ORNL-CF-57-10-41.txt

X-822

 

DATE:

SUBJECT:

TO:

FROM:

UNCLASSIFIED

OAK RIDGE NATIONAL LABORATORY
Operated By
UNION CARBIDE NUCLEAR COMPANY

ucc

POST OFFICE BOX P
OAK RIDGE, TENNESSEE

October 10, 1957
MOLTEN SALTS FOR CIVILIAN POWER

A. B. Kinzel
H. G. MacPherson

Distribution

 

S Vel
ORNL
STER C

 

 

 

ORNL
CENTRAL FILES NUMBER

57-10-k41

 

COPY NO. 9{‘ 7(—

External Transmittal

Authorized
Distribution Limited to

Recipients Indicated

1-20. A. B. Kinzel, Union Carbide and Carbon Corp.

21-35. H. G. MacPherson
3. C. E. Center
57. W. H. Jordan
38. A. M. Weinberg
39-41. ILaboratory Records
h2~hfi. C. R. Library
/4l.  ORNL-RC

NOTICE

This document contains information of a preliminary
nature and was prepared primariiy for internal use
at the Ock Ridge Nationai Laboratory. 1t is subject
to revision or correction and therefore does not
represent a final report.

UNGLASSIFIED

(077
PROPERTY OF
WASTE MANAGEMENT
DOCUMENT
LIBRARY

UCN-15212
& 2-84)

 
MOLTEN SALTS FOR CIVILIAN POWER

by

H. G. MacPherson

Oak'Ridge Netional ILaboratory
Oak Ridge, Tennessee

October 10, 1957

Prepared For Inclusion In Report To be Issued By

Technical Appraisal Task Force, Edison Electric Institute
1,

MOLTEN SALTS FOR CIVILIAN POWER

Molten salts provide the basis of a new family of liquid fuel power
reactors. The range of solubility of uranium and thorium compounds makes the
gsystem flexible, and allows.the consideration of a variety of reactorz. Suit-
able salt mixtures have melting points in the 850-9500F range and will probably
prove to be sufficlently compatible with known alloys and to provide long-1ived
components, if the temperature 1s kept below 1300°F. Thus the salt systems
naturally tend to operate in a temperature region suitable for modern stesm
plants and achieve these temperatures in unpressurized systems.

The molten salt system, for purposes other than electric power gen-~
eration, is not new. Intensive research and development over the past seven
years undexr ORNL sponsorship has provided reasonable answers to a majority of
the obvious difficulties. One of the most important of these is the ability
to handle liquids at high temperatures and to maintain them above their melt-
ing points. A great deal of information on the chemical and physical properties
of a wide variety of molten salts has been obtained, and methods are in operation
for their manufacture, purification and handling. It has been found that the
simple ionic salts are stable under radiation, and suffer no deterioration other
than the build-up of fission products.

The molten salt system has the usual benefits attributed to fluid fuel
systems. The principal advantages claimed over solid fuel elements are: (1) the
lack of radiation damage that can limit fuel burn-up; (2) the avoidance of the
expense of fabricating new fuel elements; (3) the continuous removal of gaseous

fission products; (4) a high negative temperature coefficient of reactivity; and
(5) the ability to add make-up fuel as needed, so that provision of excess
reactivity is unnecessary., The latter two factors make possible a reactor
without control rods, which automatically adjusts its pover in response to
changes of the electrical load. The lack of excess reactivity can lead to s
reactor that is safe from nuclear power excursions.

In comparison with the aqueous systems, the molten salt system has
three outstanding adventages: it allows high temperature with low pressure;
explosive radiclytic gases are not formed; and it provides soluble thorium
compounds. The compensating disadvantages are high melting point and poorer
neutron economy; the importance of these is difficult to assess without fur-
ther experience.

Probably the most outstanding characteristic of the molten salt systems
1s their chemical flexibility, i.e., the wide variety of molten salt solutions
which are of interest for reactor use. In this respect, the molten salt systems
are practically unique; this is the essential advantage which they enjoy over
the U-Bi systems. Thus the molten salt systems are not to be thought of in terms
of a single reactor - rather, they are the basis for a new clags of reactors.
Included in this class are all of the embodiments which comprise the whole of
solid fuel element technology: straight U255 burner, Th-U thermal converter or
breeder, Th-U fast converters or breeders. Of possible short-term interest is
the U255 straight burner: because of the inherently high temperatures and be-
cause there are no fuel elements, the fuel cost in the salt system can be of the
order of 2-3 mills/kwh,

The state of present technology suggests that homogeneous converters

using a base salt composed of BeF2 and either IiTF or NaF, and using UF& for
fuel and ThFh for a fertile material, are more suitable for early reactors
than are graphite moderated reactors or Pu fueled reactors. The conversion
ratio in such an early system might reach O.6. The chief virtues of this
class of molten salt reactor are that it {s based on well explored principles
and that the use of a simple fuel cycle should lead to low fuel cycle costs.

With further development, the same base salt (using IiTF) can be
combined with a graphite moderator in a heterogeneous arrangement to provide
a self-contained thorium-0235 system with & breeding ratio of about one. The
chief advantage of the molten salt system over other liquid systems in pursuing
this objective is, as has been mentioned, that it is the only system in which
a soluble thorium compound can be used, and thus the problem of slurry hand-
ling 1s avolded.

A reactor called the Alrcraft Reactor Experiment (ARE), using a
molten fluoride fuel, was operated in November 1954, The reactor had a moderator
consisting of beryllium oxide blocks. The fuel, which was a mixture of sodium
fluoride, zirconium fluoride and uranium fluoride, flowed through the moderator
in nickel alloy tubes and was pumped through an external heat exchanger by means
of & high temperature centrifugal pump. The reactor operated at a peak power of
2-1/2 megawatts, and was dismantled after carrying out a scheduled experimental
program. The ARE demonstrated again the extreme stability of a liquid fuel
reactor. The reactor power level responded automatically to changes in the rate
of heat removal from the heat exchanger, and control rods were used only for set-
ting the operating temperature level. It was demonstrated also that Xel55 came

off continuously, as did other gaseocus fission products,
i,

A recent conceptual design study has been made of a 240 electrical
megavatt central station molten salt reactor. The purpose of this study was
to examine the economics and feasibility of such a reactor using molten salts,
with an attempt to keep the technology required and the processing scheme as
simple as possible.

The reactor is a two region homogeneous reactor with a core approxi -
mately six feet in diameter and a blanket two feet thick. Moderation is
provided by the salt itself, so there i1s no need for moderator or other struc-
ture inside the reactor. The core, with its volume of 113 cubic feet, is
capable of generating 600 megawatts of heat at a pover density in the core of
187 watts/cc. The net electric power generation is approximately 240 megawatts.
The general arrangement of the core and blanket is shown in Figure (1).

The basic core salt is a mixture of about 60% 117F and 4o% BeF,.
Additions of thorium fluoride can be made if desired, and enough U255Fh is
added to make it critical. The blanket contains ThFh, elther as the eutectic
of IiF and ThFh, or mixtures of it with the basic core salt. The melting point
of the core is about 850°F, and that of the blanket salt is 1080°F or lower.

Both the core fuel and the blanket salt are circulated to external
heat exchangers, six in parallel for the core and two in parallel for the
blanket. The heat is transferred by intermediate fluids from these heat ex-
changers to the bollers, superheaters, and reheaters. The heat transfer system
is designed so that, with a fuel temperature of 1200°F, & steam temperature of

1000°F at 1800 psi can be achieved.
5.

Figure (2) gives a block diagram of the gross features of one of the
heat transfer systems of the core circuit. A factor in the selection of the
multiple stage heat transfer system involving two intermediate fluids was the
desire to have only compatible fluids in adjoining volumes where leaks could
involve radicactive materials. This avoids the possiblility of liberation of
fission products as the result of a chemical accident. As the system 1s designed,
the pressures developed by the pumps, by difference of density of the fluids, and
by over pressure, are such that any failure producing mixing of fluids would tend
to produce flow toward the reactor core, rather than from it, tending further to
confine the fission products.

An exothermic chemical reaction would result 1f the sodium and water
or steam were mixed. However, this would not involve radioactive materials and
would pose only the same danger problems as in any conventional plant handling
quantities of chemically active materials where no biological poisons are involved.

The general layout of the reactor plent is shown in Figure (3). The
primery heat exchangers and the reactor are included within the primary shield,
and the secondary heat exchangers, which are only moderately radiocactive, are
included in separate shielded compartments so they can be serviced individually.
Figure (4) shows & vertical section through the reactor and power plant. An
isolating vessel is inside the primary shield surrounding the reactor and the
primary heat exchangers. This vessel will contain any radioactive gases
liberated through leakage and will provide an inert atmosphere for remote
maintenance of this highly radioactive section.

The chemical processing method postulated uses the fluoride volatility
process‘under development at Oak Ridge. The molten salt 1s treated fiith fluorine.

UF% is recovered and is converted to UF& by a reduction process. The core salt
6.

can be fluorinated in small batches at the'rate of one or two batches per day.
The barren salt, stripped of uranium but containing most of the plutonium, fig-
sion products, and corrosion products, is transferred to tank storage where it
1s held for future salt recovery. The UF% reduction process will discharge its
UFh product directly to a fuel salt mixing pot, to which 1s also fed Presh base
salt and make-up UFh‘

Chemical processing of the blanket salt is physically much the same
as that of the core salt, except that after fluorination the blanket salt with
the Pa and fission products that it contains is returned to the blanket system.
Although research on methods of separeting the fission products from the molten
salts is under way, the possibility of doing this 1s not considered in calcu-
lating the fuel cycle costs. For purposes of calculating costs on present
technology, it is assumed that the core salt with its contained fission products
will be held in permanent storage and that new core salt must always be provided.

Assuming that the chemical plant 1s based on the capacity and the cost
of the current ORNL volatility pilot plant, a total fuel cycle cost of 2.5 mills/kwh
is calculated. Table I shows & breakdown of the fuel cycle costs. This fuel cycle
cost is equivalent to the sum of the items of chemical processing, fuel element
refabrication, inventory charge, and fuel costs for a conventional solid fuel
element reactor. The estimate is based on & $17/g cost of U235, an inventory
charge of 4%, and an 80% load factor. The 0233 produced would be burned in the
reactor. For this reason, no breeding credit appears as such, but only a reduc-

tion in the amount of U235 purchased for burn-up.
Table T

BREAKDOWN OF FUEL CYCIE COSTS

mills/kwh
Inventory charge for U235 and U253 0.3
Burn-up of U255 1.2
Replacement of core salt 0.5
Chemical plant capital charge 0.2
Chemical plant operation 0.3
2.5

 

To obtain the total power costs, the charge for operation and main-
tenance and the capital costs of the plant must be added to the fuel cycle costs.
No one knows what operation and maintenance will amount to for power reactors
since there is no operating experience, and by common consent this is placed at
1 mil1/kvh. |

The capital costs of a power reactor will in the end depend on such
broad-scale things as the physical size of the pladt, the weight of the com-
ponents, and the use of especlally expensive materials or complex methods of
construction. Molten salt reactors are certainly compact; the major hardware
is in the heat transfer systems. Since the fluids are good heat transfer agents,
these heat transfer systems are not bulky. The avoidance of a pressure vessel
and the elimination of control rods help in reducing the complexity of construc-
tion. There does not seem to be any reason why these reactors should have high
capital costs in comparison with other resctor plants.

A detailed cost estimate was attempted for the reactor plant described

above. The flow diagrams were broken down into individual components insofar as
8.

possible, and costs of purchasing, inspecting, and installing these components
were estimated on the basis of standarad engineering cost estimating procedure

as modified by experience in the nuclear power field. These modifications are
extensive, as & result of the higher standards required and the necessity for

multiple inspection of every piece that goes into a reactor.

Teble II gives a partial breakdown for the various factors in the
reactor plant cost. Conventional portions of the electrical generating plant
were not broken down, but were trested as one lump sum, since experience in
this field has indicated that such costs can be estipated to a falr degree of
accuracy. The total for the 2L0 megawatt plant of approximately $55,000,000
does not include the cost of the chemical processing plant since this has been
separately charged under the fuel cycle costs. A 4O-year 1ife was assumed for
the conventional portion of the plant and a 20-year life for the remainder.
These glve fixed costs of 14% per year on the #99/kw of the conventional plant
and 16% per year on the $133/kw of the reactor portion. Assuming a load factor
of 80% results in a fixed charge of 5 mills/kwh. The cost of power thus adds up

as follows:

mills/kwh
Fixed charges on plant 5.0
Operation and maintenance 1.0
Fuel and fuel processing cycle 2.5
8.5

The costs outlined above for the construction and operation of the
Reference Design Reactor are predicated on obtaining favorable results from a
development program which would prove-in the life of the components and provide

methods for carrying out remote maintenance.
93

At the present time, the greatest uncertainty in the molten salt
reactor power costs derives from this problem of remote maintenance. The
reactor, the primary heat exchangers, and the fuel pumps will be highly radio-
active, even after withdrawing the fuel, and maintenance operation on them
must be by remote control. Figures (3) and (4) indicate how these items could
be located in a central containment vessel. The kind of maintenance visualized
1s that of replacement of heat exchangers and pumps as. units.

The present thought is that the pump motor, bearings, shaft and impeller
could be replaced as & uniflby unbolting a flange and breaking a gasket seal located
in a gas space that 18 cool relative to the reactor temperature. The heat exchangers
would be removed by making pipe disconnections only at selected points designed for
this purpose. It is anticipated that to replace heat exchangers will require the
development of remote cutting and welding operations for speclally designed joints,
although it is possible that some sort of adequate freeze seal-type joint can be
devised. There is no scarcity of ideas as to methods for carrying out remote main-
tenance, but the detailed engineering and practical trials of such methods have
not yet been made, and thus any estimate of those capital and operating costs

related to remote maintenance is highly speculative.
Reactor Vessel and Primary System

Reactor Vessel

Fuel and blanket pumps with
motors (6 - 6000 gpm,

2 - L4000 gpm)

Eight primary heat exchangers
fuel-to-salt (21,240 sq ft
total surface)

Miscellaneous reactor and
primary circuit Piping,
tools, and other equipment

Intermediate Heat Transfer System

Eight intermediate heat
exchangers (24,000 sq ft

total surface area)
Intermediate system pumps

and motors (6 - 6000 gpm,

2 - 3000 gpm)
Miscellaneous intermediate salt
coolant circuit equipment

Sodium-Steam Generator System

Pumps

Heat exchangers, boilers,
superheaters and reheaters,
total 40,000 sq £t surface

Miscellaneous piping, vessels,
blenders, etec.

Miscellaneous Reactor Components

Reactor building, site and
improvements (reactor portion)

Reactor isolation container
Instrumentation

Remote maintenance equipment

Miscellaneous auxiliary
systems

Total for Reactor Construction
Engineering Design
Prime Contractor Fee

Table II

# 900,000
800,000

775,000

941,000

746,000
640,000

625,000

1,330,000
1,159,000

1,167,000

 

2,500,000

450,000
750,000
800,000
1,470,000

Spare Parts (pumps, heat exchangers, etc.)
Original Inventories of Salts and Sodium

Start-up_Operation
Contingency Reserve

Conventional Electrical Generating Plant

10.

g 3,416,000
2,011,000
3,656,000
5,970,000

$ 15,053,000

2,500,000

3,800,000

850,000

4,455,000

1,000,000

4,160,000

23,750,000

TOTAL g 55,568,000

 
11

UNCLASSIFIED
ORNL -LR-DWG 19330

  

r——————
PUMP QUTLET

PLAN

 

 

SECTION

Figure 1 Reference Design Reoctor
12

 

 

 

 

‘WaIskg 49jsund) J03H Jo woaboig dnowsyds g 3anSt g

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

938/ 4 ‘MO =

do “3¥NLYNIINIL =

  

 

 

 

 

 

 

 

 

 

 

 

28}
4616
3108
y 6
D9¢ 5602 40.9
]
. —¥3aN3I8
ove
402
9 9¢
3008
LY3IHIY WOYH
4098
3069
4016
- Joozi _
YIONYHIXI LvIH - YIONIY
D o
_ N Ol ©oN 40601
A b
, » 452il
I 219°2
| —¥31Y3HY3dNS To55
) X ¥IONVHIXS
29 vw ovS N L3N HIONVHOX 3
quuflmmw 01 WNIQ0S OL LY 3IH
MYNIT MUNIT4 OL
g Q1 17vS 13N 4
T 4000!
08'9¢
96621
20011

0686} 9MA ~YT-INYO
Q3131SSYTIINN

 

 

 

 

 

 

 

 

 

an3o3an

JOLov 3y

  

 

 

 
13

ol Jamod J0 MAA Uold @ FINST

 

 

 

 

 

          

 

N

HALVIHIY i

      

 

 

 

 

 

.~ 3NIBENL 3YNSSIdd HOH

 

 

  

HIDNYHIXZ LV3IH

HOIVHENID
AUVWINY - -

   

2

HOLDY3Y
QL

_‘,

 

     

 

NOIOIY S5320V

 

 

 

 

 

 

 

 

 

 

 

 

        
   

 

 

HRAVIHYIANS - Y
HOLYHINID
T ' I
AV3IHIY OL ding A
NYHIX H 2
m_mw X3 1v3 — U u\\\
TY3M 3VAYHIS YISNTONOD
.
mmozm,“m mmem HILYM
K @334
_ |
HMIONYHOXF Lv3H T ;.Fllif.ll.\. 71
AYVANQDES -

 

 

 

 

 

 

JEEGHOMG - 8 -INYO
TIASSYIINN
14

 
 

"JUDld JBMO4 PUD JOJOD3Y ybnosy| uoigo9g ¥ sandi Jq

1334 Nt 3OS
o€ 02 oF O O

LVIHIY HOd °
¥IONVHOX3 LVIH -
~ ON O ON

 

     

» - -

 

 

and
INFVAON3Y

 

                 

dANd
LIN2HID -

ON NIYN-—~ : — ] e T HOLYY3INTO

INgsnL
IUNSSIY HOIH - YOLYE3N3O

          

             

*,

“INIGNNAL 3YNSSINA MOT

       

30vds
YIINVHIXZ Lv3H
AUYANOO3S

3NIBUNL 3UNSSINd MOTT

     

dWNd LINDHID
837108 ON

 

ELE6! OMQ-HT-TINBO
CAASSVIONN