-
Instrmentirlg the mBE .."‘..".........Ol..i'..'...l.'..
Nuclear Instrument Components and. Systems tvesesesstrrsene

Process Instrument Components for Direct
Application of Molten Salt Loops N

Flcw Measux’ement OI.Q_I.I..OI.»""..l‘l"l..:....l’..*‘

salt Inventory -4.....;-65-..-;.....-...q...',--..--...,

Temperature Measurement semssvesasreriiaaniesiiinnen
Level Measurement ""';';""'"'f"'f"""""f"'
Pressure Measurement .cseceessscarsscsssscrsssconnconas
Differential Pressure Measurement «eseeveracecasesons

Process Instruments to Operate Auxiliary . ,
Su-b'systems ..q.o—.-..'--..-_-c,-qc...7-.1......--.--....y.}-..-._

Health Physics Re,di.ation 'Monitoririg . -.} ceersennaryas .. ‘e .
Steam Plant Instrumeni;at'ion ...,
Computer Control and Data Logging Ceeraseseretinetaeenees
Beryllium Monitoring reessesetesavssettesssnenanerrsenns
Component Test end_Evaluatioo rereresearensieiiitanaen
General i.eeecesercrsscetstonesctsensconstrrintanans
Electrical Control Circuit COMPONENts «seovssoncasrss
Helium Flow Elements ................................

G&B System COntrOl valve ‘.'.....00.0'.'..0...0.......

Temperature Scanners O.Ul'.l.!.....l..O.l...l't“.'..'r

‘-',Tempera.ture Alarm Swrbches R

PI‘OCESS Radiation Monitoriflg e-'q.v-eo.qoooopflonc.nocootu(

Wa.ste Effluent Monitoring

':'i..'looo.poo.ooco..n..cc"l!!

 

 

Estimate of Cost of Development Progra.m .........,........7_

T T lEeAL NOTICE

| This report was prepa.red as an account of

; States, nor the Commission, nor any person acting of behalf of the Commission:

W W

11
1k
17
18

20
20
21
22
24
25
25
26

26
.26
27
28

., 'Containment Penetra.tion Seals tressvesrearsiriertaves

a8

29

30
31

Govermnent sponsored work. Neither the United

A. Makes any warranty oy representation, expreued or lmpu,d. with respect to tho accu-

| rucy, completeness, or usefulness of the info
' of any information, apparatus,
; privately owned rights; or

rmation contsined in this report, or that the use
method, or process dsclosed in this Treport may not infringe

.. B, Auumesaayuahmuaa‘df.hmpectmtheu,eof, orlordnmnceamulung from the -

i wse of any fnformation, apparatus, method, or process disclosed in this report.
! As used in the above, “peraon acting on
© | ployee or contractor of the C issfon, o 1 of
- 1" such smployee or contractor of the Comnuuion o:I amployee of such
.. disseminates, or provides access to,
i~ with the Commission, or his omployment vrlth such eontuctoi
i — e ARG S WA,

1\,\-

.....

s

bebalf of the Commmlon" tacludes any em- |
such tor, to the extent that

sontractor prepures,

any information putsuant to llll employment or oonmct }

 

 
' n

g

)

-} (‘f )

”'pscale MBBR

 

 

vIntroduction

-

soperation.of'the'MoltenQSalt Reactor Experiment (MSRE) indicates
that inadequate instrumentation should not become a.berrier to

_further development of.moltendseltfreactors;t.Most'of the process

instruments are standerd'indnStrial‘instruments. ‘Some of them
should be upgraded to provide the greater reliability and performance
desired in & nuclear plant. ‘The nuclear instruments are ‘a new.
generation of solid state instruments. Normal evo'lut'ion should
provide even better equipment for future reactors. Operating

experience has confirmed thet,some process - instrument components

- for direct use in the mOltentflfioride'Salt are ‘still developmental.

A substantial program is required to convert those components
1nto industrial grade 1nstruments suitable for specification by
an architect ~engineer. . Development of primery sensors for measuring

Process and nuclear varisbles. in -8, highly radiocactive, high

‘temperature environment is particularly de51rable.

~ The reference des1gn of a molten salt breeder reactor (MSBR)
is described in ORNL—3996 (Ref 1) Criteria for a molten salt
breeder experiment (MSBE) and a schedule for designing and building

 that reactor are presented. 1n TM—lBSl (Ref 2). “The instrumentation
needed for those reactors hes been exemined by comparison with the
~ MSRE and with emphesis on problems assocleted with heeting the

reactor salt systems in an oven.+ A program is proposed for development

of 1nstrumentatlon for the MSBE., The proposal does ‘not include

'a discussion of control rods or other means of reactivity control,

thlS ie 1ncluded in TM 1855, Component Development Program. A goal

’ of this development is to provide 1nstrumentation thet reqnires onLy

 Instrumenting the MSEE <

It is convenlent and appropriste to discuss the instrumentation

l ngmmumm§:E¥WSQngmflIEgmmnflfifi

 
 

 

 

of the MSBE (or any other reactor which is an extrapolation of MSRE- |
developed know-how) by subdlviding the complete instrument system
thus: 7 L '
1. Nuclear ilnstrument components and systems. S
2., Process type instrumentfcomponents~for direct_applioation';.“r
| +t0 molten-salt loops. :' o N
3+ Process type instruments required to qperate auxiliaries.r
b, Rediation monitoring (Health Physics) instrumentation.
5. Steam plant instrumentation. | Lo o
6. Computer control and data logging.
Te Beryllium.monitoring. )
8. Reactivity control.
9. Component test and evaluation (new products, new methods,f
_etc,) ' o '
10. Process radiation monitoring.

The following'detailed dlscussion of these categories 7
constitutes 8 preliminary appraisal of the anticipated development
engineering needed to provide an adequately instrumented.MSBE. ‘

- Nuclear Instrument Components and. Systems

The MSRE employs two wide range counting channels and a BF3
channel for startup control and unambiguous power measurement
over the entire opersting range of the reactor. Two linear
current channels, deriving their input signals from'éompensated
ionization chambers, are equipped with range chenglng devices ‘.l
gnd are used in conjunction thh.temperature measurements for ”J
sutomatic control of reactor power. Three linear 1eve1 safety |
channels using non-compenseted ionization chambers provide high "f"
flux scram signals for the safety system. o

A majority of the electronic components which mske up these
nuclear instrument channels are ORNL's recently &eveloped line
of solid state, modular components designed specifically for
reactor control and safety. The performance of these instrument

o,
 

 

>3

=) ( )

 

| modules, as individuai units, has been uniformly excellent; the.

performance of the sub-systems or. control channels which are
formed by assemblingand interconnecting the modules has been .
equally good. These instruments, in their present state of
development and measured by today's staendards, are satisfactory;

however, continuing engineering development is foreseeanhibh‘will'

take advantage of the rapid and continuous evolution of in-

strument and control system technology. The development of new
modules and ecircuits will-be~required to meet those safety and
control reqplrements pecullar to the MSBE. 1In addition, it-is
anticipated that 1nterfac1ng the modular nuclear instrumentation with
non-modular processtequipment will require development of special |
components, The fise of a digital computer in the MSBE system

will require development. of suitable 1nterfac1ng equipment in

the modular line. _ , _ . '

A1l neutron sensors used in the MSRE are in a water-filled
venetration whose temperature does not exceed 160°F. While |
the presently available senscrs are quite satisfactory in such
an environment, they could not be used without major modification
at a considerably higher,tefiperature.f Considerable development
will be required to provide sensors and sensoi posationing

equipment capable of rellable operation in the severe temperature

environment of the MSBE reactor cell., Special shielding problems

‘appear 0 be unavoidable because of the presence of large amounts

of highly radioactive salts Circulating outside the reactor.
Although development of control rods or other means of _ |

reactiv1ty control is not a part of the dinstrument development

program, the requlrements for nuclear instrumentation will be strongly

affected by the de51gn and performance characteristics of the

‘reactivity control device used.; To insure that satisfactory overall

system performance is obtalned With minimum 1nterface equipment

these two - prOgrams must be strongly coordinated.

 

 
Process Instrument Components for Direct Application
: rto Molten-Salt Loops ‘

'Reliable,'accurate, and reasonably priced sensors to measure
flows, pressures, levels, weights and temperatures of molten salt in
pipes, tubes, and tanks will be required for the MSBE. |

Several developmental instruments have been in use on the
MSRE with varying degrees of success. Performancg of these ihsfiruments
has been encouraging; however, in some cases there is need for further
development to obtain improvements in perfbfmancé, redfiction‘in'cost,
or both. In other cases, satisfactory instrumentation is either not
presently available or the type of instrumentation used on the MSRE
would not be satisfactory for MSBE service. In particular, the use
of'the furnace type heating presently planned for MSBE reactor and
drain cells would preclude the use of some devices and technigues
that were used successfully on the MSRE. Also, the electrical
conductivity of the MSBE salts(l) will have a significant effect on
the type of primary sensing elements that can be used on the'heated
salt systems. An order of magnitude decrease in the conductivity
could preclude the use of some devices presently in use on the MSRE.
Conversely a significant increase in conductivity would result in
better performance of existing devices and possibly permit the use -
of techniques (such as magnetic flowmeter) that could not be used
in the MSRE. _

It is expected that many of the problems in instrumenting the
M3BE reactor systems will be common to those encountered in instrumenting
the MSBE chemical processing plant. To avoid dufilication of effort,
the development of instrumentation for the reactor system will be
coordinated with the development of instrumentation for the chemical
processing plant. 'Techniques and instrumentation used for measurement’
of process variables in MSRE molten salt loops and areas where additional
development may be required for the MSBE are discussed in the following

 

O

I

paragraphs.

(l) anductivities of MSBE salts are not presently known.
 

 

 

cfic )

”n

»

"

»

- Flow Measurement

 

Theffldwrate of molten salt in the MSRE coolant salt system is-
measured by means of a venturi meter section. The venturi operates
at system‘temperature and the differential pressure developed in the

venturi is measured by a high~temperature, NaK-filled differential

- pressure transmitter.--Except‘for some initial trouble with one of

the two NeK-filled D/P transmitters installed, performance of this
system has been adequate and this type of system would probably be
acceptable for similar.service on the MSBE.

| Fuel salt flow is not measured. in the MSRE becauSe there is no
acceptable flowmeter ‘available for this service at this time. The
system used for MSRE coolant salt flow measurements was not acceptable
for fuel- salt flow measurement because of the possibility of release
of NaK into the fuel bearing salt with the resultant possibility of
uranium precipitation.. This. consideration might not apply to the. MSBE
if the volume of NaX is very small in comparison with the volume of
salt; however, even if this obJection were removed additional

development would be required to use the venturi, NaK-filled D/P

- transmitter system,for measurement of fuel salt flow in either. the

MSRE or the MSBE. The problems involved are common to those. discussed
under Differential Pressure Measurement.

If measurement of flow rates of fuel salt in the MSBE is required |
a suitable flowmeter must be developed for this service. As mentioned
preViously, the present technique might be adopted if the possibility -
of a;NaK.releaseninto,the salthcan be-tolerated. .Ultrasonic techniqueS“

offer promise for molten'salttmeasurement.,rThe Dynasonios Corporation

is presently marketing an ultrasonic flowmeter which is. capable of .

measuring flows in lines from one to six inches diameter, Since this

instrument has piezo-electric transducers mounted,on‘the transmitter

'body, it is limited to process temperatures below 500°F and is probably
,'not suitable for use in high level nuclear radiation environment .

": However, a good possibility exists that these limitations can be

eliminated by using the force insensitive mount techniques (developed
by Aeroprojects Inc,, and used in the MSRE fuel storage tank level

probe) to permit.the-heat'and radiation-sensitive components to be
 

8

..%located outside the‘reactor containment and shielding. The resultant

flowmeter would be capable of operating at temperatures in excess of
1300°F and would be ccmpatible with reactor environmental conQitions
~and contairnment requirements. It would be of all welded construction
and would not require electrical or piping penetration of the meter
body or of the containment vessel. o |

Other, less promising devices that should be considered for
measurement of molten-salt flow are the turbine and magnetic type .
flowmeters. Both of these flowmeter types can be constructed for high
temperatfire service and both have been used in liquid metal systems
with varying degrees of snceess;'howéver, neither haswbeen snccessfulhy
- used in molten-salt service. ' | ,

A turbine type flowmeter was developed for the ANP program and
operated satisfactorily:at 1600°F for a short period'before'failure.i ,
The msjor problem in the development of a flowmeter of this type is.the
high temperafiure physical properties of the turbine blade and bearing
materials. Although the ANP development effort was not successful, it
is possible that the use of improved materials now availeble, together
fiith the lower MSEE temperatures might permit development of a flowmeter
for MSBE service. |

Magnetic flowmeters have been used extensively at high temperatures
(1600°F) in liquid metal system and lower temperatures for measurement
of a variety of fluid flows. This type of flowmeter could not be used
in the MSRE because of consideration of conteinment, materials |
compatibility, and 'molten-salt conductivity. Containment and material
compatibility considerations prevented the use of electrical lead- |
- through penetrations of the meter. body such as used in conventional
magnetic flowmeter construction and the relatively poor (l mho/cm)
conductivity of the molten-salt prevented_measurement of signal voltage
. &t the outside surface of the meter body as is done in liquid netal
flovmeters. If the conductivities of MSEE salts were found to be
greater than that of the MSRE salt by a fsctor of 1000 or more,
liquid-metal magnetic-flowmeter techniques could be used.for_messurement
of molten~salt flow. Development of satisfactory electrical lead-
through penetrations would permit development of msgnetic flowmeters -

 

 
 

 

7]

n

"

)y

for moltén-saltfservice-regardless'ofjsalt:conductivity.'

' Salt Inventory

Inventory of the amount of molten salt 1n ‘the MSRE drain and
storage tanks is obtained by means of pneumatic weighing systems'

r; manufactured by the A. H. Emery Company., These systems use diaphragm
.type weigh cell and null balance principles and except for some

spec1al piplng connections that permit operation under conditions of

_'_varying sub-atmospheric environment pressure, are standard commerc1al

| items. This type of system.is 1nherently radiation resistant, is not
'sensitive to the effects of varying ambient pressure, and is relatively
;insensitive to ambient temperature variations below 150 F.: The basic

principle of operation and method of installation are such that the
sens1tivity and span calibration can be checked durlng reactor operations

- at 8 control panel located outside the containment and the biologlcal

‘shield. It has the disadvantage of requiring a number of pneumatic

tubing penetrations of the containment which must be guarded by safety

fblock valves., Except for some difficulties with zero drift and with

some peripheral equipment, the MSRE systems have performed acceptably

_(Performance of similar systems used in the HRE-2 was also acceptable)

The zero shifts are thought to be caused by changes 1n pipe loading

~ rather than drift of the weigh cell system.

Although the accuracy of weighing systems may be limlted by the |
Zero shift effects produced by pipe loading, the weigh system approach

f’appears to offer the best possibility for accurate determination of MSBE
- salt inventory in those applications Where env1ronmental conditions -
i”and total tank weights are such as to permit its. use.‘ Tank 1nventories
:could also ‘be determined by measurement of level, however, this approach

.-?s?requires the use of corrections for tank geometry and salt density.

Also, as discussed below, additional level system development would

.'be required unless measurements of tank 1nventory are made under statlc
’i_pressure conditions and unless a continuous gas purge into the tanks

:;“can be tolerated.

Tt is possible that a combination of 1evel and weight measurements

_w1ll be required to Obtain a total salt inventory Present 1nd1cationsr
 

- <10-

are that the tare and live loasds of the main MSEE drain"tanks,*and" -
" the ambient temperatures in the cells in which these tanks are located
will be such as to preclude direct use of the type of system used in
MSRE for measurement of salt inventories in these tanks. e
The pneumatic weigh cells used in the MSRE are the. 1argest that
Emery has produced. Larger cell capacities are possible but a ..
considerable amount of re-design and developmental testing would be
required to obtain significant increases in individual cell capacity.
Although a number of "brute force" design techniques, such as beam
balance (leverage) systems of multiple cells with.mechanical averaging,
could be used to obtain large weighing capacities, conSiderations of
space and cost may preclude the use of such techniques._ Also, the
150 F maximum.temperature rating of the pneumatic weigh cells precludes
thelr use in the 1200°F ambient expected in the MSBE drain tank cells.
A number of other weighing devices are commercially available
which could be used for weighing of Large tanks but all of those _
considered to date have characteristics which preclude or seriously 7
limit thelr operation under reactor environmental conditions. One -
weigh system, offered by a Swedish company (ASEA), has considerable
promise.(e) The load cell in this system is essentially a misdesigned
transformer utilizing the magnetic anisotropy which occurs in &
magnetic material under mechanical stress, Desirable features of the
cell include its high loasd capacity, electrical output, solid state |
structure, low output impedance, low sensitivity to temperature effectsr
end high output signal. The standard model ASEA load cell is not .
suitable for extended service in high level radiation or high
temperature (1200°F)'environments, however, available information
indicates that adequate radiation resistance could be obtained by -
replacement of organic electrical insulation materials with inorganic
materials and that the maximumm operating temperature might be extended
to the point where satisfactory operation could be obtained by air
cooling the load cells. However, extensive laboratory and field. testing
should be performed on radietion-resistant high-temperature equipment
before committing the reactor svstem.design to the use of this device,
. TET__TEE_K§EA system!was seriously considered for the MSRE but a programr

to evaluate it was abandoned because of the press of time and
procurement problems rising from the "Buy America" act.
 

 

n

L

¥

n

-Another prom131ng techniqne (which would require development)
would be the use of a NaK. fllled (hydraulic) load cell. 0il and mercury

filled hydraulic systems hav1ng high load capacity and accuracy are
- commercially available. -Substitution of NeK for oil should permit

operation~of'the-primary,loadrcell at 1200°F. In this system, weight

~ would. be converted.to“NaKlpressure'which-would be transmitted via a

“capillary tube to a transducer located in a more hospitable enviromment.

The principle in this case would be similar to that of the NaK-filled
pressure and differential pressure cells discussed below.

In some caseSfitflmay_be'pQSSible to.ease the requirements on the
basic weighingpSystem by:arsuitable design of the reactor system. One
possible, but notparticularlypromiSing;approachiwould be to bring
suspension rods through?containment~(With;bellowsrseals)Jto‘weigh cells
located above the biological;shieldingp This approach would permit the
use of variety of'basic:weigning'systems but would introduce serious
structural, operational, and maintenanceprohlems. Another'more promising
approach would involve:weighing'of,a side tank rather than the entire
tank. This'approachlwould-ease_theproblem of load cell capacity but
not the ambient temperature problem. Possible problems with this
approach includes'rem0valrof€afterheat from the side tenk and elimination
of extraneous loadszproducednby stresses in piping connecting the side
tank to the main tank. In any'Case,'it'is'anticipated that considerable

. coordination of des1gn and development will be required to obtaln

accurate 1nventory measurements.,rgf‘“

"Temperature Measurement -

 The temperature of heated pipes and vessels 1n'the MSRE are measured

"by means of mineral-insulated Inconel-sheathed Chromel—Alumel “thermo-

couples.: Results of developmental tests and observation of field

'jperformance of this type of thermocouple indicate that an initial (hot
-junction) measurement accuracy of +2° F and a long term drlft rate of |

filess than o F/year can “be obtained at operating temperatures in the
lrange of 0—1300 F if couples are selected and calibrated and 1f |

‘attention is paid to details during de51gn, fabrication, and installation.

Errors of +8° F under static and protected condltions may result 1f a
 

12

standard grade of wire is used,and normal installation practices are
followed and errors can be even greater if the conples are exposed
to moving air or are directly exposed to electrical heaters.

Since the MSBE temperatures will not be significantly greater
than those encountered in the'MSRE, the materials and basic techniques
used for measurement of MSRErtemperatures should be adequate for MSBE
installations. The use of the furnace concept for heating of reactor a

and drain cells will, however, introduce‘problems which will necessitate

further development of in-cell 1ead~W1re, disconnects, and containment .
penetration seals. - Also, if accuracies greater than those obtained

in the MSRE are needed, additional development will be required.

' Further development effort could also be profitably applied in the
areas of thermocouple attachment, investigation of the feasibility "
and desirability of'using infra-red photography or'radiation pyrometry
techniques, and the measurement of small differential temperatures at
elevated temperatures.

The thermocouple attachment techniques used on the MSRE would
probably be satisfactory for most and possibly for all MSBE thermo-
couple attachments; however, the methods used on the MSRE are time
consuming and costly, and small improvements in technigues could yield -
large dividends where large numbers of couples are required. (Over
1000 thermocouples were installed on the MSRE.) L

Multiconductor, glass-insulated, silicone-impregnated, copper- -
sheathed thermocouple cables are used in the MSRE between the in-cell
disconnects and the out-cf-cell Junction boxes. Thesercables'penetrater

the containment and are sealed. This sealing intrdduced problems |

.because of the effort involved and because of pressure buildup produced _

by outgassing of the silicone-insulating materials. This type- of wirlng
will not be useable in the MSBE if the furnace concept is used for o
heating the reactor and drain tank cells, The expected 1200°F ambient

in these cells will require the use of inorganic insulated leadwire.. _

Furnace heating of the cells will also reqnire protective sheathing
of all 1n-cell thermocouple wiring and the development of disconnect
devices which are compatlble with the furnace atmosphere and w1th remote

maintenance requirements. Multiconductor, mineral-insulated sheathed-

O,
 

 

r

*)

”n

13

thermocouple-cableassemblieswereconsidered for in-cell leadwire and

_containment penetration service at the start of MSRE design, but this

approach was abandoned because of high costs and the difficulty of
fabricating satisfactory end seals. The cost of this type cable is-

now reduced and their use should be re-evaluated.' The major problem

1involved in the use of mineralwinsulated thermocouple cable for

Ppenetration of containments would be the development of satisfactory

methods of sealing the ends of the cable. Although the development
and fabrication of end seals_and,techniques,for~installation'of
multiconductor cable will be much more difficult than would be the
case if thermocouples wereibrought'out through'individual penetrations,
it is expected\that the reduction in the numberwof penetrations, which
would be obtained by using multiconductor cable Wlll more than justify
the additional cost. . However, other conszderatlons such as maintenance
requirements could necessitate the use of individual penetrations.»
Therefore, both approaches mnst be considered initially. |

- The use of infra-red photographic and radiation pyrometry
techniques offers the poss1bility of mapping of temperature contours
on large exposed surfaces (such as the MSBE reactor vessel) On-line
viewing of temperature distribution might be obtained by v1ewing a
heated system with a closedvcircuit television camera equipped with
an infra-red fllter. A more accurate determination of temperature
profile might be obtained by mechanically maneuvering a radiation
pyrometer to produce a scan pattern similar to the raster produced

ion a televismon screen.. Since the feasibility of us1ng these |
_,techniques would be strongly dependent on the physical geometry of }

the system viewed and of the surrounding area, investigation of

. feasibility and development of equipment and techniques should be
- performed before the start of design of the system on the fac111ty. ,

The problem of measurement of small differences in large,_

'temperatures has not been satisfactorily solved. The accuracies

obtainable by using serles-opposed (bucking) thermocouples, and

rextreme care in de31gn and 1nstallation, are barely adequate for

MSRE purposes and might be inadequate for the MSBE. There is room
for a cons1derab1e amount of orlginal work in this field. Several

-
 

14

commercially available, high-temperature resistance elements-were’i'
tested to determine the feaslibility of using such elements for -
precise measurements at high temperatures. Results were, in general,
disappointing but the performance of one unit was sufficiently
encouraging to lend support to the belief.that a suitable'resistance
element could be developed. - | ) |

A possibility exists that differential temperatures could be . -
'measured by use of ultra-sonic techniques. One ccmpany (Aeroprojects,
Inc.) is presently investigating the feasibility of the ultrasonic

measurement of absolute temperatures. Representatives of that company®-

have expressed the opinion that accurate-differential temperature._

measurements could be made by means of ultrasonic devices.

Ievel Measurement

 Several methods have been successfully used for single point and
continuous measurement of molten salt level in the MSRE. All those
methods could be used in the MSBE under similar conditiOns, hoWever,'
all have limitations which would preclude their use under certain |
conditions.

Continuous measurements of molten salt level in the MSRE coolant

system pump bowls are made by'means'ofl"bubbler" type (dip tube) and

float type level systems. Continuous measurement of molten salt
level in the fuel pump bowl is also made by means of a "bubbler"

type system and a future pump installation will include a float 1evel\

transmitter. | |
Two-level, single-point measurements of molten salt level in,the
MSRE fuel and coolant system drain tank are made by means of conduc-

tivity type'level probes. The information:obtained-from.these‘probes |

is used to check the performance and calibration of the tank weighing”
systems. The probe signale operate lamps (or other binary dev1ces)
which 1nd1cate whether the level is above or below two pre-selected
points. '

An ultra-sonic level probe is used for single-p01nt measurement

of level in the fuel storage tank. Except that the ultrasonlc probe o

presently installed is a one-level device, the information obtained_*

(3) Personal communication, Mr, Kartluke and Dr. Boyd to R. L. Mobre;.

O.

O
 

 

»

#

)

”

.

15 |

 from the ultrasonic probe;is-identical-to-thst_obtsined=from the

conductivityftypegproberend.isiuseddfor the same purpose.

' All'of_the systems-used“for measurement~of molten-salt level in
the MSRE were specially developed for the service and further development
or re-design would ‘be reqnired for other. eppliCations.'-The "bubbler"
system is basicnlly the simplest and most versatile method of measuring-
molten salt level under relatively static conditions of level and

cover gas pressure._ This type of system can be used for narrow or

:”w1de ranges and the vessel modifications required to install the system .

are simple and inexpensive._ However, since the . "bubbler” system
performance is dependent on maintainlng a gas purge flow through the
dip tube, this system canionly be used.where the purge can be tolerated.
Also, the'response‘cherscteriStics:of“this type system are dependent

on the purge flow rate which, in turn, is. dependent on supply pressure,

cover gas pressure, and other factors. In general, the low purge rate

. required for accurate measurement is not compatible with requirements
for fast response and fast variations in cover gas pressure, guch. as.

, can.occur in the drain tanks and pump bowl during fllling and draining

operations,_can render thewsystem_inoperstive unless accompanied by

corresponding changes'in'purge”flou rate. The desirability'of using -

1Apresently developed "bubbler" techniques for measurement of levels
in systems containing highly radioactive liquids and gases 1is considerably

reduced by the necessity of providing adequate means to detect and

_prevent the release of activity through the purge line. Development

:of a system wherein the purge gas would be recycled within primary
_"and secondary containment would greetly extend the usefulness of thisia
' _type of system. ' : .

i The float type level system offers the best method of continuous

-fmeasurement of molten salt levels over narrow ranges. ‘This device is

completely contained, has fast response, and. requires only electricel :

"77penetrations of secondary containment.’ Present designs are limited
to measurement spans 1ess then ten inches. The span can be extended.

-'3but this type of device is basically more suited to low span than to

high span measurements.
 

16

Q.

The conductivity type level probe has performed well in MSRE
service. With the possible exception of redesign of the tank . | |
penetration to improve containment the present probe design could be
- easily adapted to installations in MSBE tanks. . The MSRE conductivity
probe has the'disadvantagerof»having_(and requiring) thin walls in the -
 tubes extending into the tank. This:typefof-device‘is therefore not -

suitable fbr use in corrosive environments such'aSZthatiin the'MSRE-’
fuel storage tank. | | | .

- The output signal obtained from the MSRE conductivity probesf”
was much greater than had been expected. Since the output - signel
is much grester than required the possibility exists that & more -
rugged and corrosion resistant'single‘point_probe'cOuld be developed
(vy ificreasing.tube wall thickness) and that a continuous type - |
conductivity probe (similar to the "I or "J" tubes used in liduid -
metal eysteme) could be developed. An incresase in the electrical
conductivity of the salt (over that of the MSRE salt) would result
in either better 51ngle point probe design or improved prospects of
using the conductivity probes for continuous level measurement. -
Conversely a significant decrease in electrical conductivity'couhi:
prevent the use of the conductivity probe for either single point
‘or continuous measurement. o ,

Except for some problems with oscillator frequency drift which
caused the instrument to become inoperative, the performance of the
Aeroprojects ultrasonic level probe has been dependable and accurate.
Some additional development will-be required in the area of main
chassis electronics and packaging before this device can be coneidered"l
to be a reactor grade instrument; however, the remaining problems with -
this instrument appear to be amefiable'to solution by routine re-design
and development testing. - | ' o - |

. The Aeroprojects single-point-ultrasonic level probe is consider-
ably more rugged and corrosion resistant than the conductivity type
probe and, when the remaining problems are'solved,”should be the -
preferred method of single-point level measurement in 1nstallationsi‘i- |
where the technique is applicable. _ ol e T &fij

Aeroprojects has done some work on development of a continuous : .

ultrasonic level probe. Such a level probe would be very useful in
 

 

nl C ¥

v’

17
molten salt systems (and'in-aqueous and liquid metal_systems) and
development of this device should be supported.

An obvious/method:of‘leVel measurement which should be considered -

for possible use in the MBBE is the direct measurement of level by the

‘differential (head) pressurejmethod., This method was not used in the

MSRE because aISuitahle_device for measuring differential pressure was .

not available. . The problemsiassociated with measurement of differential
pressures in molten. salt systems are discussed below. .
With the possible exception of the differential (head) pressure

‘method, all of the methods discussed above are compatible with the

concept of furnace heating of teactor and drain tank cells. The float
type transmitter is especially attractive since the transformer and

other parts of this device not only can but are required to operate

at (or near) system temperature.— The conductivity and ultrasonic

probes might poss1bly be used in a furnace atmosphere in their present
form; however, some additional development work on leadwire, penetrations,

and disconnects will probably be desirable.

Pressure Measurement

 

Pressures in the MSRE molten salt loops are obtained indirectly
by measuring the pressures in gas purge or supply lines ‘connected to
drain tank and pump bowl gas spaces. The pressures measured in the
MSRE are, therefore, “those of the cover gas and ‘not of the salt.

“The - techniques used for measurement of cover gas pressures in the

MSRE may or may not be’ applicable to 1arger reactors of different

'de51gn In any event there 1s no pressure measuring device available
) 'that is suitable in its: present form.for directly measuring ‘pressures
. f(or differential pressures) of the fuel salt., Also. the system used
,,could be strongly effected by the method of heating of. 1ines and |
-'hvessels. The use of oven type heating would impose. the additional |
'1irestriction (over those encountered in the MSRE) that all components
- of the pressure transmltters be located outside the heated zone or be.

;ru.capable of operating at the temperatures ‘present in the oven._ o

| Measurement of cover gas pressure in the MSBE would require
little additional development if the 1ndirect methods cah be used as
in the MSRE. However, considerable additional development would be

~
 

18

‘required for measuring'salt'pressures since part or all of therpressure :
measuring device would have to be located within containment and |
biological shielding in close proximity- to the pressure tap and
would have to be capable of withstanding the environmental effects B

of temperature, radiation and varying ambient pressure.'..~

. The pressure transmitters presently used in the MSRE or equivalent -

devices constructed in accordance with the same'specification7shou1d

 be adequate for measurement of cover gas pressure by the indirect method .

- NaK-filled pressure transmitters offer the best prospects for direct |
measurement of cover gas or fluld pressures 1n.molten salt systems. |
Use of this type of device would require acceptance of the p0551b111ty
of release of a small quantity of NeK into the system.in the’ event of
a rupture of the seal diaphregm¢~'Also,~addit1onal development would
be required to obtain adequate containment if the transmitting'element
is located outside of the reactor secondary containment or to reduce
the environment effects of temperature, radiation;'and-varying ambient'
pressure to acceptable levels if the entire transmitter is located
inside reactor secondary containment and biological shielding.
Another, less promising, approach is offered by the possibility
of adapting a thermionic diode type of pressure transmitter, which is
presently being developed for use in high temperature 11quid.metal
systems, for use in molten salt systems. The progress of this. and
other developments in liquid metal system instrumentatlon should be
followed to determine whether they can be applied to molten salt

systems..

Differential Pressure Measurement

 

With the possible exception of the measurement of pressure drop
across. the charcoal beds, all differential pressure measurements made -
in the MSRE molten salt system were for the purpose of measuring flows
of gas or molten salt.‘.No direct differential pressure measurements
were made in‘the'liQuid portion'of-theLMSRE fuel salt system. ‘The
performance of the differential pressure ‘transmitters for measuring
gas flow was satisfactory and no additional development is ‘needed in

this area. As mentioned above, -some difficulties were experienced

4

Q.
 

 

")

.y

A o

with one ofrthe,tWOVNaKjfilled differential pressure cells used for

r_measuring salt’flowiduring initial'operations'of the MSRE. The
3performance of the other transmitter originally installed and of the

o replacement for the defective transmitter has been satisfactory and

this type transmitter would probably be acceptable for similar service
on the MSBE. While two of the three NeK filled D/P transmitters

| purchased for the MSRE have performed satisfactorily, the difficulties

experiencedrwith the third transmitter and with a spare recently
purchased (as well as with procurement on all four) indicate that an

alternate source"of supply should-be-developed. A NeK filled transmitter

z_is now being offered by Barton Instrument Corporation but this in-
strument is not reactor grade and performance has not been tested at
* ORNL.

| The problems associated w1th meking direct measurement .of
differentlal pressures in the liquid-filled portion of the MSBE fuel
salt are 51milar to those discussed above for measuring pressure under
s1milar conditions. The main dlfferences in the two problems are
that the differential pressure transmitter is not affected by variations
in ambient pregsure ‘and usually 1s required to measure much smaller
variations in pressure. In applications such as level measurements where
relatively small spans might be required the effects of variations
of_ambient temperature, process'temperature and process pressure on

the span and zero of the transmitter are of prime importance and could

,.be the. deciding factor in determining the suitability of NeK-filled -

differential-pressure transmltters for a given application. For -

. example, determination of the 1nventory of MSRE drain tanks hy >

:';.measurement of level using the NeK-filled differential pressure
_:'transmitters presently available would not be a suitable substitute
"1for the welgh system now in use.. However, if the performance of the
"'transmitters were. sufficiently improved by further development, level

| ',measurement might be considered an acceptable alternative to. weighing.
L ;_:As_in the_case;of the.pressure transmitters, another, but less -

7_ppromising,_approachfto the;prohlem of direct measurement of differ-

ential pressures in molten?salt?systems is offered by the thermionic

"diodentype elements presently under development for the liquid metals

program.
 

20

WO

Process Instruments to Opersate Auxiliary Sub-Systems-"

These are the instruments used to measure and control. auxiliaries
such as process and cooling water, cooling and ventilating air flows,
helium.purge flows, etc. In most cases, instrumenting these systems
is old art and presents no prdblems of consequence. Exceptions arise
when system containment, remote bandling, entrainment of foreign matter,
intense radiation, extreme temperatures, and very low flow rates enter
into the design. A typical example in the MSRE is the off-gas systemf"
discussed under Gas-System Control Valves. The experimental pfimp’
loops and ETU which will be built end operated in the near future will
require instruments. This instrUmentation, regardless of how‘conventional
it is, should be designed to meet the criteria and conditions anticipated
for the MSBE. The improved and tested instruments should be used on
the MSBE and the cost of the improvement regarded as development'expense;

Health Physics Radiation.Monitoring _
The MSRE is equipped with three (3) general types of radiation
monitoring equipment. These are: ' ' o
1. Monitrons to detect sources of gamma radiation.
2. Constant air monitors to detect airborne particulate sources
of gamma and beta radistion. ‘
- 3. Portable beta-gamma monitors.

A total of seven monitrons arerused with six of these being inter-:
connected o that if any two monitrons indicate & high radiation level,
a building evacuation alarm is produced. Four of the seven constant
air monitors are also used in a similar coincidence arrangement. The
evacuation alarm originating from this coincidence circuitry is- also |
transmitted to ORNL's Laboratory Emergency Control Center where
appropriate emergency actions are initiated without waiting for further
instructions from the alarmed s1te. ' - S

Several beta-gamme monitors are used to supplement the coverage:
ofithe permanent, fixed location monitors in areas less likely to
experience excessive radiation levels. The beta-gamms monitors produce
a local alarm signal only and are not_connectedto the general evacuation : -ifii .

. alarm systen. . o - '__,:__ “'"Q; o S
 

 

5 [ v

w)

21

This eqpipment and the way“it is used represent standard design
practices at ORNL _The. c01ncidence and switching circuitry is assembled
fromwmodules developed specifically for health-physics monitoring

installations. These modules areAin general use at ORNL and may be

.'interconnected to produce a wide variety of coincidence arrangements

depending on the particular installation.” ‘ .

- The components making up the system are, for the most part,
commercially available.: No unusual criteria are anticipated for the
MSBE which cannot be met by u51ng typical and available:health—phySics

instruments, hence, no development work is foreseen._
-Steam.Plant InStrumentation'

It ‘has been pointed cut elsewhere in this report that the steam
plant which extracts power from.the MSBE will-be modeled after TVA's
Bull Run system. Such a plant exemplifies the prevailing trend to
power generation with high temperature, supercritical pressure steam.

Public utilities, TVA, and the instrument manufecturers all have.
strong incentives to promotefand develop the instruments and techniques
required to operate supercritical steam systems. ORNL is reasonably

confident that, when the time arrives to deSign the system and to

specify the steam power plant instruments, this technology will be

_ approaching standard practice type of application engineering.

It is fortuitous that IVA's Bull Run plant is Virtually contiguous
to ORNL and that they must, of necessity, develop a satisfactory plant.

ii'We expect to draw heaVily on.TVA experience augmented.by developments
| T"elsewhere should they appear worthwhile._

It would be unwise at this time for us to dilute our efforts else-'

'where by engaging in development engineering for steam system in-

-;.strumentation. i

The heat exchange interface between the steam plant and the

;1i1molten salt may prov1de problems It is too early to draw firm cone-

'”cluSions but it cannot be assumed that instrument deSign for this: .

portion of the MSBE will be routine and conventional DeSign criteria,

loop and mockup tests should disclose any need to test and evaluate

critical components.
 

 

22

" Computer Control and Data Logging

. MSRE instrumentation includes a digital ‘date logger and computer
which is supplementing the more conventional instruments making up the
primary measurement and control system. '

'The data logger collects and stores information from up to 350
field-mounted instruments of all types. These instruments are scanned
at 5-seconiintervals and selected groups of the numerical readings are
automatically printed on log sheets in the appropriate engineering
units once every hour, 4 hours, and 8 hours, or at any time, should
the operator so request. Out—ofulimit alarms are actuated if any one
of the 350 data points goes beyond acceptable values in either the
' high or low direction. If the out-of-limit ‘point or points are critical
for good operation, the logger will scan a selected group of 6k data
points continuously at a repetition rate of five times per second until
the reactor system is again within its proper limits.

" The system is programmed to compute a reactivity balance once every
hour or upon request of the operator.

Other typical programs now in use are:

1. Heat balance, - |

2. Cell average temperature,

3. Salt inventory,

4. Average reactor fuel and coolant radiator outlet temperature,

5. Nuclear average temperature,. |

6. Reactor cellrevacuation volunme,

7. AT at fuel inlet and lower head,

8. Fuel pump AT and minor thermal cycles.

The machine is 8lso programmed as a temperature controller in an
experimental creep stress facility (the Surveillance Test'Rig)mwhich o
is installed at the reactor site. This facility simulates andlfollo#s'
reactor temperatures and temperature gradients in the environment
surrounding the creep test specimen. |

The potential application of a d1gita1 computer in the reactor |
control system of the MSBE will be reviewed in detail. The most |
important basic questions which will require early study are.

 
 

 

wrh

'23l

1) To what'extent~Shouldfthe'system be ‘controlled by the computer,
2)'lWhat'will be the'required reliability of the digital system, and
-3) The degree and type of backwup control to be used in the event
the primary system is renderéd 1noperative. - o

| Detailed answers to the first question must be deferred until

.'preliminary reactor system: design has been completed, however, one

basic objective may_he stated. - Control functions which have been

‘satisfactorily handled by conventional analog control devices will )

not be converted to computer control unless there is strong economic
and technical incentive and clear ‘evidence of technical feasibility.
Some pre- speciflcation testing of- the newer digital controllers would
be advantageous. - o ' '

The assessment of thefrequired reliability will also depend on what
exact control functions'are7assigned to the computer. It is likely
that high reliability will be required and consequently schemes should
be 1nvestigated which will result in a system.With higher reliability
than can be obtained with a single computer._ An example is a two’
computer system in which the second machine is normally occupied with

logging and computation of a non—vital nature and could be called on

- a priority basis to serve as a controller when the. main computer fails

or is down for maintenance.:fiQ;"'

~ The ba31c approach to these questions and the detailed de51gn
problems in general will be much better founded within the next few
months. - By mid-year (FY-68) the ‘HFIR control computer will be installed
and the results of the: preliminary control experlments w1ll be available.

oo In addition about six months of computer and. interface operating

' .experience will ‘have been logged._»

Computer control of the testing ‘process, along with computer -

,'techniques to analyze the system loops based on the ever present process
':;nOise level are attractive possibilities which should be explored
) 'iSuch techniques, if successful, would pay off well in terms of improved
'?—reliability. | o L

The application of the Bunker-Ramo 3#0 computer on the MSRE has

| demonstrated the desirability of prov1ding this type of on-line
 

 

24

date acquisition and computation system to assist with the operatiOn'
and analysis.of such large complicated engineering experiments‘,To_aid
in further analysia of the benefita of on-line computers, several
experiments and demonstrations are planned to gather additional infor-
mation and verify some theories on the use of this type of.computer for
process control of reactor system auxiliary devices_such¢as:heaters,
etc. | | | o T | C e s
- The shakedown problems experienced while putting the MSRE computer
into operation suggest several areas which should be investigated prior
to the application of a similar computer system on the MSBE. These
potential trouble areas are all in the znalog signal handling-part of
the system and have to do with 1) noise rejection and filtering of the
input signals, 2) analog input signal commutation, andu3);generation
of‘analog,output_signals._ Each of these problems is primarily economic
in nature and though the MSRE computer does an adequate job in these
three areas, camputer system costs on a larger system such as would be
required for the MSBE would be very high if more economical ways to do
thése Jobs are not found. For instance, an approximate cost to handle
input signals on the MSRE computer is more than $100 per process variable.
A conservative estimate of the number of signals to be brought into &
computer on the MSBE would be 2000, which emphasizes the need for finding
ways to reduce the input signal handling costs. : | :

The study of possible use of multiple small computers instead of
one large computer should also be done with the actual-application of one
of the smallest computers to a mock-up or test stand in the early part
of the MSEE program. In this way, it may be possible to avoid over-
- specifying or under-specifying the MSBE eomputer; The actual experience
of applying such & computer is_needed i . o o

BerylliumfiMonitoring :
| The two beryllium.monitors presently in use at the MSBE are intended
to detect airborne beryllium Prom systems not sufficiently radioactive :
to permit detection by activity monitoring. One is an air sampling
.system which requires laboratory analysis, while the other 1s a nearly
continuous monitor in the radistor air stack.

O .
 

-y

L

;General

25

The air sampler uses,a,Suctionrpump to draw. air through filter
papersfllocatedfat fifteenfstrategically placed.collecting'stations;
After a suitable-intervalJfior~Sample_collection, the filter papers -
are spectrographicallymanalyzedyfor beryllium”content.

The exhaist air from the radiator stack is monitored and the .
results recorded almost continuously by an on-line:spectrograph..

This instrument, atrlal/a ninute intervals, automatically strikes
an arc in a'small-bypass stream of air.pumped-from the stack and the
light output is examined photoelectrically for the characteristic line

'produced by beryllium.

Although the filter paper sampler has’ worked satisfactorily,

it is limited 1n scope in that the information is delayed. The on- -

line monitor is a developmental deVice which has. demonstrated a
principle but has not provided the required degree of reliability
for this type application.

The MSBE will most 1ikely employ nonwberyllium—bearing coolant
salt; if so, no beryllium monitor Will be needed for the reactor

system. However, the ETU, the chemical process1ng plant, and the

,fuel'processing facility will have_large amounts of_beryllium.

Consequently there exists a'needffor:a'monitor“having better
performance than .can be5obtained fromrthose presently:in use.
Development work will be continued toward achiev1ng a reliable

continuous on-line mDnltOI."” )

Component Test and Evaluation -

New devices are continuously appearing on the‘market, yet there N

'Jis no. general program for testing and evaluating these dev1ces at ,
'ORNL General Electrie, Motorola, Taylor and Minneapolis~Honeywell |

are. now offering systems which are in strong competition Wlth the

- f'Foxboro ECI system used extensively in the MSRE We must know more-
"about these systems if we are to evaluate bids properly and assure
'7lourselves that the components and instruments selected are the best

-choice for the Job.
 

26

Electrical Control Circuit Components .- = ,

~ Performance of all reactor systems is strongly dependent .on the
performance characteristics of all components of electrical control
and alarm'circuit. Reliability is of particular importance since’
the large number of control circuit  components dictates that the |
reliability of each.individual component be extremely.high.in order:
to obtain an acceptable overall system reliability. Poor system -
reliability will result in frequent shutdowns which in turn will..
‘result in low plant availability. To insure that adequate control.
component performance will be obtained a program.of test and evaluation
of control components 1s needed, “The scope of this program would
include testing and/or,evaluation\of'such;comPOnents as relays,.cOntrol :
svitches, wire terminals, connectors,'indicator lamps; and annunciators.

Helium Flow-Elements-

 

In the MSRE, matrix type flow elements are used for measurement
of gas flows and capillary tubes are used for restrictors. ‘The :
performance of these devices has been very good; however, the procurement
or fabrication of the devices required a great deal of engineering .
labor. A number of laminar flow elements are available commercially:
for a reasonably wide range of gas flows. These devices shouldhbe _
investigated to determine whether'they would be an acceptable substitute;
for the devices now in use and whether they offer any significant
advanteges. '

Gas-System ControlVValves"

Helium.purge flow and cover-gas pressures in the MSRE are
controlled by means of throttling valves. Due to the very low flows_
and pressure drops involved the clearances dn these'valves are ; |
extremely small. These small clearances together with 8 lubricant |
problem were the apparent causes of trim galling which caused failures
of a number of valves in MSRE service. The lubricant problem results
‘from the need for cleanliness and from the lack of any lubricant in.
the dry helium.which flows through the valves. Improvements have been

made in valve assembly procedures which have resulted in a reduction B

 
 

 

-4

-y

"_27*" }' "

_in the failure rate, but additional effort should be spent on developing

better valves for. control of small helium flows.
MSRE operating experience with the off-gas system revealed a
possible future need for the suitable means to control very low flows

"of helium contaminated w1th selid particulate material and w1th hydro-

carbons, in all phases, originating with the salt pump lubricant. It

is a foregone conclusion that qonventional valves and flow elements ,

will clog and stick in this serv1ce.- Unless it is certain that the

_ MSBE will not generate similar flow and pressure c0ntrol problems, a
B -solution is required Three lines of attack are suggested First

(and best), eliminate the problem at its source, second, “develop

--valves which will do the Job, third develop other, and prebably -

unorthodox, methods of handling small flows of dirty helium
Development work on these prohlems will be done in cooperation With

'the program of development and testing the off-gas system..'“

Temperature Scanners

A scanning system was developed for use in scanning temperatures
on heated pipes and vessels in the MSRE, displaying a profile of the

temperature on a cathode ray oscilloscope and producing an alarm 1f

»any temperature is outside the pre-set limits. Similar or equivalent

systems will undoubtedly be de31rable for the MSBE
The performance of the MSHE system has heen very satisfactory

~ The signals are scanned by a mercury jet commntator’switch4 Since

the commutator switch used in the MSRE is no longer commercially

'available, another source of supply of SW1tches or an alternative '_
~ scanning device must he:developed if this type system.is used on the -
 MSEE,. The use of reed relay or transistor switches for this service

”l?toffers promise and should be investigated. The use of solid state
?j.;commutators or multiplexers is increasing. It is 1ike1y that all

l, the components required to assemhle a scanning system'will be .
“u”commerc1ally available 1n the near future. In developing a suitable

'-gscanning system for the MSBE, the test loops and mockups of the reactor

system can be used advantageously
 

 

28

As an alternative to the MSRE type scanning system, or an
improvement, the feasibility of using a computer data-logger system
forISCahning temperatures'should be investigated. Comparison of - -
the two methods should be based on conslderations of cost and -

operational requirements.

Containment Penetration Seals | o | o
_' Electrical instrument signal penetfations in the MSRE were”made .
by means of mineral-insulated cables; by sheathed, glass-insulated
silicone-lmpregnated multiconductor cables with soldered hermetic
seals inside the cell and with organic seals outside the cell; and
by hermetically sealed_connectors velded to the containment vessels.
Labor costs to install these seals were high and their perforfiance |
has been marginal; The techniques used on the MSRE could be dsed'on
the.MBBE if envirommental conditions were similar;rhowever, development
of new techniques would probably be required if large systems were
heated in an oven rather than by heaters attached to the equipment.
The amount of effort required to develop satisfactory seals for
contained, oven-heated, areas Wlll be strongly dependent on the
design of the containment vessel and the oven. Extensive seal
development will be required unless the design is such that the,'
oven wall is not the conteinment wall (i.e. unless seals can be
located in a cool area). As s minimum, a different type of seal
will be required for thermocouple penetrations since the type of
leadwire cable used in the MSRE is not compatible with the oven
heating concept. rIa any event, a progrem is needed to study and
develop better methods of masking electrical penetrations'into'the
containment and to evaluate hardware.

Temperature Alarm Switches | |

A number of alarfi, interlock, and control actions in the MSRE _
_are accomplished by means of bi-stable magnetic amplifler-type sw1tch _
modules that were manufactured by Electra Systems, Inc. A.con51derable
amount of trouble was experienced with these devices durlng 1n1tial

operation of the MSRE. Investigations showed that several defects

Q.

O,
 

 

"

.29

- existed in'the‘modules'and;modifications-were made in-all modules.

At this time;,the,performance of the modules is being observed

to determine-whether-alljdauses offtroubles have been eliminated.

Regardless of the results of these observations, there is need for
significant improvement'in,the'design-Of sensitive switches of this
general type; For example, a sensitive switch providing a null
balance type high impedance input with cold Junction compensation
should be made available. Since Electra. Systems does not offer the
unit anymore, we should develop other sources of supply. For critical
service, a unit constructed using the input stage of a Foxboro EMF/I
converter end components of the Foxborc ECI switch would be preferable
to the Electra Systems module,

The MSRE has dlsclosed 8- general need for improved small signal
limit switching dev1ces;of this general type,- These sw1tches should
be capable of beingyused_with,input signals in the millivolt and
milliampere region; operatesconsistently’and reliably'on small.signal

changes and be-relatively;inSensitive_to.changes in line voltage and

ambient temperature. For example, on-off limit switching in the MSRE

is desirable over a 25°F or less temperature increment at average

temperatures of 1200°F, This is only a 2% change and represents a

'severe challenge to existing methods. As a very minimum, a program

to evaluate the state of the art as it is developing is needed.
~ Process Radistion Monitoring

The rediation level of many flows which- enter or. leave the MSRE

',_containment are continuously monltored These are lines- carrying

hellum, off gas, cooling water, etc., and should & monitored line

':'jyindicate excessive radiation, the usual control system action 1s to

'_close block valves as required.ry;sa,r .

A typical arrangement calls for two or more ion chambers in a-

" lead shield around the pipe being monitored. Where system safety is
'involved this redundancy 1s carried all the way fromrthe jon chamber
' through the control system output actlon., When,on-line testing o

without system perturbation is reqpired the redundancy is two out of

three.
 

30

The signals from the ion chambers areiamplified by vacuum tube
electrometers whose outputs are used for aiarm.and controi slgnals.
The foregoing is-representative.ofStandard.radiation.monitorifig-
practice today, but it is not one which is exploiting the felatively
recent advances in solid stateielecfironics.' It has the merit that
it is relatively immune to large overload signals but the disadventege
that frequent calibration and reference'point checking is reqpired.

Based on experience at the MSRE, a substantial program to improve
process radiation monitoring methods and components is under way.

The edvent of reliable solid state current sensors has enabled the
designfof a eystem.using the pulse mode of operation of G-M tubes,
thereby eliminating the need for,freQuent reference'level-calibratidn'
and permitting long, uninterrupted cable runs to a central comtrol
area. A prototype has been built and laboratory tested. Complete
englneerlng drawings and specifications are available..-This:pulse -
mode equipment appears to be particularly attractive where reliability
is the primary consideration.but its performance must be verified

by field testing. The materials of construction of the probe,; |
however, limit the environmental temperature to approx1mately 70°

for the sensor. '

For critical epplications involving high temperatures, high
radiation levels, and wide variations in radiation levels, the ion
chamber-current mode instruments will still be used. Modernization
by conversion from vacuum tube to solid'state'electrometers and
provision for analogue to digltal output conversion is required.

The need for better ceramic to metal seals and.their appllcation to
ion chambers has been established. Better vacuum seals, coaxial
connectors and radiation resistant high temperature'cabling,methOds»-'
are required to provide economical, low maintenance, long lived,

process radiation monitoring systems for the future.

Waste Effluent Mofiitorifig

Some high level, intefmediete‘level'and.low level aqueofiS'wastes will -

be produced at the MSBE site. Standard practices will be:used'fdr‘-

0.

i l‘i.
 

 

 

 

-t

-y

31

handling these'wastes. Standard monitoring 1nstrumentation such as.

| 'that in common use at ORNL should be adequate. Gaseous effluent
- (off-gas and cell ventilation) will be disposed of through a stack.
_ Standard packages for dlsposal stacks are avallable and will be used

but the nature of" the installation wouLd seem to make the final

'development of an additlonal device necessary.

The present scheme is. to. pass a sample of stack ga.s through a

mpaper filter, a charcoal trap and & clean shielded volume in succession.

Each of the dev1ces is continuously monitored by a B-y'detector and

the filter is also contlnnously monitored for alpha activity. Thls

_'scheme separates the effluent material into a particulate component,

a volatlle component (adsorbable on charcoal) and a component which

is rare gas. In case of a burst of activ1ty,_this system furnlshes

the- flrst information as to. the source and the extent of the 1ncident.

A sample taken dlrectly from the stack is next analyzed in a lab for

- more. definite infonmation. The laboratory analysis may take an ‘hour
.or longer before this informatlon is available, however, and it is ..

highly recommended that the development of & system.using a detector
at the top of the stack and a multlchannel analyzer be completed.-

ThlS W1ll allow essentially instantaneous analysis of the effluent

8&srand is & necessary addltion to the stack monitoring equlpment
for this'project; | |

| Estimate of Cost of Development Pr_gram _ 7
An estlmate of. the cost of the development is shown in Table l.
 

 

 

Table 1. Estimate of Costs for Instrument Development
FY-1968 1969 1970 1971 1972 1973 197 1975
" (costs in thousands of dolla.rs)' | ‘
Containment penetration seals 20 25 20 10 5 2 2 2
Helium control valves 15 15 13 10 10 1 1 1
| Test and evaluation of process | - a
. instruments L0 50 45 25 20 20 20 20
Level instruments 60 85 75  3%. 10 10 10 10
Pressure and differential o i o | "

- pressure instruments 30 38 37 16 3 3 3 3 ;
Ultrasonic flowmeter 15 35 32 155 5 5 5 %
Salt inventory systems . 25 25 20 10 - 5 N 2 o 2 L 2
Temperature measuring S . o T

instruments ~ 50 60 - 50 25 12 3 3 3

| 'Temperature s_cannei' -5 .30 25 12 5 3 3 2
Nuclear instruments 35 35. 3 20 20 20 20 @ 20
Test’ and evaluation of power B ' o - B | o o -
- plant instruments | o 0 6 210 . - 10 5 5 5
Computer control and data | o L E L S -
logging systems 18 18 16 10 10 ‘10 10 10
‘Beryllium monitoring | 10 10 8 5 -:; 5 - 5 o 5, 5
- Process radiation monitoring 20 | l&_O | g_g o }_Q . | _2 5 | _‘2 '5:_2 |
: . Total 343 466 397 213 125 9% 94 93
. QO

 
 

 

 

C’

1.

.4

-

33

References

Paul R. Kasten, E. S 'Bettis, R. C. Robertson, Deeign Studies of
lOOO-Mw(E) Molten-Salt Breeder Reactors, ORNL-3996 (August 1966) .

R. B. Briggs, Summary of Objectives and a Program of Development

~ of Molten-Salt Breeder Reactors, TM-lBSl.(Jfine 1967) .

 

A
 

 

n.fi;_

1)
 

c

P

o}

 

35

' internal Distribution '_ :

-1-50. MSRE Dlrector S Office ' 99. A. Giambusso, AEC-
Rm. 325, 9204-1 | - Washington
51. R. K. Adams R 100. H. E. Goeller
52. G. M. Adamson -~ - - ' 101. W. R. Grimes -
53. R. G. Affel 102, A, G. Grindell
54. L. G. Alexander = .- 103. R. H. Guymon
55. R. F. Apple 104k. B. A. Hannaford
5%6. C. F.Baes 105, P. H, Harley
5. Jd. M. Baker =~ : 106. D. G. Harman
58. S. dJ. Ball S - 107. C. S, Harrill
59. W. P. Barthold R o ~ 108. P. N. Haubenreich
60. H. F. Bauman. S 109. F. A. Heddleson
6l. S. E. Beall - - 110, P. G. Herndon
62. M. Bender R 111. J. R. Hightower
63. E. S. Bettis .~ 112, H. W. Hoffman
64. F. F. Blankenship =~ = 113. R. W. Horton
- 65. R. E. Blanco =~ 114, T. L. Hudson
66. J. O. Blomeke - 115. W. H. Jordan
67. R. Blumberg I 116. P. R. Kasten
68. E. G. Bohlman - 117. R."J. Kedl
69. C. J. Borkowski 118. M. J. Kelly
70. G. E. Boyd @~ 119. M. T. Kelley
71. M. A. Bredig- 120. .C. R. Kennedy
72. R. B. Briges 121. T. W. Kerlin
73. H. R. Bronstein T 122. H. T. Kerr
4. G. D. Brunton - - 123. S. S. Kirslis
75. D. A. Cenonico . 12k, D.J. Knowles .
6. S. Cantor o~ 125. A, I. Krakoviak
77+ W. L. Carter . 126, J. W. Krewson -
78. G. I. Cathers R 127. C. E. Lamb
79. J. M. Chandler - - 128. J. A. Lane
80. E. L. Compere o 129. W. L. Larkin, AEC-ORO
- 81. W. H. Cook . 130. R. B. Lindauer
82-83 D. F. Cope . ... . . 131. .A. P. Litman
84. L.T.Corbin 7 . . 132, M., I. Lundin -
~-85. J. L, Crowley ~: - . "-133. R. N. Lyon
86, F. L. Culler = 134, H, G. MacPherson
- .87. J. M. Dale ... 77 7 " 135.  R. E. MacPherson -
88, D.G.Davis . 136. C.D. Martin
. 89. S.J.Ditto - -137. C. E. Mathews
- 90. .J, R. BEngel .. .. .. 138. C. L. Matthews
. 91. ‘E, P. Epler .~ - 139. R. W, McClung
.~ 92. D.E. Ferguson . .. 140. H. E. McCoy -
~-93. L. M. Ferris =~~~ 141, H. F. McDuffie
94, A. P. Fraas - - = k2. C. K. McGlothlan
- 95. H. A, Friedman 143, C. J. McHargue
- 96. J. H. Frye, Jre. ... lhh 158. T. W. McIntosh, AEC-
97. C. H. Gabbard o Washington
8. S el

R. B. Gallaher
 

159.
160.
.161.
- 162.
163.
16k,
165.
166.
167.
168.
169.
170.
171,
172.
173.
(T
175.
176.
- 177.
178.
‘ 179.
- 180-181.
182.
183.
184,
185.
186.

%

 

Internal Distribution

(continued)

L. E. McNeese _ 187.
A. 8. Meyer : - 188.
R. L. Moore o 189.
J. P. Nichols , 190.
E. L. Nicholson = 19l.
L. C. Oakes | - - 192.
P. Patriarca = | 193.
A. M.Perry - - 19k .
H. B. Piper - . 195.
B. E. Prince - 196.
J. L. Redford o 197.
M. Richardson 198.
R. C..Robertson 199.
H. C. Roller 200.
H. M. Roth, AEC-ORO - 201.
H. C. Savage 202.
C. E. Schilling ) o 203.
Dunlap Scott -20k .,
H. E. Seagren 205.
W. F. Schaffer 206.
J. H. Shaffer 207.
M- Shaw, AEC" 2080

Washington - 209.
M. J. Skinner 210-211.
G. M. Slaughter 212-213.

W. L. Smalley, AEC-ORO
A. N. Smith .
F. J. Smith

214-223,
22k,

External Dlstributlon

225-239.
240.

241-242. Reactor Division (ORO)

- G.

0.
P.
W.
I.
R.
H.
R.

P.
L.
G.
F.

Smith
Smith
Smith |
Spencer

Spiewak

cC.
H.
F.

s & d..a

R.
J.
S.
C.
B.
A.

J.

W.
K.
M.

J.

L.

E.

S.
S.
P,

H.

M.
R.
Je
W,
E.
c.
V.

Steffy
Stone
Sweek, AEC, Washlngton
Thama
Watson
Watson
Weaver
Webster
Weinberg -
Welir
Werner
West
Whatley
White
Wilson

G. Young -

H. C. Young

Central Research Lib.,
Document Reference Sect.
Laboratory Records ,
Laboratory Records - RC

Division of Technlcal Information Exten51on (DTIE)
Research and Development Director (ORO)

O,