NAT_MSBRdesign.txt

THE DESIGN AND PERFORMANCE

 

FEATURES OF A SINGLE-FLUID

MOLTEN-SALT BREEDER REACTOR

E. S. BETTIS and ROY C. ROBERTSON Reactor Division

KEYWORDS: molten-salt re-
actors, design, performance,
economics, uranium-233, pow-
er reactors, fuels, MSBR, cost,
fuel cycle

Oak Ridge National Labovatory, Oak Ridge, Tennessee 37830

Received August 4, 1969
Revised October 2, 1969

 

A conceptual design has been made of a single-
fluid 1000 MW(e) Molten-Salt Breeder Reactor
(MSBR) power station based on the capabilities of
present technology. The reactor vessel is ~22 ft
in diameter X 20 ft high and is fabricated of
Hastelloy-N with graphite as the wmoderator and
reflector. The fuel is 23U carvied in a LiF-BeF ,-
ThFy mixture which is molten above 930°F. Tho-
rium is converted to “®U in excess of fissile
burnup so that bred material is a plant product.
The estimated fuel yield is 3.3% per year.

The estimated construction cost of the station
is comparable to PWR total construction costs.
The power production cost, including fuel-cycle
and graphite replacement costs, with private
utility financing, is estimated to be 0.5 to 1 mill/
kWh less than that for bresent-day light-water
reactors, largely due to the low fuel-cycle cost
and high plant thevmal efficiency.

After engineering development of the fuel puri-
fication processes and large-scale components, a
practical plant similar to the one described here
appears to be feasible.

 

INTRODUCTION

The objective of this design study is to investi-
gate the feasibility of attaining low-cost electric
power, a low specific inventory of fissile material,
and a reasonably high breeding gain in a molten-
salt reactor. As discussed by Perry and Bauman®
a molten-salt reactor can be designed as either a
breeder, or, with relatively few major differences
other than in fuel processing, as a converter. This
particular design study is confined to the single-
fluid Molten-Salt Breeder Reactor (MSBR).

190

NUCLEAR APPLICATIONS & TECHNOLOGY

The conceptual design of the MSBR plant was
made on the basis that the technology required for
fabrication, installation, and maintenance be gen-
erally within present-day capabilities. Major de-
sign considerations were keeping the fuel-salt
Inventory low, accomodating graphite dimensional
changes, selection of conditions favoring graphite
life, and providing for maintainability.

The facilities at an MSBR power station can be
grouped into the following broad categories: (a)
the reactor system which generates fission heat in
a fuel salt circulated through primary heat ex-
changers; (b) an off-gas system for purging the
fuel salt of fission-product gases and colloidal
noble metal particulates; (c) a chemiecal proces-
sing facility for continuously removing fission
products from the fuel salt, recovery of the bred
*3U, and replenishment of fertile material; (d) a
storage tank for the fuel salt which has an after-
heat removal system of assured realiability; (e) a
coolant-salt circulating system, steam generators,
and a turbine-generator plant for converting the
thermal energy into electric power ; and (f) gen-
eral facilities at the site which include condensing
water works, electrical switchyard, stacks, con-
ventional buildings, and services. These cate-
gories are not always clearly defined and are
closely interdependent, but it is convenient to
discuss them separately. The reactor, and its
related structures and maintenance system, the
drain tank, the off-gas system, and the chemical
processing equipment, are of primary interest.
The steam turbine plant and the general facilities
are more or less conventional and will be dis-
cussed only to the extent necessary to complete
the overall picture as to feasibility and costs of an
MSBR station.

In a single-fluid MSBR the nuclear fuel is 2*°U
(or other fissile material) carried in a lithium-T7-

fluoride, beryllium-fluoride, thorium-fluoride salt.

The mixture is fluid above ~930°F and has good
flow and heat transfer properties and very low

VOL. 8 FEBRUARY 1970
Bettis and Robertson

vapor pressure. This salt is pumped in a closed
loop through a graphite-moderated and reflected
core where it is heated to ~ 1300°F by fissioning
of the fuel. It then flows through heat exchangers
where the heat is transferred to a circulating heat
transport salt. This fluid, in turn, supplies heat to
steam generators and reheaters to power a con-
ventional high-temperature, high-pressure steam
turbine-generator plant. The thorium in the flu-
oride fuel-salt mixture is converted to **Uat a
rate in excess of the fuel burnup so that fissile
material, as well as power, is a valuable product
of the plant.

A low specific inventory is obtained by design-
ing the MSBR to operate at a high reactor power
density with a minimum of fuel salt in the circu-
lating system. The concentration of fissile-fertile
material in the salt results from a compromise
between keeping the inventory low and achieving a
higher breeding gain. High performance depends
on keeping the neutron parasitic absorptions low
and fuel losses to a minimum. The Li and Be in
the fuel salt are good neutron moderators, but
their concentrations are relatively low in fluoride
salts and additional moderation is needed. Graph-
ite is the most satisfactory moderator and re-
flector for the MSBR. However, radiation affects
the graphite and its useful life varies nearly in-
versely as the maximum power density in the
core. The radiation damage effect is also in-
creased by higher temperatures. Selection of a
power density is thus a balance between a low fuel
inventory and the frequency with which the graph-
ite must be replaced.

The optimization studies which equate the sev-
eral factors mentioned above are described by
Scott and Eatherly.?* These studies indicate that
an average core power density of 22.2 kW/liter
results in a useful graphite life of ~ 4 years, which
is the operating condition used in this design
study. Thus, the reactor design must provide for
periodic replacement of the core graphite with
minimal plant downtime and complexity of mainte-
nance equipment.

To attain a high nuclear performance it is
necessary to maintain low concentrations of Pa
and %°Xe in the high flux region of the core.
The protactinium is kept low by processing a
small side stream of the salt for removal of this
nuclide and other fission products. By integral
processing on-site a minimum inventory of fuel
salt is involved in transport and storage. The
xenon is removed by helium-sparging the fuel salt
on a few-second cycle; the core graphite is also
sealed to decrease the rate of diffusion of the Xe
into the pores of the graphite. The plant design
therefore includes the auxiliary systems to re-
move the nuclear poisons and the bred fissile ma-

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

MSBR DESIGN AND PERFORMANCE

terial from the salt, to purge the fission-product
gases, and to store or dispose of the radioactive
waste products; Whatley et al.,’ and Perry and
Bauman'® treat the subject of salt processing and
estimate fuel-cycle costs.

Although the economic performance of a 2000
MW (e), or larger, MSBR station would be signifi-
cantly better than that of smaller sizes, a plant
with a net electrical output of 1000 MW(e) was
chosen for this study because it permitted more
direct comparison with the results of other
studies. A steam-power cycle with 1000°F and
3500-psia turbine throttle conditions, with single
reheat, was selected because this was representa-
tive of current practice and because it afforded a
high overall plant thermal efficiency of ~ 44%.
The MSBR is adaptable to other steam conditions
and to additional reheats if future developments
lead in this direction.

All portions of the systems in contact with
salts are fabricated of Hastelloy-N. When this
material is modified with ~1% titanium or hafnium
to improve resistance to radiation embrittlement,
as described by McCoy et al.,* it has good high-
temperature strength and excellent resistance to
corrosion. No exothermic reactions of concern
result from mixing of the fuel and coolant salts
with each other or with air or water. It is im-
portant to keep water and oxygen out of the salts
in normal operation, however. The reactor graph-
ite is a specially developed type having very low
salt and gas permeability and good resistance to
radiation damage. The salt composition is a com-
promise between the nuclear and physical proper-
ties, chemical stability, etc., as discussed by
Grimes.? Some selected physical properties of the
materials important to the design study are shown
in Tables I, II, and III.

The performance features of the plant, in terms
of breeding gain, graphite life, thermal efficiency,
and the net cost to produce electric power, are all
reported here on the basis of normal full-load
operation at 80% plant factor. Very preliminary
review of the various modes of start-up and shut-
down, partial-load operation, etc., has not dis-
closed any major problem areas, however.

PLANT DESCRIPTION

A simplified flow diagram of the primary and
secondary-salt circulating systems is shown in
Fig.1. The fluoride fuel-salt mixture is circulated
through the reactor core by four pumps operating
in parallel.? Each pump has a capacity of ~ 14 000

 

a Flow sheet does not show duplicate equipment.

FEBRUARY 1970 191
Bettis and Robertson

TABLE 1

Properties of the Primary and Secondary Salts Used in
Conceptual Design Study of MSBR 1000- MW(e) Station

 

 

Primary Secondary
Salt Salt
Components LiF- BeF,- ThF,-UF, NaBF,-NaF
Composition, mole% 71.6-16-12-0.4 92-8
Molecular weight,
approximate 64 104

Liquidus temperature,
°F 930 725

Density, 1b/ft’ 205 (at 1300°F) 117 (at 988°F)
Viscosity, 1b/(ft h) 16.4 (at 1300°F) 2.5 (at 900°F)

Thermal conductivity,

 

 

 

 

Btu/(h ft °F) 0.7 to 0.8 0.27
Heat capacity,

Btu/(1b °F) 0.32 0.36
Vapor pressure at

1150°F, Torr (mm Hg) <0.1 252

 

gal/min and circulates the salt through one of four
primary heat exchangers and returns it to a com-
mon plenum at the bottom of the reactor vessel.
Use of four pumps and heat exchangers corres-
ponds to a pump size which represents a reason-
able extrapolation of the Molten-Salt Reactor
Experiment (MSRE) experience.

Each of the four coolant-salt circuits has a
pump of 22 000-gal/min capacity which circulates
the coolant salt through a primary heat exchanger
located in the reactor cell and then through steam
generating equipment installed in an adjacent cell.
The reactor can continue to operate although at
reduced output, if not all of the coolant salt pumps
are operative.

A plan of the reactor plant is shown in Fig. 2;
‘an isometric view and a sectional elevation are
presented in Figs. 3 and 4.

The reactor cell is ~62 ft in diameter and 35 ft
deep. It houses the reactor vessel, the four pri-
mary heat exchangers, and four fuel-salt circu-
lating pumps. The roof of the cell has removable
plugs over all equipment which might require
maintenance. The reactor cell is normally kept at
a temperature of ~ 1000°F by electric heater
thimbles to ensure that the salts will remain
above their liquidus temperatures. The estimated
maximum heating load is ~2500 kW. This
““furnace’’ concept for heating is preferred over
trace heating of lines and equipment because it
ensures more even heating, heater elements can
be replaced without reactor shutdown, there is no
need for space coolers inside the cell, and bulky
thermal insulation that would crowd the cell and
require removal for maintenance and inspection is
eliminated.

192 NUCLEAR APPLICATIONS & TECHNOLOGY

MSBR DESIGN AND PERFORMANCE

TABLE II
Nominal Values for Properties of MSBR Graphite

 

Density, 1b/ft> at room temperature 115
Bending strength, psi 4000-6000
Young’s modulus of elasticity, psi 1.7 x 10°
Poisson’s ratio 0.27
Thermal expansion, per °F 2.3 X 10°°
Thermal conductivity, Btu/(hr ft °F)

(1340°F) 35-42
Electrical resistivity, 2-cm 8.9-9.9 x 107*
Specific heat, Btu/(1b °F) at 600°F 0.33

Btu/(1b °F) at 1200°F 0.42

 

 

 

 

 

Biological shielding for the reactor cell is pro-
vided by reinforced concrete. The walls of the
reactor cell also furnish double containment for
the fuel-salt systems. Two thicknesses of carbon
steel plate form inner and outer containment ves-
sels, each designed for 50 psig, and also provide
gamma shielding for the concrete. Nitrogen gas
flows between the plates in a closed circulating
loop to remove ~3 MW(th) of heat. Double bellows
are used at all cell penetrations. The normal cell
operating pressure is ~ 26 in. Hg abs.

High-temperature thermal insulation is at-
tached to the inside of the containment membrane
to limit heat losses from the reactor cell. The
inside surface of the insulation is covered with a
thin stainless-steel liner to protect the insulation
from damage, to act as a radiant heat reflector,

 

 

 

 

TABLE III
Selected Physical Properties of Hastelloy-N
Nominal Chemical Composition of
Modified Alloy for Use in MSBR® wt%
Nickel 75
Molybdenum 12
Chromium 7
Iron 4
Titanium 1
Other 1
At 80°F At 1300°F
Density, 1b/ft* ~553 ~553
Thermal conductivity, Btu/(h ft °F) 6.0 12.6
Specific heat, Btu/(1b °F) 0.098 0.136
Thermal expansion, per °F 5.7 X 107%| 9.5 x 107°
Modulus of elasticity, psi 31 x10° | 25 x 10°
Electrical resistance, Q-cm 120.5%107° 126.0x107¢
Approximate tensile strength, psi 115 000 75 000
Maximum allowable design stress, psi 25 000 3 500
Maximum allowable design stress,
bolts, psi 10 000 3 500
Melting temperature, °F ~2 500 ~2 500

 

 

 

 

 

“The exact composition may be different from the nominal values
given here, as discussed by McCoy et al.*

VOL. 8 FEBRUARY 1970
MSBR DESIGN AND PERFORMANCE

Bettis and Robertson

-uonyels xomod YIS (9)MIN 000T I0F YASIN Jo weaderp mo[d T 814

$uo|d (MW 0001
._On_ _OA_O._. m._< mw._cm_ >>o_u_

A|quassy 494084

3up] uibi(g T o

 

3un} uipig

 

1 1S 9045

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

          

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1
P L.
SA|DA 929344~ suoj uioiQ o Fotes
194N A_cc_ooU|>>m“_u 3|05 Asowrid] m m " n_:com_U\..-Hu._ .
$00) (BUIJDBH) JUD|OOD) = asey Soic 1oolay 40d Lo jupjoo3 1
12045 199728 19°H $$95014 Loy
J§ - : !
NENENEAR2S VIl 7,," 1 4 JUaA 10 3uo] dnpjoH
3uo] uibiqg E I L ~ 21nssald SO m_oml.*..—o
3|05 A10puodag W - mvoraonoos
u
\ Jabuoyox3 josy .
JUBA 10O . s06-40 |
2.nssalg soo) / wo34S
swflm dung pinjJ
1,069 _4o Wod4s 0ayay 15414 oo juojo0d)
4/ o0l X G | > -—< > _
3,058 () — &
‘ juojoor) dwng [4,0601 40 Y/ql QL X 86 uisog
10}DIBUDS) o0 MM Y91 00¥9 N y240?
Woa4g ° 1aBubyox] joaH
Asowiid
SETLENEN 10101 w
3,061L1 40 Craane r
| fix /49l 01 %919 4,061 40 |
— 1058 War oixezis|| ? _ N [9559A
! b - 04009y
H | 1 < —'1 104 N_
S Wosic 10 . L 3,00€1 40 4/ 0l X 8% 6
3,00014° 410945 10943y , | 4,06L1 o ” Wl_o:cou [8AS] —
VAOLT S WAL DL ey 19pIMQ 3ur7 69
4, 00/ 10 1240p Po34 Y/q| Ol X 1 v MO| 4 x\maa:mm s00) .um_mrxm.
||||||||||||||| - .UC@ ._OEM ’mnwuo‘_&.
"3,0001 90 wosss y/qp 01 X | oA 18D 404003y ! weys!
dwnyg 4|08 %
A1owiuag
d 4/91 000§
UOIJD|1JUSA bfimmoflwm — doi} puy wiojy
JUSWIUID4UOD) | \:._On_o> 1940\
9OUDUBJUIDW ——
— uinyiay aH
204G
s0b-430 Mzfi |oACwaY
wnijid g
MO Sup | pup ‘_.vh_. 1
- pag |00210YD) ab1ng G8
s194]14 XNy MOJ3-iH el 13 pag |0o24oyD)
puy doi] auipoj MO O u”.m“w dnpjoH
|P°O °X Y Ly
W : dnpjoy 8
] AoQ-06
y0a14-314
ouy $164)14 MO ¥ 113
JUs A >uoj utoig 1ojojnWN oY
" $
abing dwnyg A1opuosag MIIMD

abing ssad01d "wayd

%

A|ddng ol

193

FEBRUARY 1970

VOL. 8

NUCLEAR APPLICATIONS & TECHNOLOGY
Bettis and Robertson

 

   

Fig. 2.

and to provide a clean, smooth surface for the
interior.

The stainless-steel ‘‘catch pan’’ at the bottom
of the reactor cell, shown in Fig. 4, slopes to a
drain line leading to the fuel-salt drain tank lo-
cated in an adjacent cell. In the very unlikely
event of a major salt spill, the salt would flow to
the tank. A valve is provided in the drain line to
isolate the tank contents from the cell during nor-
mal conditions and to permit pressurizing the
drain tank for salt transfer.

The four rectangular steam-generating cells
are located adjacent to the reactor cell. These
house the secondary-salt circulating pumps, the
steam generator-superheaters, and the reheaters.
The cell construction is similar to that of the re-
actor cell but only a single containment barrier is

194 NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

MSBR DESIGN AND PERFORMANCE

  

Plan view of MSBR cell complex for 1000 MW(e) power station.

required and heavy steel plate is not needed for
shielding the concrete. The 40-psig design pres-
sure would accommodate leakage of steam into the
cell.

Other adjacent cells provide for storage of the
fuel salt, for storage of spent reactor core as-
sembly, and for other radioactive equipment re-
moved for maintenance.

Through careful quality control of materials
and workmanship, the reactor loops containing
fission products would have a high degree of
integrity and reliability in preventing escape of
radioactive materials from the system. The cool-
ant-salt system will operate at a slightly higher
pressure than the fuel salt so that any leakage at
heat exchanger joints would be in the direction of
the fuel salt. The coolant-salt system would be

FEBRUARY 1970
Bettis and Robertson

Fig. 3.

Cutaway perspective of MSBR reactor and steam cells.

MSBR DESIGN AND PERFORMANCE

 

(1) Reactor, (2) Primary heat exchangers,

(3) Fuel-salt pumps, (4) Coolant-salt pumps, (5) Steam generators, (6) Steam reheaters, (7) Fuel-salt
drain tank, (8) Containment structure, (9) Confinement building.

provided with rupture disks to prevent steam
pressure from being transmitted to the fuel salt
via the coolant-salt system. In addition to the
double containment around all fuel-salt equipment,
mentioned previously, the building covering the
reactor plant is in itself a sealed structure to act
as a confinement for airborne material.

REACTOR

The MSBR reactor core design is based on a
maximum allowable neutron dose to the core
graphite of ~3 X 10°®> n/cm? (for E > 50 keV). The
average core power density is ~ 22 kW/liter,

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

affording a useful core graphite life of ~4 years,
a fuel yield of ~3.3% per year, and a compounded
fuel doubling time of ~ 21 years. These and other
data are given in Table IV. The dimensions of the
reactor were obtained through nuclear physics op-
timization studies discussed by Perry and Bau-
man."

The coefficient of thermal expansion of Has-
telloy-N is about three times that of the graphite.
As a result, the clearances inside the reactor
vessel increase significantly as the system is
raised from room temperature to operating tem-
perature. It is necessary to keep the graphite in a
compact array, however, to have control over the

FEBRUARY 1970 195
Bettis and Robertson

MSBR DESIGN AND PERFORMANCE

ntrol Rod Drive

 

 

 

Fig. 4.

TABLE IV

Principle Reactor Design Data for MSBR

 

Useful heat generation, MW(th)

Gross electrical generation, MW(e)

Net electrical output of plant, MW(e)

Overall plant thermal efficiency, %

Reactor vessel i.d., ft

Vessel height at centerline, ft

Vessel wall thickness, in.

Vessel head thickness, in.

Vessel design pressure, psig

Vessel top flange opening diameter, ft

Core height, ft

Distance across flats of octagonal core, ft

Radial blanket thickness, ft

Graphite reflector thickness, ft

Number of core elements

Size of core elements, in.

Salt-to-graphite ratio in core, % of core volume

Salt-to-graphite ratio in undermoderated region, %

Salt in reflector volume, %

Total weight of graphite in reactor, 1b

Weight of removable core assembly, 1b

Maximum flow velocity in core, ft/sec

Maximum graphite damage flux (>50 keV),
n/(cm? sec)

Graphite temperature in maximum flux region, °F

Average core power density, W/cm?

Maximum thermal neutron flux, n/ (cm2 sec)

Estimated graphite life, yearsa

Pressure drop through reactor due to flow, psi

Total salt volume in primary system, ft®

Thorium inventory, kg

Fissile fuel inventory of reactor system and
processing plant, kg

Breeding ratio

Yield, %/year

Doubling time, years

 

2250
1035
1000

3.2x 10"
1284

22.2

7.9 x 10%
4
18
1720
68 000

1 468
1.06
3.3
21

 

 

 

“Based on 80% plant factor

196

NUCLEAR APPLICATIONS & TECHNOLOGY

 

Sectional elevation of MSBR cell complex for 1000 MW (e) power station.

fuel-to-graphite ratios, to ensure that the salt
velocities in the passages are as planned and to
prevent vibration of the graphite. The fact that the
graphite has considerable buoyancy in the fuel
salt, and yet must be supported when the reactor
is empty of salt, is another important design con-
sideration.

Figures 5 and 6 show plan and elevation views
of the reactor. The vessel is ~ 22 ft in diameter
and 20 ft high, with 2-in.-thick Hastelloy-N walls.
The dished heads at the top and bottom are 3 in.
thick. The maximum design pressure is 75 psig.
A cylindrical extension of the vessel above the top
head permits the reactor support flange and top
head closure to be located in a relatively low-
temperature, low-activity zone between two layers
of roof shielding plugs. In this position the seal
welds and flange bolting are more accessible and
distortion of the flange due to high temperature is
kept small. Double metal gaskets and a leak-
detection system are used in the flanged closure.

The fuel salt enters the bottom of the reactor
at ~1050°F, flows upward through the vessel, and
leaves at the top at ~1300°F. The central core is
~14 ft in diameter and consists of extruded graph-
ite elements, 4 in. X 4 in. X 13 ft long. A 3-in.-
diam hole through the center of each element, and
ridges on each side of the elements to separate
the pieces, furnish flow passages and provide the
requisite 13% salt volume in this most active por-
tion of the reactor. Other shapes of graphite

VOL. 8 FEBRUARY 1970
Bettis and Robertson

MSBR DESIGN AND PERFORMANCE

 

Fig. 5. Top cut-away view of reactor vessel for MSBR 1000 MW(e) station.

elements can be considered, with perhaps the ex-

ception of cylinders, which would have poorly

defined flow passages between them. The graphite
prisms are maintained in a compact array, as
shown in Figs. 5 and 6.

Molded extensions on each end of the elements
provide a greater salt-to-graphite volume ratio at
the top and bottom of the main core to form an
undermoderated region which helps reduce the
axial neutron leakage from the core. The molded
extensions on the elements also provide the ori-

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

ficing to obtain the desired flow distribution. By
varying the salt velocity from ~3 ft/sec at the
center to ~ 2 ft/sec near the periphery, a uniform
temperature rise across the core of ~ 250°F is
obtained. The overall pressure drop in the salt
flowing through the core is ~ 26 psi.

Graphite slabs, ~ 2 X 10 in. in cross section
and 13 ft long, are arranged radially around the
active core. These slabs have larger flow pas-
sages, with ~ 37% of the volume being fuel salt.
This undermoderated region serves to reduce the

FEBRUARY 1970 197

 
Bettis and Robertson

MSBR DESIGN AND PERFORMANCE

 

Fig. 6.

radial neutron leakage. The slabs provide the
stiffness to hold the inner core graphite elements
in a compact array as dimensional changes occur
in the graphite. The slabs are confined by graph-
ite rings at the top and bottom.

Eight equally spaced vertical holes, ~3 in. in
diameter, are provided in the undermoderated re-

198 NUCLEAR APPLICATIONS & TECHNOLOGY

Sectional elevation of reactor vessel for MSBR station.

tion, mentioned above, to allow a small portion of
the incoming fuel salt to bypass the core region
and flow through graphite passages in the top head
reflector graphite to prevent overheating. These
holes are indicated on Fig. 5 as ‘‘lifting rod
ports.”’’

Since it is not expected that the core graphite

VOL. 8 FEBRUARY 1970
Bettis and Robertson

would last the life of the plant, it is designed for
periodic replacement. It was decided to replace
all the core graphite as an assembly rather than
by individual pieces because it appeared that this
method could be performed more quickly and with
less likelihood of escape of activity. Handling the
core as an assembly also permits the replacement
core to be carefully preassembled and tested
under essentially shop conditions.

The portions of the core described above are
arranged as a removable unit having an overall
diameter of ~ 16 ft and a total weight of 240 tons.
When the reactor is empty of salt, the graphite
rests on a Hastelloy-N bottom plate, which, in
turn, rests on the bottom head of the vessel. (The
bottom reflector graphite is keyed to this support
plate and when the vessel is filled with salt the
mass of the plate is sufficient to prevent the
bottom reflector from floating.) When it is neces-
sary to replace the graphite assembly, eight
Hastelloy-N rods, ~ 23 in. in diameter, are in-
serted through ports around the periphery of the
top head, down through the lifting rods ports in the
graphite, mentioned above, and latched to the
bottom support plate. The rods are then attached
to the top head so that the entire core can be re-
moved as an assembly. The reactor control rods
and the associated drive assemblies are also
removed with the head. The maintenance equip-
ment and procedures are briefly described in a
later section of this paper.

The reactor vessel also contains a 2.5-ft-thick
layer of radial reflector graphite at the outside
wall. This graphite receives a relatively low neu-
tron dose and does not require periodic replace-
ment. The graphite is arranged in wedged-shaped
slabs, ~ 12 in. wide at the thicker end and ~4 ft
high. These slabs are spaced ~3 in. from the
vessel wall to allow an upward flow of fuel salt to
cool the metal wall and maintain it below the
design temperature of ~ 1300°F. The slabs have a
radial clearance of ~1% in. on the inside to accom-
modate dimensional changes of the graphite and to
allow clearances in removing and replacing the
core assembly. Salt flow passages and appropri-
ate orificing are provided between the graphite
slabs to maintain the temperature at acceptable
levels. A system of Hastelloy-N bands and verti-
cal rods key the slabs together and cause the
reflector to move with the vessel wall as the
systems expand with temperature. Reflector
graphite is also provided in the top and bottom
heads. The amount of fuel salt in the radial and
axial reflector regions is ~ 1% of the reflector
volume. |

Design concepts for a single-fluid breeder re-
actor different from that described above are
clearly possible. Relatively little optimization

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

MSBR DESIGN AND PERFORMANCE

work has been completed to date. Hydrodynamic
testing will be an essential part of establishing a
final MSBR reactor design.

PRIMARY HEAT EXCHANGERS

Four counterflow, vertical, shell-and-tube heat
exchangers are used to transfer heat from the fuel
salt to the sodium fluoroborate coolant salt. Per-
tinent data for the exchangers are given in Ta-
ble V.

As shown in Fig. 7, fuel salt enters the top of
each unit and exits at the bottom after once-
through flow through the tubes. The coolant salt
enters the shell near the top, flows to the bottom
in a 20-in.-diam downcomer, turns and flows up-
ward through disk-and-doughnut baiffling in the
lower portion of the shell, and exits through 28-
in.-diam concentric piping at the top.

The Hastelloy-N tubes are L-shaped and are
welded into a horizontal tube sheet at the bottom
and into a vertical tube sheet at the top. The
toroidal-shaped header is stronger than a flat

TABLE V
Primary Heat Exchanger Design Data

 

 

Each of
4 Units
Thermal duty, Btu/h 1 922 x 10°
Tube-side conditions:
Fluid Fuel salt
Entrance temperature, °F 1 300
Exit temperature, °F 1050
Mass flow rate, 1b/h 23.7 x 10°
Volume salt in tubes, £t 64
Number tubes 5 900
Nominal tube o.d., in. 2
Tube thickness, in. 0.035
Tube spacing on pitch circle, in®? 2
Radial distance between pitch circles, in? | 0.6
Length, ft 21.6
Effective surface, £t? 11 000
Pressure drop, psi 129
Velocity in tubes, average ft/sec 10

Shell-side conditions:

Fluid
Entrance temperature, °F

Exit temperature, °F 1150
Mass flow rate, lb/h 17.8 x 10°
Inside diameter of shell, ft 5.4

Shell thickness, in. 2.5
Baffle spacing, ft 1.1
Pressure drop, psi 74
Velocity, typical, ft/sec 7.5

Coolant salt
850

Overall heat transfer coefficient, based on
log mean Aty tube o.d., Btu/ (h ft* °F) | 950

 

 

 

 

aTubes are not on triangular pitch to facilitate assembly of
bent tubes into tube sheets.

FEBRUARY 1970 199

 
Bettis and Robertson

Primary
Salt ¥

Fig. 7.

header and also simplifies the arrangement for
the coolant-salt flow. The design also permits the
seal weld to be located outside the heat exchanger.
About 6 ft of the upper portion of the tubing is bent
into a sine-wave configuration to absorb differen-
tial expansion. Maximum stresses in the tubing
were estimated to be <8000 psi.

200 : NUCLEAR APPLICATIONS & TECHNOLOGY

 

MSBR DESIGN AND PERFORMANCE

Sectional elevation of primary heat exchanger for MSBR station.

The tubes have a helical indentation knurled
into the surface to enhance the film heat transfer
coefficients. In the cross-flow regions of the
exchanger, a heat transfer enhancement factor of
2 was used on the inside of the tubes, and a factor
of 1.3 was applied to the outside. No enhancement
was assumed in the bent-tube region. The shells

VOL. 8 FEBRUARY 1970
Bettis and Robertson

of the exchangers are also fabricated of Hastel-
loy-N. Disk and doughnut baffles are used in the
shell to a height of ~11 ft. Coolant-salt velocity
in the unit varies but has an average value of 7 to
8 ft/sec.

The heat exchanger design was computer-
optimized and the calculations take into account
the temperature dependence of the salt properties
as a function of position in the exchanger. The
average overall heat transfer coefficient, based on
simple overall log mean temperature difference
and total area, is ~950 Btu/(h ft? °F), as given in
Table V. Limited heat transfer tests are now in
progress, but a more comprehensive program is
required to establish a final MSBR primary heat
exchanger design.

The exchangers are mounted at a point near
their center of gravity by a gimbal-type joint that
permits rotation to accommodate the unequal
thermal expansions in the inlet and outlet pipes.
The entire heat exchanger assembly is suspended
from an overhead support structure.

Through close material control and inspection,
the heat exchangers are expected to have a high
degree of reliability and to last the 30-year life of
the plant. If maintenance is required, however,
provisions have been included in the design for
replacement of tube bundles. No specific arrange-
ments are made for replacement of the shell,
although this could be accomplished but with more
difficulty.

STEAM GENERATORS AND REHEATERS

Each of the four primary heat exchangers has
coolant salt circulated through it by a coolant-salt
pump operating in a closed loop independently of
the other three pumps. Each of the coolant-salt
loops contains four horizontal U-shell, U-tube
steam-generator superheaters and two horizontal
straight-tube reheaters, connected in parallel to
the coolant-salt system. A flow-proportioning
valve balances the salt flow between the steam
generators and reheaters to provide 1000°F outlet
steam temperature from each. The coolant-salt
pump speeds can be varied in proportion to the
electric load on the plant. Design data for the
steam generators and reheaters are listed in
Tables VI and VII. |

The steam generators are supplied with ~ 10 X
10® 1b/h of demineralized feedwater at ~700°F and
3800 psia. A supercritical pressure steam cycle
was selected because it affords a high thermal ef-
ficiency and also permits steam to be directly
mixed with the high-pressure feedwater to raise
its temperature to ~700°F to guard against freez-
ing of the coolant salt in the steam generator. For
the same reason, prime steam is also used to

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

MSBR DESIGN AND PERFORMANCE

TABLE VI

Steam Generator-Superheater Design Data

 

 

Each of
16 Units
Thermal duty, Btu/h 415 x 10°
Tube-side conditions:
Fluid Steam
Entrance temperature,’F 700
Exit temperature, °F 1000
Mass flow rate, 1b/h 6300 x 10°

Number tubes 384
Nominal tube o.d., in. L

2
Tube thickness, in. 0.077
Tube pitch, in. 0.875
Tube length, ft. 73
Effective surface, £t 3628
Pressure drop, psi 150 -
Entrance pressure, psia 3750

Shell-side conditions:

Fluid Coolant salt
Entrance temperature, °F 1150

Exit temperature, °F 850

Mass flow rate, 1b/h 3.85 x 10°
Pressure drop, psi 60

Baffle spacing, ft 4

Inside diameter of shell, ft 1.5

Overall heat transfer coefficient, based
on log mean At, tube o.d., Btu/h ft* °F)| 764

 

 

 

 

temper the ‘‘cold’’ reheat steam to 650°F before it
enters the reheaters.

After encountering piping stress and other
layout problems with vertical units, it was found
that by making both the steam generators and re-
heaters horizontal, the piping was simplified,
stresses were lowered to within acceptable limits,
and cell heights could be decreased.

SALT-CIRCULATING PUMPS

Both the primary and secondary salt-circu-
lating pumps are motor-driven, single-stage,
sump-type, centrifugal pumps with overhung im-
pellers as illustrated in Fig. 8. A helium cover
gas is used in the salt circulating systems, a por-
tion of this gas being fed continuously through the
labyrinth seals on the pumps to prevent leakage of
fission products from the systems.

A design criterion adopted for the MSBR con-
ceptual study was that pressure drops in the salt
circuits be limited to a total of ~ 150 ft of head, a
pressure that can be developed by a single-stage
pump. Single-stage pumps, with overhung impel-
lers and the upper bearing lubricated by a circu-
lating oil system, have successfully operated at

FEBRUARY 1970 201
Bettis and Robertson

TABLE VII

Steam Reheater Design Data

 

 

Shell-side conditions:

Each of
8 Units
Thermal duty, Btu/h 1.25 x 10°
Tube-side conditions:
Fluid Steam
Entrance temperature, °F 650
Exit temperature, °F 1000
Mass flow rate, 1b/h 6420 x 103
Number tubes 397
Nominal tube o.d., in. 3
Tube thickness, in. 0.035
Tube pitch, in. 1
Tube length, ft 27.3
Effective surface, ft? 2127
Pressure drop, psi A
Entrance pressure, psia o'70

 

 

Fluid Coolant salt
Entrance temperature, °F 1150
Exit temperature, °F 850
Mass flow rate, 1b/h 1.16 x 10°
Pressure drop, psi 58
Baffle spacing, in. 8.4
Inside diameter of shell, in. 22
Overall heat transfer coefficient, based
on log mean At, tube o.d., Btu/ (b ft* °F)| 340

 

 

ORNL over many thousands of hours in test loops
and in the MSRE. Although development will be
required to produce the high-capacity, single-
stage pumps needed for the MSBR, no major tech-
nological problems are foreseen.

Multistage pumps having a lower bearing would
be required to produce a higher head. Salt-
lubricated bearings have been tested with some
success at ORNL, but it was judged that the
single-stage design is a more conservative ap-
proach and requires less development effort. A
further advantage of not using the salt-lubricated
bearing is that tests of single-stage units have
demonstrated that, when drained of salt, significant
amounts of nitrogen gas can be moved by the
pumps when operating at ~ 1200 rpm. Advantage
is taken of this in planning for removal of after-
heat in the reactor core in event of an unscheduled
salt drain.

DRAIN TANKS

Both the primary and secondary salt-circu-
lating systems are provided with tanks for storage
of the salt during start-up and for emergency or
maintenance operations.

202

NUCLEAR APPLICATIONS & TECHNOLOGY

MSBR DESIGN AND PERFORMANCE

 

 

 

 

! MOTOR t

 

 

 

 

 

 

 

 

   
 
  
   
  
 
  

 

 

 

 

Bf L U060,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BELLOWS
~ ASSEMBLY

 

    
  
 

CONCRETE

SHIELDING
S
4
EARING AND
EAL ASSEMBLY
13 1 2 in.
REACTOR CELL
CONTAINMENT
| ~———————SHIELD PLUG
|
|
N _— PUMP TANK
| (6 ft Oin.OD)
|
z!
i
i
——C:: t‘
O 5 10 15 20 25
INCHES
Fig. 8. Sectional elevation of primary-salt pump for

MSBR station.

VOL. 8

FEBRUARY 1970
Bettis and Robertson

Design objectives for the fuel-salt drain tanks
are that the highly radioactive contents be con-
tained with a high degree of integrity and relia-
bility and that the afterheat-removal system be
largely independent of the need for mechanical
equipment, emergency power supply, or initiating
action by the operating personnel.

The fuel-salt inventory of ~ 1720 ft’ is stored
in a single tank, ~ 12 ft in diameter X 20 ft high,
located in the drain-tank cell. Since there is in-
sufficient moderation a critical mass could not
exist. The cell is heated to maintain the salt
above 930°F, although there would be no permanent
damage in the very unlikely event that the salt
were to freeze in the tank. To limit the tempera-
ture rise after a drain due to fission-product
decay heat, the tank is equipped with internal U-
tubes through which sodium fluoroborate coolant
salt is circulated by thermal convection to a salt-
to-air exchanger located at the base of a natural
convection stack, as indicated in the flow sheet,
Fig. 1. There are two sets of U-tubes and air-
cooled exchangers, each of which alone is capable
of maintaining the tank at acceptable tempera-
tures.

The fuel-salt storage tank is connected to the
circulating system by a drain line having a
freeze-plug-type ‘‘valve.’”” At the discretion of the
plant operator, the plugs could be thawed in a few
minutes for gravity drain of the salt into the tank.
The plugs would also thaw in event of a major loss
of power or failure of the valve coolant system.
The salt is returned to the circulating system by
pressurization of the drain tank.

A coolant-salt storage tank, ~ 10 ft in diameter
x 20 ft high, is provided for each of the coolant-
salt-circulating loops and for the two heat re-
moval systems on the fuel-salt drain tank. The
six tanks have a total storage volume of ~ 8400 ft’
of coolant salt. The tanks are located in a heated
cell which maintains the temperature above 800°F.

OFF-GAS SYSTEM

The volatile fission products and colloidal
noble metals formed in the fuel salt are largely
removed by a helium-gas purge system. The gas-
handling and -disposal system 1is shown sche-
matically in Fig. 1.

About 10% of the fuel salt discharged from each
of the four circulating pumps is bypassed through
a gas separator. Swirl vanes iIn the separator
cause the gas bubbles to collect in a central vor-
tex from which they are withdrawn with nearly
100% stripping efficiency. Downstream vanes
then kill the swirl. It is desirable that the radio-
active gas flow from the separator carry some
salt with it (perhaps as much as 50%) to act as a
sink for the decay heat. The separated gas (~9

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

MSBR DESIGN AND PERFORMANCE

scfm) then flows through an entrainment separator
to remove residual salt before it enters a particle
trap and 1000-ft* decay tank. The salt removed in
the entrainment separator returns to the pump
bowl.

The main bypass flow of salt leaving the gas
separator described above then passes through a
bubble generator where the salt is recharged with
the helium gas used to strip xenon and other fis-
sion product gases from the reactor core. The
bubble generator is a venturi-like section in the
pipe capable of producing bubble diameters in the
range of 15 to 20 mils. . The fraction of xenon
remaining in the reactor core depends upon the
ratio of the surface area of the bubbles to the sur-
face area of the graphite—and on the permeability
of the graphite to xenon. Based on the belief that
the graphite coating reduces the permeability to
xenon as effectively as the tests now indicate, the
volume fraction of bubbles circulating with the
fuel salt probably need be no more than ~ 0.5%.

The particle trap for the gas leaving the gas
separator also receives about a 2-scfm flow of
helium gas from the pump bowl. The helium is
used to purge the reactor of volatile fission pro-
ducts and ‘‘smoke’’ of extremely small-sized
noble metal fission-product particles which tend
to accumulate above the liquid interface in the
bowl. The particle trap tank is designed to re-
move a fission-product heat load of ~ 18 MW (th).
The entering gas impinges on the surface of a
flowing lead-bismuth alloy where heavy particles
are trapped while the gas passes over the surface
and is cooled. The lead-bismuth is circulated by
two 900-gal/min motor-driven pumps to the shell
side of heat exchangers cooled by a closed-circuit,
boiling-water system coupled to an air-cooled
condenser. The liquid metal is circulated to a
chemical processing cell for removal of the
captured fission-product particles.

The highly radioactive gas decays for ~1h in
the tank and then passes to one of two water-
cooled charcoal holdup traps. These water-cooled
traps absorb and hold up the gases, with the xenon
being retained for ~47 h. On leaving the charcoal
beds, ~9 scfm of the gas passes through a water
detector and trap to a gas compressor which re-
circulates it to the gas purge system bubble
generator. The remaining 2 scfm is held up for
further decay and stripping of tritium and other
gases before it is returned to the off-gas system
for purging of the pump seals, etc.

With the exception of the lead-bismuth particle
trap, the components proposed for the off-gas
system have been tested in principle in the MSRE
or in experimental loops at ORNL. The off-gas
system is described in more detail by Scott and
Eatherly.?

FEBRUARY 1970 203
Bettis and Robertson
FUEL-SALT PROCESSING

The nuclear performance of the MSBR is
enhanced by continuous removal of the fission
products and nuclear poisons. There are also
advantages in plant safety to continuous reduction
and disposal of the inventory of fission products in
the system. Methods of treating the fuel salt are
described in an article by Whatley et al.® Very
briefly, the process consists of diverting a 3-
gal/min sidestream of circulating fuel salt to an
adjacent cell for reduction extraction using bis-
muth. The Pa is removed on about a 3-day cycle
and the rare earth fission products on about a 50-
day cycle. It may be noted that the reactor can
continue to operate for relatively long periods even
if the fuel processing facility is out of service.

STEAM-POWER SYSTEM

The MSBR can produce high-temperature, high-
pressure steam to make use of advanced-design
steam turbine-generator equipment.

The steam system flow sheet, Fig. 9, considers
a supercritical pressure, 1000°F, with reheat to
1000°F, steam turbine very similar to that used in
the TVA’s Bull Run plant. This is a cross-
compounded unit of 1035 MW(e) gross capacity
with the load about equally divided between the
single-flow, 3600-rpm, high-pressure and two-

 

 
  

 

 

 

 

MSBR DESIGN AND PERFORMANCE

flow intermediate pressure turbines on one shaft
and the 1800-rpm, four-flow arrangement of low-
pressure turbines on the other shaft. It seems
certain, however, that tandem-compounded units
of this capacity will be available to a future MSBR
power industry. Credit for this has been taken
in the cost estimates.

Eight stages of regenerative feedwater heating
have been shown, with extraction steam taken
from the high- and low-pressure turbines and
from the turbines used to drive the two boiler feed
pumps. The system is conventional except for the
precautions taken to preheat the entering steam or
water above the liquidus temperature of the cool-
ant salt. The 550°F steam from the high-pressure
turbine exhaust is preheated to ~650°F before it
enters the reheaters. This is accomplished in the
shell side of an exchanger heated by high-pressure
steam at throttle conditions in the tubes. The
high-pressure heating steam leaving the tube side
1s then mixed directly with the high-pressure
550°F feedwater leaving the top extraction heater
to raise its temperature to ~700°F. Motor-driven
pumps then boost the feedwater pressure from
~ 3500 psia at the mixing ‘‘tee’’ to the steam
generator pressure of ~ 3800 psia. These pumps
are similar to canned-motor supercritical pres-
sure boiler recirculation pumps in current use.

The allowable temperature margins to guard
against significant freezing of the coolant salt in

 

 

 

 

 

 

 

540 psia  1000°F 5.1 % 10% |b/h
________________________ -
o ! 3600 psia  1000°F 10 X 10° Ib/h |
T T T Tyt T T T T T T T T i |
1125°F or | Reheater | | '
S TS E 204 Mw(th) | |
< & - Preheater |
< ) 1101 MW th) _ |
< < | 600 psia  552°F _ 227 MWee)
sore| | woe|y] e - SEES———T TT
~—_[ JSteam 650°F | | —— - | 1172 psia 706 °
22T 6°F
Generator l {— r—‘ L PHe
1932 MW (th): | . ; |
| I | |
. o 61 .
3500 psia _83¢° 2.9 x10%j1b/h | Gen.| 508 Mw(e)
I l B R |
| | F Ll |1.5 in. Hg abs
Booster | | b ! '
| | t) L '
Pumps | \ | LCondenser—l | LCondenser]
4.6 MW(e) | L g |
(each) | r r-?jl—fl ij_ Ti |
| | "3 Stages L — 5 Stages | : Reactor heat = 2250 MW(th)
Mixing "Tee" | Co T T T T | Gross output = 1035 MW(e)
_J FW Heating - FW Hidtlmg li Net output = 1000 MW(e)
= ————X = O,
6950 r 5500F:_:_3_§;_‘___ fl'# L_;_Slg_g_ei_iJ Net eff. 44.4 /0
Mixing "Tee" Lo 2290 | FW Heating

FW Heating

Fig. 9. Simplified steam-system flow sheet for 1000 MW(e) MSBR station.

204

NUCLEAR APPLICATIONS & TECHNOLOGY

VOL. 8 FEBRUARY 1970
Bettis and Robertson

the heat exchangers are yet to be determined ex-
perimentally. No optimization studies of the feed-
water and reheat steam systems have been made
and there are obviously many possible variations.
For example, if 580°F feedwater and 550°F ‘‘cold”’
reheat steam could be used, thus eliminating the
heating and mixing processes suggested above, the
flow through the steam generator could be re-
duced by one-third and the overall plant thermal
efficiency increased from ~44.4 to 44.9%.

Four control rods of medium reactivity have
been provided. The effective portions of the rods
are graphite, which on removal from the core de-
crease the neutron moderation and the reactivity.

The basic principle of MSBR control takes into
account its load-following characteristics. An in-
crease in turbine load would increase the steam
flow rate and tend to lower the steam outlet tem-
peratures from the steam generators -and re-
heaters. The coolant-salt pump control would
increase the pump speed to maintain constant
1000°F steam outlet temperatures, thereby tending
to lower the salt temperatures in both the primary
and secondary salt loops. The combination of the
negative temperature coefficient of reactivity and
possible control rod movement would then in-
crease the reactor power.

The MSBR control system has not been studied
in detail, but no major problems are foreseen.
Preliminary analog computer studies indicate that
the primary, secondary, and steam-system loops
have greatly different time constants so that
power oscillating and control rod position hunting
problems can be avoided.

AFTERHEAT REMOVAL

The normal shutdown procedure after taking
the reactor subcritical would be to continue to
circulate both the salts so as to transfer afterheat
to the steam system for rejection to the condens-
ing water. The afterheat load would decrease by a
factor of 10 in two to four days. After several
days of circulation, the fuel salt could be drained
to the storage tank without concern for over-
heating of the primary loop due to the deposited
fission products remaining in the system.

In the event of excessive salt leakage or other
malfunction which would require quick drainage of
the fuel salt, nitrogen could be introduced and cir-
culated at ~50 psig in the fuel-salt loop. Pre-
liminary estimates indicate that three of the four
fuel-salt-circulating pumps (operating at 1200
rpm) can circulate enough gas to keep the temper-
atures within tolerable levels. This assumes that
at least one of the four secondary-salt pumps is
operative.

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

MSBR DESIGN AND PERFORMANCE
MAINTENANCE

Any portions of the system containing fuel salt
will become sufficiently radioactive to require
that all maintenance operations be accomplished
by other than direct methods. The coolant salt
will experience induced activity, primarily due to
activation of the sodium atoms. The coolant-salt
pumps, steam generators and reheaters, however,
will not become radioactive, and after the coolant
salt is drained and flushed, could probably be ap-
proached for direct maintenance.

Cell roof access openings will be provided over
all equipment that might need servicing. The
plugs are replaceable with a work shield through
which long-handled tools can be manipulated. A
bridge-type manipulator will be mounted beneath
the work shield to perform more sophisticated
operation. Viewing can be accomplished through

‘cell windows, periscopes, and closed-circuit tele-

vision.

Replacement of the reactor core graphite will
be a maintenance operationthat can be planned in
considerable detail. The assembly to be removed
is ~16 ft in diameter, 30 ft high (including the
cylinder extension), and weighs ~ 240 tons. After
10 days of decay time the activity level at the
outer circumference would be ~5 X 10° R/h on
contact. The first step in replacing the core
graphite would be to disconnect the attachments
and holddown devices. A transition piece and
track-mounted transport cask would then be posi-
tioned on the operating floor directly above the
reactor, and a lifting mechanism inside the cask
used to draw the core assembly up into it. The
cask would then be closed by means of sliding
shields, moved into position over the reactor
storage cell, and the spent core assembly lowered
into it.

A replacement core assembly would have pre-
viously been made ready for lifting into the trans-
port cask and subsequent lowering into the reactor
vessel. Since this arrangement requires two top
heads and control rod drive assemblies for the
reactor vessel to be on hand, these have been in-
cluded in the direct construction cost of the plant.
Very preliminary estimates indicate that the capi-
tal cost of special maintenance equipment would
be in the range of $6 million. These studies give
promise that the core graphite can be replaced
within the usual downtime allowance for other
plant maintenance. :

Although there is some background of exper-
jence in maintenance of the MSRE and other re-
actor types, obviously the above comments
regarding MSBR maintenance procedures and
costs are speculative and will require further
evaluation in continued study and testing. |

FEBRUARY 1970 205
Bettis and Robertson
COST ESTIMATES

Methods used in making the cost estimates for
an MSBR power station conform largely to those
used by the USAEC in reactor cost evaluation
studies. It is important to note that the MSBR
estimated construction cost is not that of a first
plant but that of a new plant in an established
molten-salt reactor industry. It is further as-
sumed that there is a proven design essentially

the same as the one described here, that develop-

ment costs have been largely absorbed, and that
the manufacture of materials, plant construction,
and licensing are routine. On this basis the con-
tingency and other uncertainty factors assumed
for MSBR costs are little different from those
applied to other reactor types.

Costs are based on the early 1969 value of the
dollar. It is assumed that it would require five
years to construct a plant. Escalation of prices
during the construction period was not included in
these estimates but could amount to ~ 12% addi-
tional cost. Interest (at 6%) on capital during the
5-year construction period was included, however.
Expenses of an extended (54-h) work week were
applied, as was a 3% local sales tax.

TABLE VIII

Estimated Construction Cost of 1000-MW (e)
MSBR Power Station

 

 

$10°2
Structures and improvements 11.5
Reactor
Vessel 5.9
Graphite 6.0
Shielding and containment 3.1
Heating and cooling systems 1.3
Control rods 1.0
Cranes 0.2
Heat transfer systems 22.2
Drain tanks 4.3
Waste treatment and disposal 0.5
Instrumentation and controls 4.1
Feedwater supply and treatment 5.2
Steam piping 5.6
Remote maintenance and equipment 6.6
Turbine-generator plant equipment 25.0
Accessory electrical equipment - 5.2
Miscellaneous 1.4
109.1
Total + 6% allowance for 54-h work week |$115.6
Total + 3% local sales tax $119.1
Total + 33.5% indirect costs (see Table
IX) $159.0

 

 

 

 

?Based on early 1969 prices.

206

NUCLEAR APPLICATIONS & TECHNOLOGY

MSBR DESIGN AND PERFORMANCE

 

 

TABLE IX

Explanation of Indirect Costs Used in Table VIII
(For direct cost of $1) Percent | Total Cost
General and administrative 4.7 1.047
Miscellaneous construction 1.0 1.057
Architect-engineer fees 5.1 1.111
Nuclear engineering fees 2.0 1.134
Start-up costs 0.7 1.142
Contingency 2.7 1.172
Interest during construction 13.5 1.331
Land ($360 000) - 1.335

 

 

 

 

 

Substantive cost data are available for making
MSBR estimates on about half the items in the
MSBR plant. The remaining costs are less certain
because of the preliminary nature of the reactor
system designs, and because the special graphite
and Hastelloy-N materials have accumulated little
large-scale production experience. The fuel cycle
costs are discussed in a companion paper of this
series. The estimated construction costs are
shown in Table VIII and the breakdown of indirect
costs in Table IX. The power production costs are
given in Table X.

The reactor vessel, heat exchangers, drain
tanks, etc., were estimated on the basis of an in-
stalled cost of Hastelloy-N parts of $8/lb for
simple cylindrical shapes, $12/l1b for vessel
heads, etc., and $20/1b for tubing and other special
shapes. The assumed installed graphite cost was
estimated at $8/1b for simple slabs and $10/1b for
the extruded core elements. The cost of an alter-
nate reactor head assembly is included.

Fixed charges for the capital, which include
interest depreciation, routine replacements (other
than reactor core), taxes, and insurance, total
13.7% for a plant owned by a private utility.

The estimated cost of replacing the core
graphite every 4 years is given in Table XI. It is
assumed that the replacement can be accomplished

TABLE X

Estimated Power Production Cost in 1000-MW (e)
MSBR Station

 

 

Depreciating capital® 3.1 mills/kWh
Fuel cycle costb 0.7
Operating cost 0.3

Total 4.1 mills/kWh

 

 

 

“Based on fixed charges of 13%/year and 80% plant fac-
tor.

bIncludes graphite replacement cost of 0.1 mil/kWh(e),
based on 4-year life of graphite and capitalization
costs totaling 8% for replacement expenses.

VOL. 8 FEBRUARY 1970

 
Bettis and Robertson

TABLE XI

Explanation of Graphite Replacement Cost
Used in Table X

 

 

 

 

 

$10°
Cost of graphite per replacement $ 3.3
Cost of Hastelloy-N per replacement 0.5
Special labor costs per replacement 0.3
$ 4.1
Replacement cost factor for 4-year life to
convert to present worth (at 6%), based on
30-year plant life (3.04)
Total 30-year replacement cost $12.4
Production cost of graphite replacement,
based on fixed charges of 8%, 80% plant
factor, mills/kWh 0.1

 

within the usual plant downtime for other mainte-
nance that is accommodated within the 80% plant
factor.

The general conclusion has been reached that
the direct construction cost of an MSBR is similar
to that of a PWR. This cost position of the MSBR
is largely due to the high thermal efficiency of the
plant and the use of modern high-speed turbine-
generators having a relatively low first cost per
kilowatt. The above features, combined with a low
fuel-cycle cost, permit MSBR stations to operate
with low power production costs and with high load
factors throughout their life.

In summary, this study of a conceptual design

NUCLEAR APPLICATIONS & TECHNOLOGY VOL. 8

MSBR DESIGN AND PERFORMANCE

for a 1000-MW(e) MSBR power station has indi-
cated that it is feasible to design practical ther-
mal breeders that would contribute significantly to
conservation of fertile fuel resources and be
capable of generating low-cost electric power both
in the near and distant future.

ACKNOWLEDGMENTS

This research was sponsored by the U.S. Atomic
Energy Commission under contract with the Union Car-
bide Corporation.

REFERENCES

1. A. M. PERRY and H. F. BAUMAN, ‘‘Reactor Physics
and Fuel Cycle Analysis,”” Nucl. Appl. Tech., 8, 208
(1970).

2. DUNLAP SCOTT and W. P, EATHERLY, ‘‘Graphite
and Xenon Behavior and Their Influence on Molten-Salt
Reactor Design,”’ Nucl. Appl. Tech., 8,179 (1970).

3. M. E. WHATLEY, L. E. McNEESE, W. L, CARTER,
L. M. FERRIS, and E. L, NICHOLSON, ‘Engineering
Development of the MSBR Fuel Recycle,”’ Nucl. Appl.
Tech., 8, 170 (1970).

4, H. E. McCOY, W. H. COOK, R. E. GEHLBACH, J. R.
WEIR, C. R. KENNEDY, C. E. SESSIONS, R. .. BEATTY,
A. P. LITMAN, and J. W. KOGER, ‘‘Materials for
Molten-Salt Reactors,’’ Nucl. Appl. Teck., 8, 156 (1970).

5. W. R. GRIMES, ‘“Molten Salt Reactor Chemistry,”’
Nucl. Appl. Tech., 8, 137 (1970).

FEBRUARY 1970 207