ORNL-TM-1855.txt

MR

i

/

3 4456 051339, 4

 
 

N

 

: R LEGAL ROTICE = s e ey

This regort was preparsd as an sccount of Government spensorad wack, Neirher the Urited States,

nor the ‘Ceismiseion, nor any person acting on brehalf of the Commission: :

A, Makes aony warranty or representction, expressed st implied, with respest to the gccuracy,
complzteness, or usefplness of the informction contained in this repert, or that ths use of
any informarion, apparatus, method, or process disclosed in this repert may not infringe
privovaly owned rights; or ' ‘

B. Azsumes any licbilities with respect fo the use »f, or for damages resulting from the use of
any information, apparatus, methad, or process disclc?sed in this report, L

As usad in the aheve, “persus neting v behaif of the Commission® includes ony cmployes or

contractor nf the Commission, or employee of such sontrsctor, o the exteny that such employes

or coatcacter of rhe Commizssion, or smployee of such contractar prepares, disseminates, or
providas access to, any informetien porsuont to his employmert or noatract with the Cammission,

or his employmmnt with such contractor.

 
INTRA-LABORATORY CORRESPONDENCE

OAK RIDGE NATIONAL LABORATORY

Qctober 3, 1967

To: Recipients of the subject report.

 

ORNL-TM~1855 Unclassified

Report No.: Classification:

 

 

Author(s): Dunlap Scott and A. G. Grindell

Subject: Components and Systems Development for Molten=-Salt Breeder Reactors

 

 

For your information:

Reference 7, on page 52 of subject report is incorrect. It should read as follows:

’p. G. Smith, Experience with High~Temperature Centrifugal Pumps
in Nuclear Reactors and Their Application to Molten=Salt Thermal
Breeder Reactors, ORNL-TM~1993 (September 1967).

N. T. Bray, Supervisor

Laboratory Records Department
Technical Information Division

 

NTB:WCB:dbp

UCMN-6638
(3 5-68)

UCN-430
(3 5-61)
iii

TABLE OF CONTENTS

 

Page

INTRODUCTION 1
GENERAL STATUS OF TECHNOLOGY 1
PURPOBE : AND SCOPE OF DEVELOPMENT PROGRAM 2
REACTOR CORE 2
Review of Hydraulic Tests of the MSRE Core 2
Hydraulic Tests for the MSBE Core | L
Fuel Cell Tests in Molten Salt 5
CONTROL ROD AND DRIVE 5
Control Rod System for the MSRE 6
Control Scheme for the MSER 6
SALT PUMPS FOR MOLTEN SALT - 7
Present Technology T
Requirements for Pumps for Breeder Reactors 8
Program Scope 10
Design and Development Program 1l
Effects of Physical Properties of Breeder Salts 11

on Pump Design
opecific Design Problems and Goals 12
Separation Requirement 12
Hydraulic Design 12
Rotordynamic Analysis ' 12
Plastic Btrain 13
Purge Gas Requirements 13
Size-Scaling Requirements 13
Specific Development Problems and Goals 13

LOCKHEED MARTIN ENERG

i

3 445k 0513396 &

 
iv

TABLE OF CONTENTS -~ cont'd

 

Page

Fuel and Blanket Pumps 1k

Shaft Damper Tester 14

Molten-Salt Bearing Tester 14

Pump Test Facility 1L

Coolant Salt Pumps 15

HEAT EXCHANGERS 15
Review of Heat Exchangers in the MSRE 15
Fundamental Molten-Salt Heat Transfer 17

Heat Exchangers for Breeder Reactors 20
Ffuel~Salt Heat Exchanger 20
Blanket-35alt Heat Exchanger 21
Steam Reheater 21
Reheat Steam Preheaters 21
Boiler-Superheater 21

Heat Transfer Enhancement 23
PRESSURE RELIEF IN COOLANT SYSTEM 25
DRAIN AND STORAGE TANKS 25
Drain Tank System for the MSRE 25
Drain Tank System for the MSBE 26
VALVES FOR MOLTEN SALTS 28
SALT SAMPLERS 29
GAS SYSTEM 31
MSRE Cover and Offgas Systems 31

MSRE Gas System Performance 32

MSBR Offgas System 3k
TABLE OF CONTENTS - cont'd

 

MSBE Offgas System
Development Program
Gas Separator
Mass Transfer to Circulating Bubbles
Salt-Powered Injector
Cyclone Beparator
Salt-Mist Removal
Fission Product Filt;r
Kr-85 and Tritium Removal System
Gas Compressor
Gas Sampler
MBSBE Gas System Tegt Loop
CELL FURNACE AND SHIELDING
STEAM SYSTEM
TURBINE GENERATOR
ENGINEERING TEST UNIT
SCHEDULE AND COST

ACKNOWLEDGEMENTS

Page
37

37
38
38
38
40
Lo
41
42
L2
L2
43
43
L5
46
46
49
51
INTRODUCTIOCHN

The conceptual design of a 1000-Mw(e) Molten-Salt Breeder Reactor
(MSBR) is described in ORNL-3996.1 The Molten-Salt Reactor Experiment
(MSRE) ,2now operating, represents a first step in the development of
such a reactor. A Molten-Salt Breeder Experiment (MSBE) is proposed as
the next step. This reactor would be a 100~ to 150-Mw(th) model of the
MSBR designed to demonstrate all aspects of the breeder technology under
conditions at least as severe as those proposed for the full-scale -
breeder. Components and systems for the MSBE would incorporate all the
features of the full-scale units so that "scaling-up" the equipment to
higher power level would be the major task in building the reference
breeder.

The purpose of this report is to describe the present status of
development of components and systems for molten-salt resctors and to
present a development program for the MSBE. Since no design has been
made for the breeder experiment, the program is based on a study of the
problems of the reference design assuming that the MSBE would be a
"scaled-down" version of the modular concept desecribed in the reference
report. For purposes of organizing this report and the development
program, the plant was subdivided into components, systems, and general
problem areas. The design, the problems, the present status of the
technology, and the required development sre discussed for each sub-
division.

GENERAL STATUS OF TECHNOLOGY

 

The initial technology development for molten-salt reactors was
done in the early 1950's in the Aircraft Nuclear Propulsion (ANP) Pro-
gram at Oak Ridge National Laboratory. In carrying out this program,
much information on the physical, chemical, and engineering characteris-
tics of molten-salt systems was obtained from studies of fluoride salt
chemistry, and materials compatibility, and from development of compon-
ents, materials, fabrication methods, and reactor mainienance methods.
In 1954 the Aireraft Reactor Experiment (ARE), a 2-1/2 Mw(th) molten-
salt reactor--fueled with UF, dissolved in a mixture of zirconium and
sodium fluorides, moderated with beryllium oxide, and contained in
Inconel-~was built and operated successfully at outlet salt temperatures
up to 1650°F. '

The present molten-salt reactor program was initiated in 1957, draw-
ing upon the information developed in the ANP program as well as
beginning new investigations. By 1960 enough favorable experimental
results were obtained to support authorization for design and construc-
tion of a 10-Mw(th) Molten-Salt Reactor Experiment (MSRE). The MSRE
initiated power operation in early 1966, and provides facilities for
testing fuel salt, graphite, and Hastelloy N under appropriate reactor
operating conditions. The basic reactor performance to date has been
outstanding, and indicates that the desgirable features of the molten-
salt concept can be embodied in a practical reactor that can be
constructed, operated, and maintained with safety and reliability.
PURPOSE AND SCOPE OF DEVELOPMENT PROGRAM

 

The purpose of the program is to provide components and systems
with demonstrated reliability for use in the MSBE. A1l components and
systems must be of a design that can be scaled up to the higher power
level of the MSBR. The development of new types of sguipment and
improvement of existing equipment require that life-tests be performed.
Such tests provide information on limiting operational characteristics
and assist in predicting maintenance requirements. In addition, certain
performance tests must be made on components when operated as part of
a system to provide information for evaluating the compatibility of the
component with the system.

The development of new types of equipment such as the steam genera-
tor, the off-gags dispcsal system, the salt-ccoled control rod, and the
long shaft molten-salt pumps will regquire separate test facilities of
significant size. In addition, there will be numerous small tests con-
ducted to assist 1n resolving design features as well as to establish
the expected life of some components. These small tests may be conducted
in separate facilities but in many cases they can be incorporated into
one of the larger test facilities. In many areas the technology is
reasonably well established, butl conservative engineering requires
performance and life testing of the components to make sure they will
operate satisfactorily with the reactor.

For a final demonstration of the reliability and compatibility of
all molten~salt connected components and systems, an Engineering Test
Unit (ETU), a full-scale operating model of this MSBE, will be constructed
and operated, essentially isothermally, over the ranges of temperature
and salt flow proposed for the MSBE. As described in the appropriate
sections, the final evaluation of the components will be made while
operating as a part of this system. The model will also be used to train
operators for thne reactor and to demonstrate the maintenance procedures
and equipment.

REACTOR CORE

Review of Hydraulic Tests of the MSRE Core

 

The MSRE reactor vessel is a S5~ft-~diam by B—ft-high tank that con-
tains a 55-in.-diam by &7-in.-high graphite core structure. Under design
conditions of 10 Mw of reactor heat, the fuel salt would enter the flow
distributor at the top of the vessel al 1175°F and 20 psig. The fuel
1s distributed evenly around the circumference of the vessel and then
flows turbulently downward in a spiral path through a l-in. annulus
between the vessel wall and the core can. The salt loses its rotational
motlon in the straightening vanes in the lower plenum and turns and
flows upward through the graphite matrix in the core can. The graphite
matrix is an assembly of vertical bars, 2 in. by 2 in. by about 67 in.
long. The fuel flows in O.4-in. by 1.2-in. channels that are formed
by grooves in the sides of the bars. There are about 1140 of these
3

passages. Fuel was to leave the top of the reactor at 1225°F. Addi-
tional description of the MSRE core is given in the MSRE Design Report.2
The core development program was divided into two phases. The first
phase consisted of building and testing a 1/5 linearly scaled plastic
model. This model was operated with water and was relatively inexpensive.
It was used as a rapid method of checking the preliminary design to
establish the acceptability of major concepts.

The second phase consisted of bullding and testing a full-scale
model of the core at the rated flow. This model was used to establish
the design. The core vessel was made of carbon steel and the moderator
bars were extruded from aluminum. This model was used for a final and
much more detailed look at the hydrailic and thermal characteristics of
the core. ©BSome of the major items studied were:

1. Overall pressure drop and distribution of this pressure drop
among the core components.

2. Flow distribution by the volute.
3. Efficiency of the swirl killers in the lower vessel head.

L. . Heat transfer ccefficients in the lower and upper heads to
agssure adequate vessel wall cooling.

5. Flow distributions in the lower and upper head to assure that
no stagnant salt pockets were present.

6. Tendency of particulate matter to settle out in the lower
vessel head, on the tops of the core bars, and on the core support
Tlange.

T. Various other more minor phenomena.

Most of the measurements were made with water in the loop, and at
flow rates from the design flow down to 25% of the design flow. With
water, however, the Reynolds number was several times higher than would
be expected for fuel salt at the noted flow rate. To attain Reynolds
similarity, a thickening agent was added to the water to increase its
viscosity, and therefore decrease the Reynolds number. Several -of the
items in the above list were then rechecked. The agreement between
measurements in the 1/5 scale model, the full-scale model with water,
and the full-scale model with thickened water, was good where equivalent
measurements were made. None of these measurements were checked in a
molten-salt system. It was believed that the heat and momentum transfer
analogles were adequately well established to extrapolate water data to
a molten-salt system with a degree of reliablility much greater than was
required to insure adequate performance of the MSRE.

During the course of MSRE core'development, seversl small models
were made to check some hydraulic phenomena. Generally, these models
were made of plastic and operated with tap water.
Hydraulic Tests for the M5BE Core

The reactor core for the MSBE is expected to be about L4-ft-diam
by 5-ft-high and composed of re-entrant type graphite fuel cells through
which the fuel salt flows. The graphite tubes are attached to two plenum
chambers at the bottom of the reactor with graphite-to-metal trsnsition
sleeves. Fuel from the entrance plenum flows up through the outer
annulus of the fuel cell and down through the central passage to The
exit plenum. The fuel flows from the exit plenum to the pump then
through the heat exchanger and back to the reactor. A 2-ft-thick blanket
of a thorium-containing sali and graphite surrounds the core. The
blanket salt also permeates the interstices of the core lattice soO
fertile material flows through the core without mixing with the fissile
fuel salt.

Generally speaking,the MSBE core will be studied more critically
than the MSRE core because of its much higher power density. The pro-
posed development program for the MSBE core will be,in many respects,
similar to that for the MSRE core, and can be thought of as a two-phase
program.

The first phase will be directed toward making plastic models as
necessary for rapid and preliminary checks on major design concepts.
This could take the form of a complete scaled-down plastic medel as in
the MSRE, but probably not. Rather, small plastic models of individual
core components will be built and tested. Probable examples are:

1. A small model of the fuel salt distribution plenums would be
built and tested for proper flow distributionm to the fuel cells.

2. A small wmodel of the blanket salt distributor would be built
and tested for proper flow distribution.

3. A single full-scale model of a fuel cell would be built and
tested with tap water to measure the pressure drop and check for ade-
quite degassing on startup.

L. Other models as needed to provide confidence in the design.

Assembled units do not always behave as one might expect from
observing individual components. It is therefore necessary to test
hydraulically a complete prototype of the core. A full-scale prototype
is available in the ETU and it is planned to run the BETU with water
for a period of time, thus fluid measurements could be easily obtained.
However, it may take 6 months to a year to make all the measurements
necessary in the core. Tt would certainly be undesirable to restrict
the ETU to water operation for this long a period of time. We there-
fore plan, as phasse two, To build another and much less expensive
prototype of the core suitable for operating in a circulating water loop.
This special loop will alsc allow us to start testing the core sooner than
in the ETU, possibly in FY 1969. The loop will simulate both salt
systems. The core size will be half teo full scale, although full scale
is probably more desirable. The principal objectives of this model
will be: '

1. To demonstrate the required flow distribution of fuel and
blanket salt throughout the core.

2. To insure adequate flow for cooling structural members of the
core.

3. To demonstrate that no stagnant fuel and blanket salt regions
exist. '

L. To insure complete degassing of all fuel tubes during filling
and startup. :

5. To show that fluid induced vibrations are below acceptable
levels.

These measurements will be made over a range of flow rates both
above and below the design values. Water will be the fluid used in
most of these tests and Reynolds similarity will not hold. Where
necessary, the measurements will be repeated with a thickening agent
added to the water to attain Reynolds similarity.

Meagsurements made in the ETU would then be limited to those thought
necessary to confirm results of the water model. Certainly some data
will be taken with water in the system. OSome direct measurements with
salt in the system may be necessary, although this will be a more
difficult task and might have to await development of additional
instrumentation. WNevertheless, if some fluid dynamic characteristic of
the core is sufficiently eritical, it could be checked out in the ETU
while circulating salt. : :

Fuel Cell Tests in_Molten Salt

 

Demonstration of the performance of full~sgcale MSBE fuel cells
without radiation is an important part of the early phases of core
development for the MSBE. As soon as practicable, representative
graphite fuel cells will be operated with the full design salt flows,
temperatures and pressure differences. These tests will be run in the
pump development loop and the off-gas test locop. Removal and replace~
ment by the remote means will be demonstrated as part of this test pro-
gram.

CONTROL ROD AND DRIVE

 

The design of the MSBR takes advantage of the ease of adding fuel
while the reactor is operating to minimize the excess reactivity in the
core, the ability to drain the fuel to effect complete shutdown and
safety functions. However, a control rod or rods, as yet undesigned, are
included to permit short-term adjustments to the reactor temperature.
Control Rod System for the MSRE

 

The control rod system for the MSRE consists of a flexible poison
rod that is moved in and out of a re-entrant thimble by a continuous
link~-chain drive mechanism. This chain drive is controlled by a servo-
motor through a magnelic~clutch arrangement which permits rapid inser-
tion of the poison rod. In addition there are electrical synchros and
potentiometers for remote Indication of position, limit switches for
control of the range of motion, and a shock absorher to stop the rapid
insertion. The drive unit and the control element are cooled by cir-
culating alr through the drive housing and through the center of the
control rod.

The poison elements cperate at a temperature in the 1200 to 1LO0°F
range. The electromechanical drive unit, mounted well above the reactor
vessel, is slightly above ambient cell air temperature which does not
exceed 150°F. Two conditions dominated the design, high-thimble tem-
perature and maintenance-free operation. The electromechanical design
of the drive unit is straightforward, complicated principally by space
restrictions. It was not expected to be troublesome. The service
record of these M3RE rods and associated drive units has been good but
only because the final design was preceded by over a year of concen-
trated developmental testing of a prototype unit. As expected, these
tests disclosed a number of defects and confirmed the guality of the
final version.

Control Scheme for the MSBR

 

Although experience with the control rods for the MSRE provides
a useful background, the control recd and drive for the MGBE will be
considerably different. Design of the drive should be straightforward
but airceoling of the rod would not suffice at the much higher power
density, and the metal thimble would abscrb too many of the neutrons
needed for breeding 233U. A likely control scheme for the MSBE involves
the insertion of the control element directly into the center of the
reactor core, without using a thimble, and letting the fertile blanket
salt provide the necessary cooling. If it proves necessary to provide
cooling for the portion of the graphite rod which is in the gas space
above the fertile salt, a small stream can be diverted from the inlet
line and directed over the rod. One problem of this scheme is that
the drive mechanism must also be within a gas space directly connected
to the blanket salt. Not only does this radioactive environment present
a problem of electrical design but it also makes the repalir of the drive
difficult. A system of gas seals and buffer control should be developed
to permit the drive to operate in a clean gas atmosphere.

A thorough development and prototype testing program will be
required. Components of the drive and rod will be tested separately,
and then tested in assembly in a simulated reactor environment.
Finally, the rod will be operated in the ETU.
SALT PUMPS FOR MOLTEN SALT

 

The approach which will be followed to provide the pumps reguired
for molten-salt breeders is outlined. A brief resum€ is presented of
the present status of molten salt pump technology at ORNL and the con-
siderations given to using the MSRE pump configuration in the bhreeder
concept. A more desirable pump configuration is broached., and the prob-
lems anticipated with it are listed. Finally, the specific design and
development problems for the new configuration as they are presently
envisiconed are discussed in more detsil.

Present Technology

 

The present status of the technology of meolten salt pumps at ORNL
is set forth in References 3, U, 5 and 6. In brief, we have developed
the sump pump configuration in which the impeller is mounted on the
lower end of the pump shaft below the lower shaft bearing. Conventional
ball bearings and shaft seals, lubricated and cocled with a petroleum
base turbine oil, are utilized. The vertical shaft is mounted in a
bearing housing to support and guide the impeller in the pump veolute,
which is an integral part of the pump tank. The pump tank alsc serves
as the expansion tank for the molten salt system and is used in the MSRE
for the removal of gaseous fission products such as 135%e.

These pumps have been built in sizes from 2 to 1600 gpm to develop
heads to 400 ft of fluid. They have been used to pump molten salts
and liquid metals to temperatures of 1500°F. The MSRE fuel salt pump
circulates 1200 gpm normally at 1210°F against 49 £t of head, and the
coolant salt pump circulates 850 gpm normally at 1020°F against 78 ft of
head. The MSRE prototype fuel pump was operated at temperatures up to
1500°F. '

Four 5 gpm pumps, one 750 gpm and one 1200 gpm pump, were operated
at temperatures above 1200°F for periods greater than one year. The
750 gpm pump was operated with molten salt for 25,000 hr at 1200°F in
a regime of cavitation. Another test pump which was equipped with a
submerged journal bearing lubricated with molten salt was operated for
12,000 hr, during which it was started and stopped approximately 100
times. ' :

Two' pump characteristics of concern to operabtion of the MSRE were
determined in somewhat special fashions. Techniques were developed using
85K» to measure the back diffusion of gaseous fission products against
a flow of purge gas in the shaft annulus of the MSRE fuel pump. The con-
centration of undissolved gas in the circulating molten salt was measured
for the MSRE fuel pump in the prototype pump test facility using radia-
tion densitometry devices and techniques adapted to the task.

Larger pumps, of designs similar to those proposed for large molten
salt pumps, have been built and operated in liquid metal cooled reactors.
Operating conditions for three such pumps are given in Table 1. Experi-
ence with these pumps bears directly on the development of pumps for the
molten salt breeder reactors. A survey7 of the pertinent design features
and the operating experiences with these larger pumps is being made to
stimulate and enhance their contribution to the design of the breeder
salt pumps.

Teble 1. Pumps for Liquid Metal Reactors

 

Hallam Fermi ERR-2

 

Flow, gpm 7200 11,800 5500
Head, ft 160 310 200
Temperature, °F 1000 1000 800
Speed, rom 900 300 1035
Pumping power, bhp 350 1060 350

Operating experience:
Hallem pumps accumulated several thousand
hr of pump operation with sodium from 300
to 950°F, of which at least 1000 hr was at
950°F.

Ferml pumps accunulated over 7000 hr operation
inecluding two weeks at 1000°F.

 

Regquirements for Pumps for Breeder Reactors

 

The presently envisioned requirements of the fuel, blanket, and
coolant salt pumps for a 1000 Mw(e) Molten-Salt Breeder Reactor (MSBR)
plant_ and for the Molten Salt Breeder Experiment (MSBE), a 150 Mw(th)
experiment, are presented in Table 2. Tentative values for pertinent
hydraulic design parameters, e.g., speed, specific speed, are given
also.

The centrifugal sump pumps developed and used in the Aircraft
Reactor Experiment (ARE), the Aircraft Reactor Test program (ART), and
the Molten-Salt Reactor Experiment (MSRE) received first consideration
for application to the Molten-Salt Breeder Reactor (MSBR). There are
at least three differences between the MSRE and the MSBR concept whose
effects on the thermal and nuclear radiation environments will influence
the choice of the pump configuration for the MIBR.

Depending on the type of designs,the power rating for the MSBR is
fifty to 200 times greater than MSRE design power. In addition, the
separation distances between its reactor, heat exchanger, and pump are
equal to or smaller than the corresponding MSRE distances. Thus the
intensity of the nuclear radiations in the viecinity of the fuel pump
will be very much greater for the MSBR than the MSRE.
Table 2. Pumps for Breeder Reactors

 

 

Fuel j ‘Blanket | Cooclant
2225 Mw(th) MSBR
Number required La La | L&
Deéign'temperature, °F 1300 1300 ' 1300
Capacity, gpm 11,000 2000 16,000
Heat, ft 150 80 : 150
Speed, rpm 1160 | 1160 1160
Specific speed,.Ns 2830 2150 3400
NPSH, required, ft 25 8 32
(Net positive suction head)
Impeller input power, hp - 990 250 1440
150 Mwt MSBE
Number required 1 1 1
Design temperature, °F 1300 1300 1300
Capacity, gpm 4500 ' 540 4300
Heat, ft 150 | 80 150
Speed, rpm 1750 1750 1750
Specific speed, N_ | 2730 1520 2670
NPSH required, ft 27 5 | 26
(Net positive suction head)

Impeller input power, hp L10 61 ‘ 390

 

aJThe same total number of pumps is requlred for a 1000 MW( ) plant
of the MSBR reference design or modular design.
10

Another difference concerns the manner of heating the MSBR. One
off the features of the MSBR concept is the use of large furnaces to
contain the fuel, blanket and cooclant salt systems and to maintain them
at elevated temperatures during reactor power operation. The temperature
in the furnace for the fuel and blanket salt systems will range between
1050 and 1150°F. The temperature in the coolant salt system furtace may
range between 700 and 1150°F.

A list of the conditions and circumstances under which the MSRE
pump configuration may be used in the fuel and blanket salt systems
include:

a. Provide a specially constructed and cooled pit, both
to maintain the ambient temperature for the bearing
housing in the range 150 to 1T75°F,and reduce intensity
of nuclear radiations.

b. Develop suitable shaft bearings and seals and the
associated lubricant for operation at a higher ambient
temperature which, although still requiring the con-
struction of a pump pit, would materially reduce the
heat load on the pump pit cooling system.

c. Return to the concept of local preheating of the salt
system componentg,with attendant use of local nuclear
radigtion shielding end space cooling,k to maintain the
ambient temperature below 200°F.

In a more desirable pump configuration, the thermal and radiation
damage sensitive drive motor 1s separated from the pump, per se, by a
sufficiently large distance to provide both reascnable thermal gradients
in the pump structure and adequate amounts of radiation attenuation
materials. The approach is to separste the drive motor from the hostile
environments by as large a distance as practicable within the limits
of rotordynamic, fabrication, and reactor layout considerations. Pre-~
liminary study indicates that the required separation probably cannot
be obtained with the MSRE pump configuration using a reasonable shaft
diameter. Thus initial consideration will be given to a pump con-
figuration that features a rather long slender ghaft and utilizes
molten salt lubricated bearings and probably shaft dampers.

Program Scope

The pump program will provide for the study and design of the fuel,
blanket, and coolant salt pumps for the larger MSBR, and for the design
and development of those pumps for the smaller MSBE. The study will
include the evaluation of the feasibility of the long shaft pump con-
figuration and the practicability of scaling it down by a factor of four
to suit the MSBE pump reguirements.

Our present approach is to use one basic pump rotary assembly
design and to accommodate the differences in the hydraulic requirements
for the three pumps with appropriate changes in the hydraulic designs of
the impeller and volute and in the characteristics of the drive motors.
11

If, for reasons of reactor system layout the coolant salt pump
requires separate treatment, then either the long shaft configuration
will be modified or the MSRE pump configuration will be used, depending
upon the results of further study.

One each of the fuel, blanket, and coclant salt pumps will be pro-
vided for 1) development, 2) the Engineering Test Unit, and 3) the Molten-
Salt Breeder Experiment. -

In essence, the study portion of the program will be focused on
identifying pump configurations that are feasible for the MSBR, and
the development portion of the program will be concerned with producing
pumps for the MSBE, which will be scaled-down versions of the MSBR
configurations. During the development of the MSBE pumps, attention
will be given to the problems of scallng—up components for use in the
MSBR pumps.

 

Design and Development Program

Because of the importance of the pumps and the close relationship
between their design and development, these two activities are considered
to be one. In this activity the major problems are expected to include:

l. selecting a hydraulic design,

2 choosing a satisfactory rotordynamic configuration,

3. controlling the total plastic strain in the pump caused
by temperature cycling of the system,

L specifying purge gas reguirements to prevent back
diffusion of gaseous fission products to radiation
sensitive regions of the pump.

5. controlling adequately any flow which passes through
the pump tank, a) to prevent the re-entrainment of
xenon-laden gas in the reécirculating salt, and b)
to prevent stoppage of purge gas flow by freezing of

- salt splash or aerosol in the pump shaft annulus,
6. designing and proof testing an adequate shaft damper,
a molten-galt lubricated bearing, and any shaft seal
that is larger in diameter than now used,
T. wverifying the adequacy of the hydraulic and rotor-
- dynamic designs,

8. providing pump reliability, and

9. obtaining confidence in scaling-up the MSBE pumps to
- fit the requirements of large-scale plants.

 

Effects of Physical Properties of Breeder Salts on Pump Design

Density and viscosity are the two physical properties of molten
salts which strongly affect pump design. 8alt density mainly affects
the torque requirements for the pump impeller and requires that the
shaft has sufficlent torsional strength and that the drive moter pro-
duces the required torque. Both of these items, which are under the
control of the pump designer, should present no untoward problem in
the design of the breeder pumps.
12

Viscosity strongly affects the life characteristics of the hydro-
dynamic bearings, which we anticipate will be used in the breeder pumps.
The values of the viscosity for all three breeder reactor salts are
similar and greater thean wabter and should present no untoward problem
in the design of molten salt lubricated bearings.

 

Specific Design Problems and Goals

Separation Requirement. The principal feature of the long shaft
pump configuration is the use of sufficient separation distance and
shielding to provide for ten years of operation of the drive motor. Such
a configuration requires a long, slender shaft guided at its lower end by
a molten-salt bearing and at its upper end by a more conventional bearing
and using, hopefully, a conventional, easily replenishable lubricant.

A shaft damper will prohably be necessary to provide for operation at a
speed above the first critical frequency of the shaft-bearing system.

 

Estimates of the separation distance between the pump impeller and
drive motor will be made based on the anticipated flux of neutrons and
gamma radiation at the motor and the shielding required to provide
ten-year life for the radiation damage sensitive materials in the fuel
and blanket pumps.

Hydraulic Design. We plan to select the hydraulic design which
will provide the reguired head (H) and capacity (Q) at as high shaft
speed (N) as good practice and the available net positive suction head
(NPSH) in the system will permit. This approach should permit the use
of a relatively small diameter impeller and volute and should minimize
the parasitic volume of salt. Vanes for the back side of the impeller,
suitable for reducing hydraulic thrust,will receive consideration.

 

Rotordynamic Analysis. The principal snalytical problem we antici-
pate concerns the selection of satisfactory pump rotordynamic configura-
tions. These should provide reliable and economic pumps for the fuel,
blanket, and coolant salt circuits in the MSBR, which can be scaled-
down by a factor of approximately four for use in the MSBE.

 

The rotordynamics of the proposed "long shaft" configuration are
new to us and will be analyzed extensively to determine a suitable
arrangement of shaft, bearings, and shaft damper. We plan to select two
or three out of several promising shaft-bearing configurations which
provide the required separation, and to subject them to rotordynamic
analysis. The location and performance characteristics of both shaft
dampers and bearings necessary to provide a safe margin of fatigue life
for the shaft during ten—-year operation will be determined for several
values of shaft diameter. The configuration which has the best chance
of providing reliable pumps will be chosen for development. Appropriate
rotordynamic analysis of a reduced scope will be performed for the
coolant salt pump, if a different configuration is required.
13

Plastic Strain. Temperature cycles in a salt system can impose
increments of plastic strain in the high-temperature portions of the
pump due to changes in either thermal stresses associated with steep
temperature gradients or mechanical stresses associated with pipe
anchor forces and moments exerted on pump nozzles. We noted that the
largest temperature gradients assoclated with the nuclear operation of
the MSRE fuel pump were caused by heat deposited in the pump walls by
gaseous fission preducts. It is likely that more fission products will
be present and heat will be deposited in larger quantities in the salt
reservoir in the MSBR fuel pump. We plan to use a small portion of
the circulating fuel salt to remove the heat. The total plastic strain
in the pump nozzles,resulting from the forces associated with heating
and cooling the system and changing reactor power levels, will be
estimated for & specified number of cycles. Measures will be taken to
keep the total strain within the plastic fatigue strength of the con-
tainer material.

Purge Gas Requirements. An inert purge gas will be used in the
MSBR, as in the MSRE, to: (1) remove, dilute, and transport to an
appropriate trap system the xenon and other gaseocus fission products
from the fuel salt; (2) reduce the back diffusion of these gaseous
fission products into radiation sensitive regions of the pump; and
(3) remove any lubricant that leaks past a shaft seal without per-
mitting the leakage to enter the pumped salt. The amount of purge
gas required for the fuel end blanket salt pumps will be much larger
than for the MSRE fuel pump, and a recycle system will be used to
conserve gas. The recycle system is treated in the section on the
offgas system. The smaller purge gas flow for the coolant salt pump
may permit open-cycle operation. ‘ :

 

Size-8caling Requirements. Pumps for the MSBR and MSBE should use
the same genersal configurations. The feasibility of scaling up the
MSBE pumps by a factor of four for application to the MSBR will be one
criterion for acceptance of MSBE pump design. The scaling of the rotor-
dynamic configuration will be made a part of the analysis of the MSER
pumps, which, in turn, should establish the reguirements for scaling
the molten salt bearings and dampers. Fabrication, inspection, handling,
assembly, and installation of the MSBR pumps will receive study to
determine that the long shaft configuraticn will not impose expensive
solutions for large molten~salt systems. We plan to have the MSBE
pumps fabricated by industry, and to discuss extensively with them
during this time the fabrication problems of the MSBR pumps. The neces-
sity for proof-testing the large pumps in molten salt will receive much
attention during similar tests with the MSBE pumps.

 

Specific Development Problems and Goals

 

The development of the MSBE pumps will entail the testing of pumps
and certain pump components and the feedback of information from these
tests to pump design, and will include all the aspects of detsiled hard-
ware fabrication. The main problems anticipated with the long shaft
punp configuration include: (1) demonstrating the adequacy of the design
1k

and the reliabllity of tne shaft dsmper and the molten-salt bearing

in component testers; (2) providing adequate control of the bypass salt
flow which carries fission-product laden heliuvm into the pump tank:; and
(3) verifying the adequacy of the hydrsulic, rotordynamic, and purge gas
designs for each complete pump. The long-time reliability (availability)
of the pumps will be demonstrated in endurance tests. The feasibility
of verifying the rotordynamic characteristics of large pumps in room
temperature shaker tests of small models of shaft-bearing-damper con-
figurations will be studied also. 8Since many of these development tasks
are of routine nature, only those problems whose resolutlion meets
specific and significant goals are discussed below:

Fuel and Blanket Punmps

 

Shaft Damper Tester. The hydraulic performance and the mechanical
design of the demper will be verified in what we anticipate will be a
room-temperature tester using a fluid which approximates the kinematic
viscosity of the damper working fluid. The damping coefficient reguired
to reduce shaft flexure stress to a value satisfactory to provide ten~
year pump life will he deduced during rotordynamic analysis of the pumps.
In the tester, we anticipate imposing on the damper a sinusoidal trans-
verse motion of known amplitude and fregquency and deducing the damping
cecefficient from measurements of the force necessary to sustain that
motion. Satisfactory correlation between predicted and experimental
values of the damping coefficient would provide confidence in extrapo-
lating the hydraulic and mechanical designs of shafti dampers to pumps
for the large-scale systems.

 

Molten-3alt Bearing Tester. The operating stability of the bearing
and the start-stop wear resistance of the bearing materials will be
verified in a component tester. We anticipate first operating a suitable
bearing configuration at room temperature with a fluid having the
approximate kinematic viscosity of the appropriate salt Ian order to
insure stable operation of the bearing. Next, the bearing will be
operated in the appropriate molten salt in the tester for more than
the anticipated number of starts and stops for the pump in the MSBE.
Then, the btearing will be thermally cycled over the temperature range
and the number of cycles anticipated for the MSBE and operated in
an endurance test to obtain confidence in the adequacy and reliability
of its mechanical design. The tester will be designed to accommodate
the larger diameter bearings anticipated for pumps for large-scale
systems. Sufficient tests will be made with mockup fluid to establish
the stability of the larger bearings.

 

Pump Test Facility. We plan to verify the hydraulic and rotor-
dynamic designs and to establish control of salt bypass flow, i1.e., to
eliminate salt splash and re-entrainment of xenon-laden gas in the
recirculated salt using a fluid which has kinematic viscosity similar
to the appropriate salt. Then the hydraulic and rotordynamic designs
and the functions of the purge gas system will be checked in molten
sall operation, after which the reliability of the pump will be investi-
gated during endurance tests. Maintainability of the pump will be
demcnstrated during these tests also.

 
15

It will be necessary to perform several kinds of room temperature
tests with each of the three salt pump designs using a suitable fluid.
These tests include: 1) checking the hydraulic design performance,

2) developing appropriate controls for required bypass flows through
the pump tank, 3) providing adequate capacity for degassing the liquid,
and 4) determining the adequacy of the pump design to meet special,
transient or emergency conditions encountered in reactor operation or
of revisions to pump design deemed necessary. Preliminary study indi-~
cates these tests can be performed for all three pumps in a single room
temperature facilily using an appropriate fluid.

The differences in the chemistry and physical propertlies of the
three salts and in the flow rate requirements for the individual salt
circuits indicate the desirability of using three high temperature
pump test facilities, one each for the fuel, blanket, and coolant salt
pumps.

A facility for meking a room temperature rotordynamic test of the
full-scale rotary assembly will also be provided, 1f the results of the
rotordynamic analyses indicste this necessity.

Coolant Salt Pumps. In the event that the MSKE pump configuration
is chosen over the long shaft configuration for the coolant salt pump,
the following development program will be carried out.

 

The leaksge performance and life characteristics of the shaft seal
will be obtained at an esrly practicable date. The bearing-shaft-seal
configuration will be mocked up and operated at design speed and tem-
perature using a good grade turbine oil to lubricate the seal. Follow-
on tests will be made with larger diameter shaft seals suitable for the
coolant salt pumps in large-scale systems.

In addition to the molten salt pump test facility alluded to
previously, a room temperature pump test facllity suited to the water
test development of the MSRE pump configuration will be provided. The
hydraulic design of the pump will be checked and satisfactory controls
for the bypass flow of salt in the pump tank will be developed. Subse-
quent ly, performance and endurance tests will be performed with molten
salt in the high temperature facility.

HEAT EXCHANGERS

 

Review of Heat Exchangers in the MSRE

 

Although 10 Mw(th) was the nominal power level for the design cal-
culations for the MSRE, +the actual capability is limited to about
7.5 Mw(t). Both the primary heat exchanger (fuel salt to coolant salt)
and the radiator {coolant salt to air) contribute to this reduced capacity.
A review at each heat exchanger will be very briefly summarized.

The primary heat exchanger is a conventicnal cross baffles, U-tube
exchanger, with fuel salt on the shell side and coolant salt on the tube
16

side. For a more detailed description see Ref. 8. The observed overall
neat transfer coefficient was about 60% of the estimated design coefficient
(refs. g and 10). The initial design of this heat exchanger was reviewed
in detail and the following significant items noted:

1. eat transfer coefficients and pressure drops were computed
from conventional relationships for normal fluids.

2. The physical properties of salt used in the design were
those believed to be correct at that time.

3. A total contingency factor of something over 20% was
included in the heat transfer area.

In trying to determine why the measured coefficient was so low, many
things were considered. The following are the most pertinent.

1. Since salt does not wet Hastelloy N, the guestion of
a helium gas film on the tubes was considered. This
was discounted by pressure release tests discussed
in ref. 9 and other considerations.

2. The question of an insulating scale was considered. The
very high resistance of Hastelloy N to attack by fuel
salt in a good many loops and the reactor made this a
negligible consideration. Also, there has been very
little, if any, drop in U with time in the reactor.

3. The physical properties of salt were looked at
critically, and herein lies what we believe to be
the discrepancy.

Specifically,it involves the thermal conductivity. Shown in the
table are values of thermal conductivity used in the heat exchanger
design. These values were estimated from data on similar but not
identical salts. Also shown are recently measured values (ref.li) for
fuel salt and an estimated value for coclant salt.

Thermal Conductivity
(Btu/hr ft °F)

 

 

Fuel Salt Coclant Salt
Values used in original design 2.75 3.5
Measured value (ref. 11) 0.83 Not yet
measured
Estimated for cooclant salt = (0.83)(3.5/2.75) -- 1.06

 

Now 1f the overall heat transfer coefficient of the primary healt exchanger
is recomputed with these new values of thermal conductivity, then the
measured and computed values agree quite well.
17

The primary heat exchanger was tested with water after it was
built. The tubes vibrated excessively in the baffles at flows greater
than about 2/3 of the design flow through the shell, and the shell side
pressure drop was excessive. Modifications were made to tighten the
tubes in the baffles and to reduce the pressure drop at the outlet
nozzle. The exchanger has been operated for about 9000 hr at high tem-
perature with salt without obvious difficulty.

The radiator consists of 120 tubes, each about 30 £t long, in the
form of an S-shaped bundle. Alr blows across the tube bundle st right
angles to remove the heat. For a more detailed description, see ref. 8.
The observed overall heat transfer coefficient was about 68% of the
design coefficient (refs. 9 and 10 ). The initial design of the radiator
was reviewed in detail and the following significant items noted:

1. The salt side coefficient was computed from conventional
relationships for normal fluids. In the radiator this is
a negligible consideration, however, because only about
2% of the resistance to heat transfer is through the salt
film.

2. A design error was found in the calculation of the air
coefficient. This resulted in the estimated outside
coefficient being about 14% too high.

3. A contingency factor of only about 4% was included in
the heat transfer area.

If the overall coefficient is recomputed to account, for the error
mentioned in item 2, then the observed coefficient is gbout 75% of the
design coefficient. We believe that a discrepancy of this magnitude
is not unreasonable when the unconventional configuration of the
radiator is considered. A 20 to 30% contingency factor should have
been included in the design heat transfer area to be certain of 10 Mw
capacity.

The conclusion we have arrived at after looking at the performance
of both heat exchangers in detail, i1s thal no unique healt transfer
problems exist in the MSRE that can be attributed to some unusual
behavior of the molten salt. '

Fundamental Molten Salt Heat Transfer

Heat transfer with molten-salt mixtures was extensively studied
during the period 1950-195T7 as a part of the ANP effort, and the period
1958~1961 as a part of the MSR program. The mixtures investigated are
tabulated below along with certain pertinent parameters:
18

 

 

. Composition Reynolds Test Section
Salt Mixture (mole%) Modulus Material
NaOl 100.0 6000-12,000 Nickel
NaF-KF-LiF 11.5-42.,0-46.5 2300-9000 Nickel,
Inconel,
316 S8
NafF-KF-LiF-UF), 11.2-b1.0-45.3-2.5 2800-13,000 Inconel,
316 S8,
Hastelloy B
NakF-7rFL-UF), 50.0-46.0-4.0 7000~10,000 Inconel
NaF-ZrF),-UF), 53.5-40.0-6.5 3500-14,000 Inconel
LiF-BeF,-UF), 53.0-46.0-1.0 4000~5000 Inconel
LiF-BeF,-UF 62.0-37.0-1.0 3000-7000 Inconel
LiF-BeFp-UF),-ThF) 67.0-18.5-0.5-14.0 5500-25,000 Inconel,
Hastelloy N
LiC1-XC1 41.2-58.8 6000-18,000 Inconel,
347 88
NaNOp~NaN03- LT7.0-7.0-53 4900-25,000 Inconel,
KNO4 (wt %) 316 S8

 

While these experiments indicated that the molten salts have the
heat transfer and fluid mechanical properties of common fluids (0.5 <
Np, < 100) ,* the same studies showed that such phenomena as non-wetting
and interfaclal deposits could drastically reduce the heat transfer.
Since these effects are difficult to predict, the heat transfer charac-
teristics of the molien salts for critical applications should be ex-

perimentally established.

A bibliography covering ORNL studies on

molten-salt heat transfer is given in references 1. through 8.

1. H. W. Hoffman, Turbulent Forced Convection Heat Transfer in
Circular Tubes Containing Molten Sodium Hydroxide, USAEC
Report ORNL-1370, ORNL (1952); see also Heat Transfer and
Fluid Mechanies Institute, p 83, Stanford University Press,

 

Stanford, Calif. (1953).

 

2. H. W. Hoffman and J. Lones, Fused Salt Heat Transfer-
Forced Convection ieat Transfer in Circular Tubes
Containing NaF-KF-LiF Hutectic, USAEC Report ORNL-1T7TT,
ORNL (1955).

Part Il:

3. H. W. Hoffman and S. I. Cohen, Fused Salt Heat Transfer-
Forced Convection Heat Transfer in Clrcular

Part IIT:
Tubes Containing the 3alt Mixture NaNOp-NaNO5-KlO4,

USAEC Report ORNL-2433, ORNL (1960).

L. H. W. Hoffman, fused Salt Heat Transfer, Reactor Heat
Transfer Information Meeting Oect. 18~19, 1954, p. 23,
USAEC Report BNL-311, Brookhaven National Laboratory
(classified).

*The Prandtl modulus, NPr for the tabulated mixtures ranges from 1 to 10.
19

H. W. Hoffman, Molten Salt Heat Transfer, Reactor Heat
Transfer Conference of 1956, p. 50, USAEC Report TID-7329
(Pt. 3), (classified).

H. W. Hoffman, Molten Salt Heat Transfer, USAEC Report
ORNL-CF-58-2-40, ORNL (1958).

ANP Quart. Prog. Repts. Period Ending Sept. 10, 1955,

p. 149, USAEC Report ORNL-1947; Period Ending Dec. 10,

1955, p. 170, USAEC Report ORNL~-2012; Period Ending

Mar. 10, 1956, p. 171, USAEC Report ORNL-2061; Period

Ending June 30, 1957, p. 99, USAEC Report ORNL-2340;

Period Ending Sept. 30, 1957, p. 103, USAEC Report ORNL-2387;
Period Ending Dec. 31, 1957, p. 57, USAEC Report ORNL~2LL0.

MSR Quart. Prog. Repts. Period Ending Jan. 31, 1958,
USAEC Report ORNL-24Th; Periocd Ending June 30, 1958,
USAEC Report ORNL-2551; Period Ending Oct. 31, 1958,
USAEC Report ORNL-2626; Period Fnding Jan. 31, 1959, 67,
USAEC Report ORNL-2684; Period Ending Apr. 30, 1959, p. 41,
USAEC Report ORNL-2773; Period Ending July 31, 1959, p. 39,
USAEC Report ORNL-2799; Period Ending Oct. 31, 1959, p. 23,
USAEC Report ORNL-2890; Periods HEnding Jan. 31 and Apr. 30,
1960, p. 27, USAEC Report ORNL~-2973; Period Ending

July 31, 1960, p. 86, USAEC Report ORNL-301ll4; Period Ending
Feb. 28, 1961, p. 140, USAEC Report ORNL-3122; Period
Ending Aug. 31, 1961, p. 132, USAEC Report ORNL-3215.

37,
.13,
46,

g 3 oo

For added assurance and for other reasons, a molfen salt heatl
transfer loop will be built. Generally the loop will be small and
oriented toward fundamental heat transfer rather than toward component

testing.

It will be capable of operating with a wvariety of salts and

container materials. Some of its objectives will be to evaluate:

1.

Transfer coefficients and pressure drop with salt
flowing axially on the inside and outside of the tubes.

Thermal effects of possible corrosion and scale deposits
on the tubes.

Effects of circulating gas bubbles on heat transfer.

This could be significant if the salt does not wet the
tubes, and is of concern in the MSBE because gas bubbles
will be intentionally injected into the fuel salt at

the inlet to the primary heat exchanger to aid in stripping
Xenon-135.
20

Heat Exchangers for Breeder Reactors

 

Molten-salt breeder reactors of the reference design make use of
five different heat exchangers: (1) the fuel salt heat exchanger,
(2) the blanket salt heat exchanger, (3) the boiler superheater,
(4) the steam reheater, and (5) the reheat steam preheater. These heat
exchangers will be largely designed and built by commercial manufacturers.
Fabrication procedures will be devélo%ed by the manufacturers or as part
of the materials development program. 2However, any heat transfer or fluid
flow studles necessary to assure adequate performance of the units is a
part of this component development program.

Fuel Salt Heat Exchenger

 

The function of this heat exchanger is to transfer heat from the
fuel salt to the coolant salt. The heat exchangers in each of the MSBR
circuits has a capacity of 528 Mw. The MSBE will have one heat exchanger
with a capacity of 150 Mw.

The fuel salt is in the tubes and makes two passes through the ex-
changer. First it goes downward through an annular bundle of tubes near
the center. It reverses direction in a floating head and then passes up
through another annular tube bundle at the outer periphery. The coolant
salt on the shell side flows generally countercurrent to the fuel sslt.
Baffles are incorporated to attain cross flow. The coolant salt outlet
is through a pipe located at the center of the heat exchanger and running
its entire length. This heat exchanger, and all other heat exchangers
in this report, are in the conceptual design stage and subject to change
as more design work is done on the reactor.

A heat exchanger of full MSBE size will be built and tested in
the ETU. The ETU is conceived of as an isothermal loop; nevertheless,
it will have a large capacity for heating fluids. With this heat source
available and with water in the system, we expect to measure reasonably
well the overall heat transfer coefficient. This will characterize the
heat exchanger physically; that is, the combined effects of baffles, un-
conventional geometry, etc. The overall heat transfer coefficient can
then be extrapolated to a salt-salt system. While water is in the ETU,
the pressure drop through the heat exchanger will be measured. In
addition, the unit will be checked for fluid induced vibrations. As
many of the measurements as possible will be checked again with salt in
the systemn.

Because of the unconventional configuration of this heat exchanger,
various small models may have to be built to serve as a check of the
hydraulic and structural design. For instance,a relatively small plastic
model to run on water would be used to look at the adequacy of the flow
distribution produced by the coolant salt inlet volute. Also, because
of the complexity introduced by combining the pump with the heat ex-
changer, models may have to be built to measure thermally induced
stresses.
21

 

Blanket Sdlt Heat Exchanger

The purpose of this exchanger is to transfer heat from the blanket
salt to the coolant salt. The MSBR exchangers each have a capacity
of 28 Mw. The MSBE would have a 5- to 8-Mw exchanger. They are similar
to the fuel salt heat exchangers and the same comments and program apply.

Steam Reheater

The function of this exchanger is to reheat steam from the high-
pressure turbine before it is admitted to the intermediate pressure
turbine. The MSBR plant has eight reheaters, each rated at 36 Mw; the
MSBE would have a 12- to 18-Mw unit.” The reheater would be single pass
of steam and cooclant salt with the steam in the tubes and the salt flowing
countercurrently through the disc-and-doughnut-baffled shell. Control
of the steam outlet temperature is obtained by varying the salt flow.

Generally, this heat exchanger is conventional, and no fundamentsl
heat transfer problems are foreseen. Again, the overall heat transfer
coefficient should be measured in the ETU, possibly with salt on the
shell side and low-pressure superheated steam on the tube side.

Reheat Steam Preheaters

 

The function of this heat exchanger is to preheat the high pressure
turbine exhaust to 650°F before it reaches the reheaters. This is
necessary to minimize the chances of freezing the coolant salt In the
reheaters. The heating fluid is steam at 3500 psia and 1000°F. The
exchanger is U-shaped containing a single pass of U tubes. Ixhaust
steam from the high-pressure turbine mekes a single pass through the
shell side; it is superheated throughout. Throttle steam makes a single
pass through the tubes countercurrently; it is supercriticzal throughout.
The heat transfer coefficient on the tube side is high compared to the
overall coefficient, so that even the boundary layer temperature is
fairly well above the critical point. No development work is planned for
this hedt exchanger. The unit for the MSBE will be designed with ample
capacity and the overall heat transfer coefficient will be measured
during the operation of the reactor.

Boiler-Superheater -

 

The function of these heat exchangers is to take heat from the coclant
salt and generate supercritical steam for the turbines. Each of 16 ‘
exchangers in the MSBR plant has a capacity of 121 Mw. Two exchangers
of about 60-Mw capacity each or one of 120 Mw capacity would be used in
the MSBE. Physically, the heat exchangers are U shell with a single pass
of U tubes. The shell side has cross flow baffles with a variable pitch.
The pitch is greatest in the central portion of the exchanger where the
bulk fluid temperature difference is highest. Salt on the shell side and
supercritical weter on the tube side flow countercurrently. This heat
exchanger is conventional except for the supercritical aspect of the
water. Coolant salt is supplied through a throttling valve which permits
control of the outlet steam temperature.
22

The choice of the supercritical steam cycle for the MSBR steam
generator was based on considerations of the high melting point of
the fuel and coolant salt, the large thermal stresses which are produced
in the tube wall during normal ocperation, and the higher thermal efficien-
cy which may be obtained with the supercritical cycle.

We prefer to operate the fuel and coolant salt system such that even
local freezing would not occur. The fuel galt, which freezes at aboul
850°F, will be kept from freezing in the fuel heat exchanger by maintaining
the temperature of the coolant salt above 850°F. and the coolant salt will
be kept from freezing by maintaining the feedwater temperature at about
TOO°F. Conventionally, a feedwater temperature of less than 5T5°F is used
but the higher temperature is possible with the supercritical system.

Also, by raising the feedwater temperature, the temperature gradient
across the tube wall is reduced to about 1/2 with a corresponding reduc-
tion in the thermal stress for normal operation.

Rapid thermal cycling of the tube wall which is produced by a rapid
oscillation of the steam-water interface in subcritical systems will be
greatly reduced by the virtual elimination of the phase change in going
through the critical temperature in the supercritical pressure system.

The thermal efficiency consideration is consistent with the trend
of the power industry toward the use of supercritical steam systems to
obtain the highest efficiency for the temperature. Since the high nickel
alloy is to be used to contain the cooclant salt anyway, there will be
no additional penalty from the materials standpoint. The study of the
compatibility of Hastelloy N with supercritical steam will he done as a
part of the materials progra.m.12

Several investigations of heat transfer to supercritical water
have been recported in the literature. Notably are two investigations,
one by Babcock and Wilcox (ref. 13) and the other by Westinghouse (ref.ili).
Both programs are quite extensive. The programs were carried out
independently and yielded practically ldentical heat transfer coefficient
correlations. As a result of this, we believe that extensive work
directed toward fundamentally measuring and correlating coefficients is
unwarranted at this time. Nevertheless, the technology of supercritical
heat transfer is relatively new and there are gquestions of corrosion,
scale deposits, thermal stresses, etc. and how these affect tube per-
formance. One of the most potentially serious unknowns concerns super-
critical heat transfer whistle. '"Whistle" is a phenomenon much like the
more conventional boiling songs, but confined to supercritical systems.
Scometimes this whistle is associated with extensive tube vibrations
and flow oscillations. Little is known about the cause and effect of
this phenocmenon.

Two developmental programs are planned concerning supercritical
heat transfer. The first will be a rather fundamentally oriented study
in a small laboratory loop. One or possibly two supercritical steam
generating loops will be involved. This program 1s intended primarily
to lock at whistle from a mechanistic point of view. Calculated heat
transfer coefficients will also be confirmed.
23

The second program involves a larger two fiuid (salt and water)
system, and can be thought of as s component development lecop. This loop
will feature two or three full-sized ftubes from the boiler superheater.
The tubes will be operated at their rated heat load with circulating salt
on the shell side transferring heat to water in the tubez at supercritical
conditions. ©Some of the principal objectives of this loop are as
follows:

1. To assure the absence of supercritical whistle in a nearly
exact duplicate of the boiler superheater tubes and under
similar conditions.

N

To confirm the overall heat transfer coefficient calculated
for the boiler superheater.

3. To gain experience in corrosion and scale bulldup in &
once~through system of this type, and to obgerve the feed-
water chemical contrel necessary for acceptable resulis.
Additional information will be gotten on the compatibility
of Hastelloy N in steam, however, the major program for this
study is a part of the materials development program.

k. Circulating bubbles would be added to the salt loop to
see 1f any reduction in overall heat transfer coefficient
results.

5. As one of the finsl experiments, one of the supercritical
steam tubes would be intenticnally ruptured to see 1f the
salt side pressure release system is adequate. This test
should yield valuable information on the effect of a
similar incident in the MSBE.

Figure 1 is a diagram of this loop.

The above program should provide an adequate demonstration of the
heat transfer performance of the beoiler superheater, even though it is
confined to two or three tubes. The entire boliler superheater will
be tested iscthermglly in the ETU. As in the case of all other heat
exchangers, its pressure drop, fluid induced tube vibrations and
overall heat transfer coefficient (salt to air or low pressure steam)
will be measured.

Heat Transfer Enhancement

 

As a long-range goal, heat transfer enhancement may be a method
of reducing capital cost and fuel salt inventory. Particular reference
is made to a tube shape devised by C. G. Lawson {ref. 15). Thisz tudbe
features a "wavy" surface and is unique in that inside heat transfer
coefficients have been doubled in such a way that Colburn's analogy
holds. That is to say, the inside heat transfer coefficient increazes
in the same proportion as the fluid friction factor. A given enhanced

tube will transfer the same quantity of heal as a longer smooth tube
ORNL-DWG. 66-TETOR

Heat Recovery Letdown Valve

  
   
  

Cooling Water

 

e e, e — i — i t— — — ———

-

A |
Condenser ——/

Variable Speed

Salt Pump With | Temperature

Pump Bowl Surge 1 Control Heater 77
|

  

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 |
Volume ‘ l
(Approx. 30 GPM at 60 PSI)
A~ l I
| 1. | |
Q) ’ l i i . Demineralized
I | | l Water
I | 4000 PSI |
ouper i l ‘ S
1DE . | urge
Critical , i I Tank & - Chemicadi
Boiler l ] ! Deaerator " Control
Test _—
cection ! | ; (_
(3 Tubes) | | | |
| | 1 | Atmospheric
| Heater (~1.5 Mw) l | 50-100 psI P Pressure
i
\A |
k \<§E£Eéjj f |
A8 0000804 ! : =
X ]
E Jal.
reeze Valve ?PTeeze Valve Boiler Peed Pump
(0-12 GPM Room Temperature Water)

 

| -Fill and

[\_;:f Drain Tank

 

Fig. 1. ©Super Critical Heat Transfer Test of Prototype Tubes

¥C
if both have the seme flow and pressure drop. Both tubes would require
the same pumping power. The tubes tested so far are of brass, and the
"waves" are spiral grooves produced by drawing the tubes through a
planetary swaging head contalning four balls or needles. The indentions
left by the balls sre the spiral grooves. Tests have besn conducted only
with water. Tubes of this type should be useful In any hest exchanger

in the reactor system including the turbine condenser, but would be very
important in the fuel to coolant salt heat exchanger in providing a
significant decrease in the fuel inventory.

A study will be made of the.use of the enhanced heat transfer in

the fuel heat exchanger and 1f the concept is compatible , then measure-
ments will be made using salt in the special tubes. ~

PRESSURE RELIEF IN COOLANT SYSTEM

 

One of the purposes of using an intermediate salt to separate the
fuel and steam systems is to prevent the high-pressure steam from leaking
into the highly radioactive fuel region and creating a gross rupture cf
the low-pressure pilpe or vessel walls. A pressure velief system is pro-
posed for the coolant salt side of the steam generator and reheater to
handle the total flow of steam resulting from the rupture of several
steam tubes without putting an excessive pressure on the heat exchangers.
There is no eguivalent reguirement in the MSRE.

This problem can be mostly solved by design studies and tests using
compressed gas and water. However, some tests may be reguired of the
discharge of supercritical steam into coolant salt and the release of
steam and salt mixtures through rupture discs and blow-down lines.

 

DRAIN AND STORAGE TANKS

The drain and storage tank system provides for the szafe storage of
the fuel, blanket, and coolant salts when they are not in the circulating
systems. In addition to mainteining the salts above the ligquidus tem-
perature, there are provigions for removing the after-hest from the fuel
and blanket salts and for maintaining an inert cover gas overpressure to
prevent the inleakage of moisture and oxygen. A method for determining
the level of salt in the tank is also provided. '

Drein Tank System for the MSRE

 

Four tanks are provided for storage of the salt mixtures when they
are not in use in the fuel- and coolant-salt circulating systems of the
MSRE. Two fuel-salt drain tanks and a flush-salt tank are connected to
the reactor by means of the fill and drain line which contains freeze
valves to control the galt flow. One drain tank is provided for the
coolant salt.
26

A fuel drain tank is 50 in. in diameter by 86 in. high and has a
volume of about 80 £t3, sufficient to hold in a noncritical geometry
all the salt that can be contained in the fuel circulating system. The
tank is provided with a cooling system capable of removing 100 kw of
fission-product decay heat, the cooling being accomplished by boiling
water in 32 bayonet tubes that are inserted in thimbles in the tank.

The flush-salt tank is similar to the fuel-salt tank except that
it has no thimbles or cooling system. New flush salt is like fuel salt
but without fissile or fertile material. It is used to wash the fuel
circulating system before fuel is added and after fuel is drained, and
the only decay heating is by the small quantity of fission products
that it removes from the equipment.

The coolant-sall tank resembles the flush-salt tank, but it is
40 in. in diameter by T8 in. high and the volume is 50 ft3.

The tanks are provided with devices to indicate high and low
liguid levels, with weigh cells tc indicate the weight of the tanks and
their contents, and with thermocouples to indicate the temperatures at
several points on the tank surface.

These tanks are installed in insulated furnaces which are heated
with electrical resistance heaters. The experience with the electrically
heated furnaces in the MSRE has been very good. A test unit was brought
up to a temperature of 1200°F in 1963 and has operated essentially without
interruption for 25,600 hours. The fallure of an electrical lead, which
occurred shortly after the startup, was traced to a poor weld at a
Junction. This junction was redesigned to permit inspection and no
further trouble has been encountered from either the test unit or the
heaters in the drain tanks of the MSRE.

Another problem encountered with the MSRE drain tanks was the
startup of the steam cooling system. In the standby condition, the
bayonet tubes are at the drain tank temperature of greater than 1000°F.
To start the cooling system, water is allowed to run down the center-
most tube, cocling as it goes. There is no method in the present system
of reducing the thermal shock to the tube that results from this sudden
cooling. The tests run on MSRE prototype cooling tubes demonstrated
(1) that there was no significant hazard even if one of the tubes did
fail since it was easily detected, and (2) that there were sufficient
thermal cycles available before feilure to operate through the expected
life of the MSEE.

Drain Tank System for the M3BE

 

The fuel, blanket, and coolant systems will require separate drain
or storage areas. Because of critical mass considerations, the fuel
storage area will contain several tanks each of which will have a cooling
system that may be similar to the one in the MSEE drain tank system.
Although this cooling system would not be necessary if the fuel could be
held in the heat exchanger-fuel pump system for after heat decay, it would
not be practical Tc delay a maintenance operatiocn just for this purpose.
Therefore, the storage tank cooling system must have sufficient capacity
to take care of the heat exigting shortly after reactor shutdown.

The blanket salt system does not have an automatic drain feature
in the circulating system but will have a drain tank system similar to
the fuel storage system. The amount of after-heat removal needed will
be considerably less than for the fuel system but scme provision must
be made. The coolant salt drain tank area will be separate from the fuel
blanket salt storage area to take advantage of the lower temperature
required for the coolant salt. The inventory for both the blanket and
coclant salt is large enough to require several tanks.

The design of the drain tank system for the MSBR hag not been
carried far enough to give good definition to the development program.
If the MSRE drain tank system can be used, a2 new method of start-up and
operation must be developed. The major thermal shock in the MSRE cocling
system occurs when the water is first introduced into the centermost
tube thereby suddenly reducing the tube temperature from above L000°F
to the operating temperature of sbout 212%. A more gentle startup could
be accomplished by repiping the system to permit the use of low pressure
steam initially,and then by reducing the steam quality a transition to
water could be made. The use of wster as the normal opersting coolant
would permit the use of a storage system for emergency cooling.

If a different system is devised for the MSBR, some development
and testing, as yet undefined, will undoubtedly be required. TIn any event,
the test stand used for the MSRE coolers will be reactivated or a new test
stand will be built for this program.

Another area which must be investigated is the method of maintain-
ing the drain tanks at temperature. The system in the MSRE uszed z
furnace for each tank, removable heater-insulation packages for the
piping, and space coolers to remove the heat which leaks through the
insulation, thereby maintaining a low temperature in the drain cell.
One advantage of this arrangement is that it allowed much leeway in
the choice of materials and equipment that would be used in the cell.
The disadvantages are that a failure of the space cooler in the cell
would permit the cell temperature to rise excessively and that the
heater~insulation packages involve many intricate shapes which are
difficult to fabricate and maintain. The MSBE proposes Lo ingtall the
tanks and piping in hesated cells, one cell for the fuel and Wlanket
drain tanks and one cell for the coolant salt tanks. In addition to
the cell wall, insulation and containment membrane, thers are cell
heaters, vessel supports, instrumentation, and service line disconnects
all of which must be examined for compatibility with the high tempera-
ture enviromment. Many of these problems are common to those found in
the reactor cell. |
28

VALVES FOR MOLTEN SALTS

 

Means of controlling the flow of molten salt will be needed in all
of the systems of the MSBR. The requirements range from absolute shut-
off of the flow in the £ill, drain, and transfer lines of the fuel,
blanket, and coolant systems to the throttling of tne flow of the ccolant
gsalt to the steam reheater and to the boiler-superheater. The coolant salt
flow will be normally 14,600 gpm to the boiler superheaters and 2200 gpm
to the reheaters. The control scheme for operating the plant during
startup and other off design conditions has not been firmly established,
but it is estimated that the throttling range may be down to about
20% of the full flow but in no case will it have %o go to zero flow.
Methods of achieving these requirements include variable flow pumps,
mechanical throttling and cutoff valves, frozen salt cutoff valves, and
balanced pressure barometric legs to provide and maintain flow interrup-
tion. A brief discussion of the status of these methods is given below.

Some development work was done on an emergency drain or dump valve
for use in the molten salt of the Aircraft Reactor Test. This valve con-
sisted of a spherical metal plug which fitted into a similar metal seat
to form a seal. The leakage permitted through this valve was less than
1 cc/hr. The positioning of the plug was accomplished by the movement
of a stem through a bellows seal. The guiding was done on the helium
gas side of the bellows. Among the accomplishments of this work was
the development of a method of attaching a Kenametal plug and seat to an
Inconel stem and valve body. One of the difficulties encountered was
the tendency of the valve to stick after an extended period in the closed
position. Although the valve tended to stick even after the Kenametal
was adopted, there were indications that the sticking then was due to
galling of the exterior guides. Unfortunately, thils program was termi-
nated before it was completed and the question was not resolved. During
the tests in connection with this development, none of the valves would
consistently leak less than 10 cc/hr, .although some of them did approach
this.

Valves for use with liquid metal systems have been developed and
successfully demonstrated. One feature of these valves which prevents
their being adopted directly for use with molten salt is the frozen
metal shaft seal. Static seals using frozen molten salts have been used
with freeze flanges, access ports, and freeze valves, however, the glass-
like character of the frozen salt makes its use in a dynamic seal
doubtful.

An alternate to the mechanicsl throttling wvalve for controlling salt
flow would be to provide variable flow salt pumps, one for the steam
recheater, and a second for the boiler-superhester. While it appears
that such a scheme is feasible, there are other considerations which
make the use of a mechanical valve more attractive for controlling the
salt flow to the steam reheater, and a variable flow pump for circulat-
ing the coolant salt.
29

O

d life because
of the stress produced in thermal cycling during the freeze-thaw opera-
tion, these valves have performed well and should be adaptable for use
in the MSBR. The thermal cycling problem will have Lo be studied.
Briefly, the valve consists of a flattened section of pipe, cooled at
the center by an insulated air stream, and & heated furnace which
encloses the entire valve. This arrangement ccontrols the manner in
which the salt freezes and thaws 530 as to prevent damage resulting from
trapped liquid expansion during the thaw. Modes of operation possible
for these valves include: (1) open, (2) closed but ready to open on s
power failure, (3) closed and to remain closed on a power Tailure, and
(4) closed but ready to open rapidly (less than 15 minutes) on demand.
The time required to freeze a valve varies from 5 to 30 minutes depending
on the starting temperature and the time required to thaw varies from
less than 15 minutes up to several hours depending on the mode of
operation at the time of the demand.

Another method of controlling the flow is the use of the balanced
pressure barometric leg to provide and maintain the flow interruption
except when g controlled transfer 1s needed. Briefly, the flow is
controlled by differential gas pressures between the source and the
receiver and the guantity transferred is limited by the volumes available
to the transfer lines. This system is particularly attractive for use
in eliminating the thermal stress associated with freeze valves where
transfers are very freguent but involve only a small guantity of salt
in each transfer.

The development program initially will consist of a design study to
determine more exactly the requirement of the:system for flow control.
A mechanical throttle valve for use in high flow streams will be
degigned and tested in connection with one of the salt pump loops. A
nechanical valve for the shutoff of small flows will be designed and
studied for possible use in itransfer lines. The development of the
valve operator will be a part of this program. A study will be made of
the £ill, drein, and transfer flow rate regquirements to determine the
size of wvalves needed and the appropriate freeze valve will be designed
and tested. Finally, a study will be made of the use of balanced
pressure method of flow control to determine the limitations and if the
problems are regsonable, a system for controlled transfer will be
developed for process systems.

SALT SAMPLERS

The salt sampler consists of a mechanism for lowering a capsule
into a free surface of the salt from which the sample is to be taken,
Since the sample capsule must pass through both the primary and secon-
dary containment boundaries, it is necessary that suitable valves,
accegs ports, and transport mechanisms be provided to prevent the
accidental release of activity. In addition, it is necessary that a
30

high level of reliability of the mechanisms involved be established,
and that a reliable maintenance scheme be devised to replace all com—
ponents of the sampler system with a minimum delay.

The sampler in the MSRE contains three primary sesl closures
through which the sample is transferred in sequence. The first closure
consists of two gate valves with helium buffer and leak detectors
at the valve seats. This closure isolates the sampler from the tank
that is being sampled except when the sample is actually being withdrawn.
These valves are replaceable but require that the reactor be shut down
during replacement. '

The second closure consists of a single door with a helium buffered
and leak detected seal. This door provides access to the capsule
lowering mechanism and is opened only after the first closure is closed
and sealed. The sample capsule is inserted and removed through this
door with a hand~operated mechanical manipulator.

The third closure is a ball valve with a helium buffered and leak
detected seal. A cylindrical transport container is lowered through this
valve, the sampler installed, and the sealed container is removed into a
shielded carrier.

The capsule lowering mechanism, which i1s between the first and
second closures, consists of several electrically operated components
including a motor and a cable drum with cable and capsule latch.

In general,the sampler system has operatedsatisfactorily through
the removal of 239 samples and the insertion of 6 enriching capsules.
The only difficulty with the system, which required complete shutting
down of the reactor to repair, cccurred when an electrical connection
on the innermost closure opened due to a shorted connector. This
equipment was designed to insure that it could be maintained but with no
strong emphasis on speed of maintenance. Approximately one week was
required to disassemble, repair, and replace the faulty item, under
conditions of medium radioactive contamination (60 R/hr at surface of
the component).

A more objectlve apprcach is needed for the maintenance of the
sampler for the MSBBE. First, it should not be necessary to shut down
the reactor to repair the sampler. Secondly, the time required to
repair any portion of the sampler system should be less than the reguired
sampling interval. With the experience gained from the MSRE sampler as
a guide, the entire sampler design will be reviewed for changes which
will improve the maintainability and the success of these changes will
be demonstrated in a prototype sampler to be used on the ETU.

One simplifying feature of the MZBE is that the sampling need not
e carried out in the reactor cell. Since a portion of the fuel and
blanket stream will be bypassed continuously to the chemical process
plant, the sampling station might be installed in a cell adjacent to the
analytical facility such that the salt coming from the reactor to the
Process

31

plant and from the process plant back to the reactor could be

analyzed for material control and inventory purposes. Development needed
is to enable one sampler to serve several stages in the chemical process

plant.

GAS SYSTEM

The design of the MSBR gas system will be based in part on experience
gained from previous liquid fuel reactors, the ART, the HRT, and the MSRE,
and in part on design criteria peculiar to the MSBR.

As

1.

in the MSRE, the MSBR gas system will:

Strip xenon-135 from the fuel system.

Dispose of radioactive fission gases and daughters.
Protect the salt from oiidizing atmospheres.
Provide motive power for salt transfér.

Furnish miscellaneous services such as purging, venting,
and leak detection.

Compared with the MSRE, the MSBR cover gas supply system will have
a larger capacity and will utilize helium recycled from the offgas

system.
features

1.

The MSBR offgas system plans or contemplates the following
which were not a part of the MSRE system.

Rapid stripping of xenon-l135 from the fuel system
by injecting helium at the heat exchanger inlet

and removing xenon-enriched helium by means of a

gas separator downstream of the heat exchanger. The
helium from the separator will be returned to the
fuel stream after passing through a holdup system
designed to reduce the concentration of Xe-13%5 by a
factor of LO. -

Removal of Kr-85 and tritium from the charcoal bed
effluent, with collection, recompression and reuse of
the helium, and concentration and storage of the Kr-85
and tritium. '

 

MSRE Cover and Offgas Systems

A detailed description of the MSRE cover and offgas system may be

found in

the MSRE Design and Operations Report.Z Only a brief descrip-

tion will be presented here.
32

A helium cover-gas system protects the oxygen-sensitive fuel
from contact with air and moisture. Commercial helium is supplied in
a tank truck and is passed through a purification system tc reduce the
oxygen and water content below 1 ppm before it is admitted to the
reactor systems. A flow of 200 ft3/day (STP) is passed continuously
through the fuel pump bowl fto transport the fission product gases to
activated charcoal adsorber beds. The radioactive xenon 1s retained
on the charcoal for a minimum of 90 days, and the krypton for T7-1/2
days, which is sufficient for all but the 85Ky to decay to insignifi-
cant levels. The 8%Kr is maintained well within tolerance, the
effluent gas being diluted with 21,000 c¢fm of air, filtered, monitored,
and dispersed from a 3-ft-diam by 100 £t high stack. Some tritium is
also produced in the reactor. Its Dbehavior has not been studied but
we assume that it too is discharged to the atmosphere. The quantity
produced is such that the concentration in the air does not exceed AEC
limits.

As described in the section on pump development, the MSRE pump
also contains an arrangement for spraying the salt Into the gas space
of the pump in fine streams such that the gas 1s permitted Lo separate
from the salt. Experiments conducted on this arrangement in a small
test loop indicated that the efficiency of this separation would be
less than 15%, probably due to the extremely short contacting time
available in the gas space. The experience gained in the MSRE has
indicated that the apparent efficiency may be as high as 75%. It is
believed that the apparent higher separation efficiency is the result
of a small volume fraction of gas bubbles circulating in the salt. On
the average these bubbles make 20 trips around the fuel loop and then
pass through the spray ring into the pump bowl where they can separate
from the salt. The xenon concentration in the bubbles should be in
eguilibrium with the concentration of xenon dissolved in the salt.
Since the solubility in salt is very low, most of the xenon will be
in the bubbles. This type of bubble-liquid contactor is proposed for
use 1n the M3BR except that a bubble will not reside in the loop for
more than 1 circuit.

The cover gas system is also used to pressurize the drain tanks to
move molten salts into the fuel and coolant ¢irculating systems. Gas
from these operations is passed through charccal beds and filters before
it 1s discharged through the offgas stack.

MSRE Gas System Performance

 

Performance of the cover gas supply and distribution system has
been satisfactory. Buildup of oxide in the fuel salt during
10,000 hours of operation has been negligible, indicating that the
purity of the helium gas has been adeqguate. The helium purifier
system, the pressure and flow controls, and the leak detector system
have offered no mechanical problems.

Minor difficulties have been experienced as follows:
- 33

Gas purification -~ the indicated moisture content of
the purified gas has varied from less than the 1 ppm
control point to as high as 10 ppm. It is not known
if the variation is real or perhaps is a systematic
error in the analytical Iinstrument. Some development
work is under way to evaluate this problem for the
MSRE.

Back diffusion - pressure variations in the drain
tanks caused back diffusion of activity into the
helium supply line with resultant high radiation
level in the adjacent work area. The situation was
corrected by providing a small, continuous purge
of clean gas through the lines. '

Rupture discs - the supply line rupture disc (protects
salt vessels from excess pressure) failed several
times during the pre-nuclear operating period. The
trouble was ascribed to the lower ratio of rupture to
operating pressures, and the problem was corrected by
reducing the header operating pressure.

Delay tests were run on the MSRE charcoal beds during the pre-
nuclear pericd. The krypton delay time was found to agree with the
predicted value within acceptable limits. This fact, coupled with
experience during MSRE power operation, indicates that the basic
technology of charcoal bed design is adequate.

The following difficulties were experienced with the MSRE offgas

system:

1.

Foreign matter, which resulted from inleakage of oil
from the fuel pump lubricating system, caused plugging
of control valves, flow restrictors, and filters. The
resulting loss of control of pump bowl over-pressure
and off-gas flow caused several interruptions in reactor
operations. The problem is being corrected by modifi-
cations in fuel pump design and by the use of filters
of improved design. The investigation of this

problem yielded valuable experience in the design of
filters for radiocactive gases (e.g., the importance

of adequate heat transfer), and in the hot cell
examination of components. The work also served to
emphasize the potential problem of salt-mist carryover
in the offgas. '

A flow restriction resulted in the drain tank vent
header when the poppet in an in-line relief valve
became disengaged from the valve body and became
lodged in the piping at the inlet to the auxiliary
charcoal bed. '
34

3. Control valves, which operated in the very low flow
of 4 liters/min of dry helium, tended to gall in the
sliding surfaces between the valve plug and the valve
seat. This problem has been reduced by using dif-
ferent materials and by opening the clearances between
the plug and seat.

4. Intermittent plugging occurred at the inlet to the main
charcoal beds. This trouble, which has persisted
throughout the power operation, was at first relieved
periodically by blowing back from the charcoal bed
into the keoldup volume. As a result of hot cell
examination of offgas system components, the difficulty
was ascribed to accumuations of polymerized organic
solids. In recent months the rate of plugging has been
appreciably lower, possibly due to the improved filters
mentioned in 1. above, and restrictions, when excessive,
have been cleared by heating the charcoal bed inlet
lines to about 800°F.

M5BR Offgas System

 

The offgas system proposed for the MSBR is shown schematically
on Figure 2. The fuel system offgas stream 1s a combination of helium
flows from the gas separator and from the pump bowl vapor space. The
pump bowl effluent is a collection of miscellaneous purge gases from
places such as the pump shaft annulus and the level indicator instrument
bubblers. The two offgas streams combine at the cyclone separator where
entrained salt i1s removed. The offgas stream then combines with similar
streams from the other three reactor modules,¥* and the total offgas
stream passes through the 48-hr holdup system where the gas is filtered
Lo remove solid daughter products and the xenon-135 concentration is
reduced by radiocactive decay to sbout 2-1/2% of its initial value. At
the outlet of the 48-hr holdup system, the offgas is split into two
streams. One, comprising about 96% of the total flow, is returned to
the fuel system at the fuel pump sucticon line by means of a gas injec—
tion system. The injection rate is calculated to maintain the volume
fraction of gas at the point of injection equal to 1% of the fuel flow.
This value for the circulating void fraction is tentative, however,
pending results of development studies to be made on the diffusion rate
of xenon from the salt to the circulating bubbles.

The balance of the fuel system offgas flow 1s combined with blanket
pump offgas from the four modules and the combined stream passes in order
through long delay charcoal beds, a tritium removal system and a
krypton-85 trapping system, and is then recompressed and returned to the
purge gas supply systen.

¥The term module is used interchangegbly in this discussion to represent
one heat exchanger and pump circuit in the 1000 Mw(e) MSBR of reference
design or one complete reactor system in a 1000 Mw{e) MSBR plant contain-
ing four reactors of modular designs.
e u 5 » e ',

 

BLAMNKET |
PUME

 
 

&

 

MISC. PURGE GAS

 

 

 

 
 
  

 

  

 

 

   
  

 

   

 

 

 

 

 

 

 

 
 

 

    
  

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 
  
  
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o \coouwe WATE'R-\ CHEM. PROCESS PURGE —m,
\ HELIUM i o
o &P_P,u COOLANT UMP PURGSE
HOLD-uP ' o e
N l , BDS."““T_]/ S ALLUMULATOR
! 1 5 / o
o / 3o
FILTERS &
PRE-TRZAT
“CCOLING
i T WATER
|
- ) C :
U o NEADER
SEPARATOR - - HOLDUP ¢ t— Y
FISSIGn T._H CHARCLAL
JET Pump PRCGOULTS | BI0
¢y
[
i " .
- Kr-83 &£ TRITIUM
{ } ] REMOVAL . .
A o ! 5 U:?
— . 1 STMAINTENANCE ' ~ v
Ty CONTAINMENT ASUINE TRASBD
48 HR. HOLDUP (HARCOAL BEDS VENTILATION FARS FILTERS STACK
B —— ——— TR
— DATA —+—
FOINT FTSEC PSIA TEMPF POINT SCFM PSIA TEMPF POINT SCEM BSIA TEMPE OTE -
J | 9 2.00 (80 100 TUTAL SEAT LOAD ON 48HR DELAY
g 25 27T 1000 N 675 24T 1000 8 2.00 165 70 SYSTEM IS~ 5 Mw.
% 2700 240 1000 9 2.00 150 70
T 27.00 23¢ 230 & 2.00 447 70
B 26.00 180 230 2y 200 47 1o
25 641 160 A3 €.50 70 1000 85 028 2477 —
o 25 457 1006 % Q25 247 — % S0MAX. (TTmAy, 100
3 25 327 1000 G 025 247 1000 B SOMK {T2ZMAX 100
7 025 — — ® 50 MAx 147 70
7 1.00 — -—
Figo 2

MOLTEN SALT BREEDER REACTOR
(4) 250 MWrfe) MODULES
GAS FLOW DIAGRAM
36

Thus, the offgas system consists of two recirculating loops, one a
high flow system for rapid removal of xenon-135 from the reactor fuel
stream, and the other a low flow system for removal of long-lived activity
and recovery of helium. Current estimates of the various flow rates are
as follows:

Flow Rate, scifm Helium
Cne Module Four Modules

 

 

Input to Fuel System

Stripper Gas 6.5 26

Miscellaneous Purge 0.25 1

Total Input 6.75 27
Effluent from Fuel System

Gas Separator 6.5 26

Miscellaneous Purge 0.25 1

Total Effluent 6.75 27
Throughput for 48 hr Holdup System ——— 27
Stripper Gas Recycle 6.5 26
Throughput for Blanket System 0.25 1
Throughput for Long Delay Charcocal Bed ——-- 2
Throughput for Tritium and Krypton —_——— 2

Removal Systen

Clean Hellum Recycle e 2

 

The anticipated temperature and pressure levels at various points in
the system are tabulated on Figure 2.

In addition to the main offgas system, provision is made for handling
the following gas streams:

1. Intermittent venting of large volumes of helium from the
fuel blanket and coolant system drain tanks.

2. A continuous flow of about 0.8 scfm of helium from the
coolant pumps. This stream will contain some short
lived, induced activity plus a high partial pressure
of BF3.

3. A more-or-less continuous low flow offgas siream con-
taining gaseocus fission product activity from the chemical
processing plant.
37

These gases are pretreated to remove undesirable contaminants such
as BF3 and fluorine and then dirvected through a high flow, auxiliary
charcoal bed. Depending on the outcome of future feasibility studies,
the effluent from the suxiliary charcoal bed may be diluted with air
from the building ventilation system and discharged to the atmosphere
through an iodine trap and absolute filter system, or it may be purified,
recompressed and returned to the helium supply header. The present flow
sheet shows this material as being vented to the stack.

MSBE Offgas System

 

The MSBE offgas system will be a scale model of the proposed MSBR
system and as such will serve as a final proving ground to check the
adequacy of the individuval components and of the integrated system.

Since the fuel flow rate in the MSBE is about 40% of the flow in one MSBR
module, components common to each module, such as the gas separator and
the gas injector, will be about 2/5 MSBR scale. The scale of the other
components will vary from 1/t to 1/10 as shown on Table 3.

Table 3. MSBE Offgas System Scaling Factors

 

 

Svstem Flow Rate MSBE
Y MSBR  MSBE Scale
Fuel, gpm 11,000% LLo0 2/5
Gas injection & stripping, scfm 6. 5% 2.6 2/5
Cyeclone separator:

Gas, scfm 6.5% 2.6 2/5

Salt, gpm S50% 20 2/5
48-hr holdup system, scf 27 2.85 1/10
Biological charcoal beds, scfm 2 0.5 1/4
Kr-85 trapping system, scfm 2 0.5 1/h
Helium compressor, scfm 3 0.75 1/b

 

*¥Per reactor module.

Development Program

 

Gas system components which reqguire development are the gas separa-
tor, the salt powered gas injector, the cyclone separator, the filter
system for fission products, the gas compressor, the 8%Kr stripping
system, the tritium removal system and the gas sampler.
38

Gas Separator

The gas separator will be located in the fuel pipe between the heat
exchanger outlet and the reactor inlet. In this region the reactor inlet
line surrounds, and is concentric with, the reactor outlet line, forming
an annular duct. The geparator for the MSBR will consist of about
15 identical vortex bubble separator units of the same diameter as the
annulus width and distributed about the annulus as shown in Figure 3.

The separator for the MSBE will use the full-scale MSBR separator unit,
but only six of the vortex units will be required because of the lower
salt flow rate.

For the MSER, ezch individual unit should have the capability of re-
moving all bubbles larger than 0.01 in. diam and the capacity to separate
0.4k sefm of gas from a flow of T50 gpm of salt. The gas takeoff lines
will be connected to a ccmmon header leading to the cyclone separator.
About 50 gpm of salt, equal to about 40% of the total volume flow will
be discharged to the cyclone separator along with the gas.

A considerable amount of theoretical and experimental work was done
in connection with the gas separation problem in the agqueous homogeneous
reactors.16-17,18,13 The progrem will include (1) application of
those results to the design of a full-scale unit, and (2) proof testing
of a prototype unit in water and salt loops.

Mass Transfer to Circulating Bubbles

 

One of the uncertainties in estimating the efficiency of the proposed
system for rewoving 135%ec and all other noble gases from the system is
the rate of mass transfer from the fuel salt to the circulating bubbles.
This rate is intimately associated with the mass transfer coefficient to
the bubbles. Only a moderate amount of information on this mechanism
ig available from the literature and generally it is not directly appli-
cable. A program involving a water loop and some relatively insoluble
gas (perhaps oxygen) will be used to measure the mass transfer coefficient
to small bubbles circulating in a turbulent stream. This information
will then be extrapolated to the fuel salt-helium bubble system by mass
transfer analogy. It is hoped to be able to confirm these results in the
gas system test loop and the engineering test unit.

salt-Powered Injector

 

The Xe~135 stripping system requires injection of helium into the
fuel stream at the heat exchanger inlet. The required gas flow is about
6.5 sefm. The available gas head, which is the pressure at the ocutlet
of the 48-hr holdup system, is about 2 psig. The static pressure of
the salt at the point of injection is ahout 9 psig. We believe that the
gas injection can be effected by utilizing the energy in the salt stream.
The device might be either a commercial, liguid powered gas injector or
a full-flow venturi section. As a backup, the possibility of using a
mechanical compressor will also be investigated. The program will include
small-scale tests to evaluate the alternates, followed by construction
    

FROM M. 44x

BEPARATOR UNT (1)

 

 

 

 

 

 

 

 

 

 

 

 

PLENUM.

SEPARATOR UAIT.

LRPARAGUR UNMT
WALE = 1T

—GAS  PILKYP TUBE.

YORTE ™,
40

and testing of an MSBE prototype unit. The main problems will be to

demonstrate (1) ability to perform at the prescribed MSBE design con-
ditions, (2) compatibility with the overall gas handling system, (3)

adequate service life and (4) replaceability.

A cyclone separater is provided to remove salt which is entrained
by the gas coming from the gas separator. The underflow is returned to
the fuel stream by way of the pump bowl. The design of the unit will
be based on existing technology, but this work will be deferred pending
development of data on the operating characteristics of the gas separator.
The program will consist merely of proof-testing the final design.

Salt-Mist Removal

 

Offgas system components must be protected from salt mist. A program
will be required to determine the nature and severity of the salt-misi
problem, and to develop a satisfactory entrainment separation system.
Tests will be conducted in a loop incorporating flow of fuel and offgas
at design conditions. The program will include (1) characterization of
offgas contaminants and (2) proof-testing of various possible mist
trapping devices. The MSBR mist trap will be located at the outlet of
the cyclone separator.

The problem with the entrainment of a salt mist in the gas lines
of the MSRE was first encountered in one of the offgas lines coming
from the pump bowl. The gas flows from the pump bowl at about 4 liters/
min and travels through a holdup volume of 170 liters before entering a
filter and valve assembly. It was found that very fine particles of
salt were carrying through the holdup volume, going through the filter
and depositing in the copening of the valve ultimately causing a flow
stoppage. ©Since the quantity of salt was very small and did not seem
to present other problems, the difficulty was solved by changing the
pore size of the filter from 25 u to one closer to 10 u. At the same
time tests were initiated to characterize the particles coming from
the bowl of a similar pump being operated in a test loop. It was found
that the particles were formed as a mist of the circulating salt and
were not the resull of condensation of a salt vapor. The proposed
solution to this problem, which is to be tested for use in the MSRE replace-
ment pump, is to install a heated metallic filter element in the pump bowl
such that any sall caught on the filter will run back into the pump
bowl. The MSB# could potentially have a more severe problem since the
gas leaves the pump bowl at about 30 times the rate for the MSRE. A
similar sclution is proposed for the MSBE but it must be tested at
the higher gas flow.
41

Fission Product Filter

 

At a power level of 2200 Mw thermal, fission products will be pro-
duced at a rate of about 0.1 y gm-moles/day, where vy is the yield in
per cent. BSince yields of kryptons and xenons range up to 6.5%, the total
production of xenon and krypton will be on the order of several hundred
grams per day (see Table 4 ). Except for Kr-85, essentially 100% of the
radicactive kryptons and xenons will decay to solid daughter products
which will deposit in the reactor system or at some point in the offgas
system. The location and rate of deposition will be a function of the
decay constant and the local flow rate and geometry. In the MSBR, a
filter system is required upstream of the 48-hr charcoal beds to remove
solid daughters from the offgas stream. The purpose is to minimize the
deposition of solids in undesired places, such as on valve trim, and to
minimize the carry-over of solids into the charcoal beds. The equivalent
system for the MSBE will be designed utilizing experience gained in the
design and operation of filters for the MSRE offgas system. The prime con-
siderations are (1) adequate heat dissipation, (2) adequate service life
and (3) ease of repair or replacement.

Table 4

Production of Gaseous Fission Products in M3BR¥*
Power Level - 2200 Mwt

 

 

Isotope, i (tl/2)i | Yi egms/day
Kr 87 78 m 2.7 23
Kr 88 2.77 hr 3.7 33
Kr 89 3.18 m b.6 41
Kr 90 33 s 5.2 L7
Kr 91 9.8 s 3.7 34

178

Xe 133 5.27 4 6.5 86
Xe 135 9.13 hr 6.2 83
Xe 137 3.9 m 5.9 81
¥e 138 ' 17T m 5.5 76
Xe 139 L1 s L.7 65
Xe 140 16 s 3.7 _o2
443

Total Kr and Xe 621

 

#¥Kr-85 and short lived (t 1/2 < 5 secsg) are not shown.
Kr-85 and Tritium Removal System

 

At a power of 2200 Mw thermal, the MSBR will produce an estimated
2000 curies/day of Kr-85 (10.27 yr half-life) and 2800 curies/day of
tritium (12.26 yr half-life). About 25% of the tritium will be produced
in the blanket. Since there is no information available at present
which would indicate an alternate fate for the tritium, it is assumed
that it, as well as the Kr-85, will pass essentially undecayed through
the charcoal bed retention system. A cleanup system will be provided
which will reduce the Kr-85 and tritium to a level which will permit
heljum to be returned to the purge gas system or vented to the stack.
The current concept of the cleanup system provides for a series operation
wherein the helium will pass first through a tritium removal station and
then through a cryocgenic absorber to remove Kr-85. The Kr-85 system will
be of the regenerative {type, with periodic transfer of the stripped
material to storage cylinders.

The development program will include:

1. An evaluation of possible methods of tritium removal followed
by mockup and testing of selected units. Two systems are currently under
consideration: (1) an oxidizer, such as hot Cu0, followed by a chemical
trap to remove the %fifi)and (2) a2 chemical reactor which will bind the
tritium in the form of a hydride.

2. Mockup and testing of the Kr-85 removal system.

1t is planned to use the MERE offgas system as a test station for
this work.

Gas Compressor

The helium effluent from the Xr-85 trapping system will be recycled
to the purge gas header. A compressor will be required to recompress
the gas to about 30 psig in order to permit adequate flow control to
the various purge and bubbler linez. Operating conditions for the MSBR
are: flow - 3 scfm, inlet pressure - 0 psig, discharge pressurc -~ 30
psig. One of the main problems will be to provide a system which will
not add contaminants to the gas. Tt is anticipated that regquirements
will be met by a commercially-available diaphragm compressor. The pro-
gram will include (1) a survey and paper cvaluvation of commercial units
and (2) proof-testing of selected units both on test stand and as part
of the integrated gas system.

Gas Sampler

Control or check samples will be required from the recirculating
gas and the main offgas stiream. Analyses of these gas samples will be
used to provide additionsl information about the chemistry of molten-
salt reactors. The major problem will be to develop and proof-test
adequate sampling techniques. The offgas sample unit currently being
installed at the MSRE will provide the basic technology for this work.
L3

MSBE Gas System Test Loop

 

It will be necessary to demonstrate the adeguacy and reliability of
the xenon stripping and offgas system components when operating as an
integrated system. ©OGuch a demconstration can best be made by means of a
test system incorporating a szalt circulating loop with suitable connec-—
tions for the various gas circulating locps. Figure U shows a schematic
diagram of the proposed lcop layout. Provisions are made for:

1. Gas injection and separstion.
2. Salt powered gas circulation.
3. Recompression-recycle system.

The salt loop contains an extra test section for materials testing
and a tank to simulate reactor holdup. The gas system contains test
sections for component test or component simulaticn. The sglt flow rate
will be 1000 gpm.

This loop will also be used to evaluate the performance of scale
models of the components. The net mass transport of the gas to the
entrained bubbles injected into the salt will be determined and the
efficiency for removal of the bubbles by the separator will ‘be measured.
The use of the salt driven exhauster as a prime mover for the gas
recirculation system will be demonstrated as will the effect of the gas
injection system on the stability of the liquid level, the stability
of the gas recirculation, and the xenon removal rate. In short, this
test loop will serve to proof test the concept.

CELL FURNACE AND SHIELDING

 

The cell containing the fuel and blanket salts is to be maintained
above the liquidus temperature of both salts (agbout 1040°F). The cells
containing coolant salt only are to be operated above the liquidus
temperature of the coolant (about TOC°F). However, there are times when
the coolant cell must also operate above 10LO°F so the furnace will be
designed for 1150°F. The cell temperatures are to be maintained by
radiant heating surfaces. Thermal insulation and cooling will be applied
as reguired to protect concrete, equipment supports, instrumentation, etc.

The concept of maintalining the entire cell at the elevated tempera-
ture differs from the method used in previous molten salt work. The
closest approach to the proposed method is in the thermal-shield reactor-
furnace arrangement in the MSRE. Here the reactor is installed in a
water-cooled thermal shield that has thermal insulation on the inner
surfaces. The heat for the furnace is supplied by electrical heater
vhich are inserted through the top of the furnace. The cooling system
in the thermal shield removes the heat deposited by the nuclear source
and also the heat that leaks through the thermal 1nsulation. This
system has operated satisfactorily except for the minor problem of
keeping radiolytic ges from accumulating in some parts of the thermal
ORNL DWG. 66-7671

 

 

Gas
1
Compressor !

 

 

 

 

 

Surge Tank

 

 

 

 

Gas Injection Loop
L sefm oo e Yr——Salt Powered Gas Injector Liguid Loop

 

| )

|

<— L8 hr Xenon delay
; Gas circulation loop
16 scfm -

' Simulated i
s L8 hr Ye- e m

Delay Sec.

% .
—TEm Salt Pump \\\_-
Mist Collection

3% Test Section

Shaft Purge
C.1 scfm

    

Gas Separator Loop F

L sefm Gas 50 gpm Salt ——1\\ |
] Simulated |,
{ Heat Exch. |

 

 

 

 

 

 

 

 

 

 

 

 

 

 

     
     

 

 

 

 

 

 

 

 

Simulated
3 Reactor =
Volume

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Salt

Storege
Tank I

 

Fig. 4. MSBE Gas System Test Loop.

1
L5

shield. This problem was at least partially solved by installing an
additional circulating pump to increase the flow through the water
passages and by msking provisions for the removal of the entrained gas.

One basic probiem which eveolves from this system results from the
necessity of using water as the heat removal medium. The maximum credible
accident for the MSRE involves the intimate mixing of salt leaking from
the fuel system into water leaking from the thermal shield system. Steam
is generated and this produces a high pressure in the reactor cell,

In the MSRE, a pressure suppression system is used to limit the pressure
and to contain the fission products should such an accident occur.

Studies will be made of designs of furnace and shield .combinations
for the MSBE with the objective of eliminating the possibility of mixing
of salt and water.

The heat for these cells might be supplied by gas or oil burners
mounted outside the cell but with the hot gases passing through the
cell in radiating fin tubes or eléectrical heaters mounted in re-entrant

tubes in the cell, The selected method and the proposed design will
require testing.. - . '

The arrangement for cooling the supports for the reactor vessel
mist be tested. The effect of temperature changes in the supports on
the vessel alignment must be examined under the conditions expected
during the system startup and during a failure of the support cooling
system.

Some small scale mockups will be built to give ald in design
arrangement and evaluation. The final checkout of the system will be
done in the IETU.

STEAM SYSTEM

The steam conditions in the MSBR steam cycle are almost identical
to those in the Bull Run Plant of the TVA except that a reheat steam-
preheater, a mixing tee on the feedwater line, and a feedwater pressure-~
booster pump have been added. The technology involved in these three
components is such that the design and construction should be straight-
forward. These units will be specified in sufficient detall so that
they can be designed and fabricated by an experienced manufacturer.
While some testing would be done by the manufacturer, the final testing
for performence and serviceability could be done elsewhere, perhaps at
the Bull Run Plant. Preliminary inguiries to vendors have been met
with sufficient interest and confidence to indicate that this approach
to the procurement problem is a reasonable one.
L6

TURBINE GENERATOR

 

A 60 Mwe turbine-generator is smaller than is normally supplied by
the major manufacturers for a 3500 psi-1000°F -~ 1000°F reheat cycle.
Here too, a preliminary inquiry indicated that such a unit is technically
feasible and could be made, however, interstage leakage in the small
high pressure end would reduce the efficiency. The cost for this small
supercritical pressure unit would run from 10% to 25% more than for a
comparable unit at a lower, more conventional pressure. It is desirable
to produce supercritical steam in order to reduce some of the thermal
stress problems in the steam generator. A study will be required to
decide whether to use a supercritical turbine or to throtile the steam
to lower pressure and use the more conventional unit.

In sunmary, it appears that the steam system can be tailored to fit
the MSBE requirements and that no extensive development 1s necessary to
produce the equipment or to insure that it meels the program require-
ments.

ENGINEERING TEST UNIT

 

The final demonstration of the components and systems for the MSBE
will be done in a full scale engineering test unit wmodeled after the reactor.
The components for this model are to be exactly the same as those intended
for use in the MSBE and in fact some of these units, after demonstrating
a satisfactory performance in the ETU, might serve as spare parts for the
reactor experiment. It 1s expected the experience gained in bullding
and operating the ETU will considerably shorten the startup time for
the MSBE.

The primary purposes for the Engineering Test Unit are three:

1. It will provide an integrated test at full temperatures, pressures,
and flows, but in an unradicactive environment of the reactor fuel,
blanket, and coolant systems: the fuel and blanket processing systems;
and their major auxiliaries.

2. The equipment will be the first manufactured of Hastelloy N by
commercial fabricators. This will provide an opportunity to qualify
the manufacturers to build reactor equipment and the modified Hastelloy N
as a structural material. The same manufacturers will make a second
group of components for the MSBE, and they should be of superior quality.

3. It will provide g plant for thoroughly training the operations
and maintenance staff and for thoroughly testing the maintenance equip-
ment and procedures before the startup of the MSBE.
REACTOR

VESSEL

  

} 963 ft¥sec
| ‘g

1050°

     

   
   

Y} BUANKET SALT HEAT
E B EXCH AND PUMPS
M 75 Mwil) -540 GPM

FUEL SALT HEAT

H25" g
{ EXCH. AND PUMPS [f
: 1425 Mwit) -4320
e __f te G

 

 

BLANKET SALT FUEL SALT
DRAIN TANKS DRAIN TANKS
LEGEND
FUEL e % —  i0fib/py
BLANKET e o aren P — psio
COOLANT —-==— h — Bt/
STEAM  ——=~=— A
M0 —{}- — Freeze Vaoive

Fig. 5,

  
   
 

 

    

 

 
 
     

ORNL DWG. 66-7383

   
  

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

   

     

COOLANT SALT
_ORAIN TANKS

  

 

Molten Salt Breeder Experiment Flow Disgram.

 

 

______________ ). e
feigsn
'850, 6844 486 #
100G f T T T 04
i i » 35150
' : 000°
= ———— v e e e e - i _
r ' ,; } ! - GEN
' i liagan | ]
| g 136000 | . TURBINE
: ! BOILER jio0o” | | —
| rencaters. | SUPE:*;E:TERS ‘ | 1256.h,
4 [ :
25 Mwt 594 i"_;‘_&&& _?_‘aw N
i25* z‘ BT o
N-a554 ' f | e ’
; f f ! TURBINE
.. : : g . ° = -
i | f' 13078 P !
oo :!all 1 1
! _BSOPE : REMEAT STEAM [gggop | Lo '
T PREHEATERS 4 It i
o CONDENSER 8 FEEDWATER
- A2 et : 6.7 Mw(t}
T | 1 COOLANT saLY rea2h | SYSTEMS
‘ A PUMP ;
: : t 3800p
| | 1 4340 oPm {ofwooe 54630
= bl 133350 4 A 4 13654 35009
o _tE 5709-650'!=_§ | — $509°
' BOOSTER
i PUMPS

L%
L8

The design of the ETU will be based on the MSBE flowsheet shown
in Figure 5. The test unit will be complete through the beiler-super-
heater and the steam reheater but will not include the supercritical steam
system. It will include those parts of the offgas system that are required
to inject gas into the circulating fuel stream, remove gas from the fuel
stream, and separate entrainment from the gas before it goes to the
carbon beds. The fuel salt, blanket salt, and ccolant salt draln tank
system will be included.

A pilot plant will be provided for processing the fuel and blanket
salts. How complete that plant will be has not been decided. The
equipment to be used for effecting transfers between the reactor and the
processing plant will certainly be included. We also expect to provide
equipment for fluorinating, distilling, and reconstituting the salts but
have not yet fully decided whether this is fully necessary.

We plan to install the EIU in an existing building, probably in
the Y-12 arca. This will involve removing existing equipment and
floors, installing the cell liners and cooling system, insulation, vessel
supports, and coolers, cell heaters and representative sections of water-
cooled concrete shielding. The method of penetrating the furnace-
containment wall will be used in representative locations such as salt
piping, instrument lines, and gas supply and vent lines.

The program will use the designs and drawings prepared for the
reactor and will prepare only those additional designs necessary to
adapt the reactor to the experimental installations. Much of the
vendor's development and testing of fabrication methods will be done
in building components for the ETU.

The systems will operate isothermally during most of their experi-
mental 1ife although it is planned that some representative temperature
differences across components may be generated by greatly reducing
the salt flow and throttling small quantities of steam out of the steam
generators. In addition, some thermal transients will be produced to
help in developing a startup procedure which will minimize the thermal
shock to the system.

As outlined in the program on heat transfer equipment, it is
planned that initially the individual salt loops will be operated with
water so that the hydraulic characteristics of the loops may be evaluated.
Since there will be considerable heat input capacity available to the
loops, it is planned that some heat transfer data will be obtained
during the periocd of coperation with water.

Another area, in which some preliminsry experience will be gained,
is the control of the coolant salt flow through the two steam generators
and reheater to help adjust the system for carrying reduced power loads.
As described in the design proposal, this will be done by varying the
coolant pump speed and the position of the valve which controls the salt
flow to the reheater.
49

The adequacy of the reactor and component support system will be
evaluated during thermal startup. The stresses at critical vessel
nozzles and penetrations will be measured with the system at reduced
temperature by reproducing strains measured during thermal startup.

While the purpose of the ETU will be to evaluate as many of the
components asnd systems as are ready at the time the unit is scheduled
to startup, it is anticipated that some expedients will be necessary to
meet the startup date. Such things as direct access maintenance will
be used initially with demonstration of the remote technigue developed
for radiocactive systems to be made at a later date. Where avallable,
representative reactor-grade instrumentation will be used, however, if
the actual instrument 1s not ready, some substitute measure will be
taken with provision made to install the reactor grade equipment later.

The operation of the ETU will be continuous during the testing
period with molten salts. During this time the plant will be staffed
largely by operators and engineers in training for assignment to the

MSBE. This will provide a well trained staff for the startup of the
breeder experiment. ' '

SCHEDULE AND COST

 

The schedule and costs are shown on Table 5 and were estimated
on the basis that the program begins with FY 1968 and that the ETU and
MSBE installations should be completed by the end of FY 1971 and
FY 1974, respectively. The main effort of the development program is
centered on building components of MSBE size, assembling them into
systems and testing them in the Engineering Test Unit.
Table

5.

Cost Summery - Component and System Development

 

 

 

($ thousands) FyY
68 69 70 71 70 73 7l 75 Totals
Core flow tests 100 150 150 150 100 50 700
Control rod tests 100 250 150 100 80 8¢ 80 40 B8C
Pumps 600 1400 600 350 280 200 150 150 3730
Heat transfer studies 100 L00 100 100 80 - - - 780
Steam system vendor tests 60 30 50 1ko
Selt samples 100 150 120 100 50 50 50 620
Cell furnace 120 100 100 100 50 50 50 50 620
Drain tank and storage 50 80 50 50 230
tank cooler
Sait valves 80 200 150 100 50 50 20 20 670
Gas systen 130 400 200 150 100 100 50 50 1180
ETU Shown separately in Summary CF 66-7-39
Totals 1280 3140 1680 1270 840 580 400 360 8550

 
o1

ACKNOWLEDGEMENTS

 

The work of assembling the program for engineering development of
components and systems for the Molten-Salt Breeder Reactor was done by
some of the same people who have had the job of developing equivalent
components for the MSKE and even for the original Aircraft Reactor
Experiment. This continuity is considered very important in the success
of a future program.

P. G. Smith and L. V. Wilson prepared the section on pumps.
R. J. Kedl and A. N. Smith prepared a substantial portion of the remainder.
R. B. Briggs provided guidance and editorial help in the preparation of
the document.
5e

REFERENCES

lp. R. Kasten, E. S. Bettis and R. C. Robertson, Design Studies of
1000-Mw{e) Molten-Salt Breeder Reactors, ORNL-3996 (August 1966).

 

 

2R. C. Robertson, MSRE Design and Operations Report; Part I, Descripition
of Reactor Design, ORNL-TM-T28 (January 1965).

 

3A. G. Grindell, W. F. Boudreau and H. W. Savage, "Development of
Centrifugal Pumps for Operation with Liquid Metals and Molten Salts at
1100-15C0°F," Nuclear Science and Engineering, Vol. T(1): 83-91
(January 1960).

 

“H. W. Savage and A. G. Grindell, "Pumps for High-Temperature Liquid
Systems," Space-Nuclear Conference, Sponsored Jointly by American
Rocket Society and ORNL, Gatlinburg, Tennessee, May 1961.

SORNL MSRP Semiann. Prog. Rept. July 31, 1964, ORNI-3708, pp 147-167
(November 1964).

 

bP. G. Smith and L. V. Wilson, "Development of an Elevated-Temperature
Centrifugal Pump for a Molten-Salt Nuclear Reactor,' ASME 65-WA/FE-27
(November 1965).

  

Breeder Reac tor», ORMNL-TM=19%3 (Septem"mf 1967).
of Reactor Design, ORNL-TM-T728, pp 162-172 (Januvary 1965).

°C. H. Gabbard et al., Heat Transfer Performance of the MSRE Main Heat
Exchanger and Radiator, ORNL-CF-67-3-38 (March 21, 1967).

 

 

100RNI, MSRP Semiann. Prog. Rept. Aug. 31, 1966, ORNL-403T (January 1967).

 

11y, W. Cooke, Experimental Determinations of the Thermal Conductivity
of Molten Fluoride Salt Mixtures, ORNL-CF-66-5-41 (May 12, 1966).

 

 

124, B. McCoy, Jr. and J. R. Weir, Jr., Material Development for Molten-
Salt Breeder Reactors, ORNL-TM-1854 (June 1967).

 

 

13H. 5. Swenson et al., "Heat Transfer to Supercritical Water in Smooth-
Bore Tubes," Trans. of the ASME - Jour. of Heat Transfer (November 1965).

 

44, A, Bishop et al., High Temperature Supercritical Pressure Water
Loop, Parts I through IV, WCAP-2056 (Westinghouse Electric
Corporation) Pts. 1-6 (February 1964).

 

 

15¢, ¢. Lawson et al., "Enhanced Heat Transfer Tubes for Horizontal
Condensers with Possible Applications in Nuclear Power Plant Design,”
Trans. - 1966 Winter Meeting of ANS, Pittsburg, Pennsylvania.

 
53

REFERENCES - cont'd

 

18F. N. Peebles, The Motion of Gas Bubbles in the HRE Reactor Core,
ORNL-1171 {(January 4, 1952).

 

175, A. Hafford, Development of the Pipe Line Gas Separator, ORNL-1602
(February 18, 1954L).

 

187. A. Hafford, HRT Gas Separator, ORNL-CF-54-6-112 (June 1k, 1954).

 

197, 4. Hafford, large Scale Gag Separators, ORNL-CF-55-7-67
(July 20, 1955).

 
20

 

Internal Distribution

MSRP Director's Office 95.
Room 325, 9204-1 96,
R. K. Adams g9T.
G. M. Adamson 98,
R. G. Affel 99.
L. G. Alexander 100.
R. F. Apple 101.
C. I'. Baes 102.
J. M. Baker 103.
S. J. Ball 10k,
W. P. Barthold 105.
H. F. Bauman 106.
S. E. Beall 107.
M. Bender 108.
E. S. Bettis 109.
F. F. Blankenship 110.
R. E. Blanco 111.
J. 0. Blomeke 112.
R. Blumberg 113.
E. G. Bohlmann 114,
C. J. Borkowski 115.
G. E. Boyd 116.
J. Braunstein 117.
M. A. Bredig 118.
R. B. Briggs 119.
H. R. Bronstein 120.
G. D. Brunton 121..
D. A. Canonico 122.
5. Cantor 123.
W. L. Carter 12L4,
G. I. Cathers 125.
J. M. Chandler 126.
E. L. Compere 127.
W. H. Coock 128.
L. T. Corbin 129.
J. L. Crowley 130.
F. L. Culler 131.
J. M. Dale 132,
D. G. Davis 133.
5. J. Ditto 134,
A. S. Dworkin 135.
J. R. Engel 136.
E. P. Epler 137.
D. E. Ferguson 138.
L. M. Ferris 139.
A. P. Fraas 140,

Hee sSSP0 eER s D HE

aEF@OmTQOQEEEDErrYQarunmSoRRag s iy dadaoaoddd = HYaQnm

AEHErEHoEEROEpHERO3E 0SB 49T

ORNI~-TM-1855

Friedman
Frye, Jr.
Gabbard.
Gallaher
Goeller
Grimes
Grindell
Guymon
Hannaford
. Harley

. Harman
Harrill

. Haubenreich
. Heddleson
. Herndon

. Hightower
. Hoffman

. Horton

. Hudson
nouye

. Jordan

. Kasten
Kedl

. Kelley
Kelly
Kennedy
Kerlin
Kerr
Kirslis

. Krakoviak
Krewson
Lamb

Lene

. Lindauver
. Litman

. Lundin

. Lyon

. MacPherson
. MacPherson
. Martin
Methews

. Matthews
. McClung

. McCoy
McDuffie
. McGlothlan

-
1h1.
1h2,
143,

1Lk,

1hs,
146.
147.
148,
149,
150.
151,
152,
153,
154,
155.

156.

157.
158.-162.
163.

164.
165.
166.

167.
168.
169.
170.

207 .

211.

227

231.

20T,

SWHEHPpNHNCOEHSG HBE O

C.
Dun
H.
W

J

M.
G.
A

.
G.

~-208.
209,
210.
~225.
226.

.—228.

229.
230.
-245,
2L6.
~2UL8,

TG = s o E

. McHargue
. McNeese

. Meyer
Moore
Nichols
Nicholson
Oakes
Patriarca

Q"B nmEYy

. M. Perry

B. Piper

. BE. Prince

L. Redford

. Richardson

C. Robertson
C. Roller

C. Savage

E. Schilling
lap Scott
Seagren
Schafler
Shaffer
Skinner

. Slaughter
. Smith
omith
Cmith

= = g

w == A

56

171.
172.
173.
174,
175.
176.
177.
178.
179.
180.
181.
182.
183.
184,
185.
186.
187.
188.
189.
190.
191.
192.-193.
19h,..195,
196.-205.
206.

ORNL~-TM--1855

I.. Smith
G. Smith
. Spencer
Spiliewak
C. Steffy
H. Stone
R. Tallackson
. Taylor
. E. Thonmsa

S. Watson
F. Weaver
H. Webster
M. Weinberg
R. Weir
J. Werner
W. West
E. Whatley
C. White
V. Wilson

TPy ERRESg oSy MO S5O

Central Research Library
Document Reference Section
Laboratory Records {(LRD)
Laboratory Records (ILRD--RC)

External Distribution

 

F. Cope, Atomic Energy Commission, RDT Site Office
Giambusso, Atomic Energy Commission, Washinglton

Jd. Larkin, Atomic Energy Commission, ORO

. W. Mcintosh, Atomic Energy Commission, Washington
. M. Roth, Atomic Energy Commission, ORO

. Shaw, Atomic Energy Commission, Washington

. L. Bmalley, Atomic Energy Commission, ORO

F. Sweek, Atomic Energy Commission, Washington

Division of Technical Information Extension (DTIE)
Research and Development Division, ORO
Reactor Division, ORO