FFR_chap24.txt

CHAPTER 24

LIQUID METAL FUEL REACTOR DESIGN STUDY*

24-1. CompARISON oF Two-FLuip AND SINGLE-FLUID
LMFR DeEsignNs

In Chapter 18, the two-fluid and the single-fluid externally cooled LMFR
concepts were discussed in a general way. It was pointed out that the two-
fluid design has the better breeding possibilities but is somewhat more
complex than the single-fluid reactor. In this chapter a complete design
study of a two-fluid full-sized LMTR reactor is deseribed and discussed,
and a shorter discussion of a single-fluid design study follows. This does
not mean that one design is necessarily favored over the other. In fact
both of these designs are being studied very extensively.

24-2. Two-Fruip Rracror DESIGN

24-2.1 General description. The two-fluid externally cooled LMFR
concept consists of a relatively small core surrounded, for the most part, by
a blanket containing fertile material. The core is composed of high-density,
impervious graphite through which vertical channels are drilled to allow
circulation of the fuel coolant. The fuel in the core is dissolved U233 or
U233 dissolved and suspended in liquid bismuth. The fluid fuel also acts as
coolant for the core system. The required coolant to moderator ratio is
obtained by proper size and spacing of the fuel coolant channels.

The blanket is constructed of high-density graphite through which flows
a liquid bismuth slurry containing the bred U?2*3 fuel and thorium, the
fertile material. In this study, thorium is assumed to be suspended in bis-
muth as thortum bismuthide, although thorium oxide particles could be
used. The blanket is wrapped around the core as completely as possible
for good neutron economy. An important economic consideration is the
degree of end blanketing which can be achieved while keeping coolant
velocities below the allowable limit. Several blanket designs were in-
vestigated, but a complete study for obtaining the best end blanket design
has not yet been carried out.

*This chapter is based on studies made by Babeock & Wilcox Company for the
USAEC, BAW-1046, March 1958, and on a 17 company report BAW--2, June 30,
1955, for which Brookhaven National Laboratory contributed information and sup-
plementary design studies.

866
24-2] TWO-FLUID REACTOR DESIGN 867

24-2.2 General specifications. Unless otherwise noted, the specifications
listed below are common to all calculations performed in this design.

Total power 825 mw (thermal)
315,000 kw (electrical)

Coolant to moderator ratio in core, Vpi/Vc 1.22

Coolaut to moderator ratio in blanket, Vaury/Ve — 0.50

Core-blanket barrier material graphite

Blanket thickness 3.0 ft

Blanket slurry composition:
Bismuth 90 w/o
Thorium, as ThzBis 10 w/o

C'oolant inlet temperature 750°F

Coolant outlet temperature 1050°F

Nuclear ealeulations utilizing latest cross sections and multigroup diffu-
sion theory indicate that the values 1.22 and 0.50 listed above are close
to the optimum.

The =everal factors which dictated the choice of a bismuth-to-carbon
volume ratio merit some attention. There are some losses of neutrons due
to capture in graphite. Hence, one would wish to use only enough graphite
to <utliciently thermalize the reactor. If too little graphite is used, the
eritien] mass will be large. It is suspected that the % value for U23? may
be lower in the epithermal than in the thermal energy range. This would
nuke it desirable to keep the reactor thermal. It was found that bismuth-
to-carbon volume ratios in the range of 0.5 to 2.0 satisfy these various re-
quirements quite well. It may be further observed by referring to Iig. 24-1
thut breeding improves with an increase in the bismuth-to-carbon volume
ratio. However, the maximum bismuth-to-carbon volume ratio acceptable
on the basis of structural limitations was 1.22) and consequently this core
ditneter 15 155.7 em (61 in.) at a bismuth-to-carbon volume ratio of 1.22;
assiutning a cvlinder with its height equal to diameter.

Bianlt slurry-to-graphite volume ratio and blanket thickness. A series of
crliulitions were made to estimate the most economical parameter values
for the blanket. Blanket slurry-to-graphite volume ratio and blanket
thicknes=s were varied to give the best breeding ratio consistent with reason-
able hismuth holdup. Figures 24-2 and 24-3 demonstrate the effects of
varving blanket composition and thickness on breeding ratio. The slurry-
to-graphite volume ratio was set at 0.5 and the blanket thickness was set
at o1t

Studiy of design parameters. The parameters investigated in the following
analvsi= are (11 end blanket design, (2) power fraction in the blanket, and
(31 tizsion product poison level in the core.
868

 

0.14

0.12

0.10

o
o
®

Breeding Guain
o
o
>

0.04

0.02

 

 

     

 

 

LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24
[ I | T I I
Noa/Ng. ~ 15 x 10 4
23" 7B
] ] ] | ] ] 1 1
0.25 0.50 075 1.00 1.25 1.50 175 2.00 2.25

Bismuth te Carbon Volume Ratio

F1a. 24-1. Breeding gain vs. bismuth-to-carbon volume ratio in core.

Breeding Ratio

 

1.1

 

 

0.9 l ] L

2.0 2.5 3.0 3.5 4,0
Blanket Thickness in Ft.

 

Fia. 24-2. Breeding vs. blanket thickness for slurry-to-carbon volume ratio
= 1.00 and bismuth to carbon volume ratio in core = 1.00.

Breeding Ratio

 

 

 

 

i i ! [ i

 

1.04

0.50 0.75 1.00 1.25 1.5¢ 175

Slurry to Carbon Volume Ratio

Fia. 24.3. Breeding vs. slurry-to-carbon volume ratio in blanket for bismuth to
carbon volume ratio = 1.00 and blanket thickness = 3 ft.
24-2] TWO-FLUID REACTOR DESIGN 869

24-2.3 End blanket effects. A series of nuclear calculations were per-
formed to determine the effects of end blanket design upon breeding ratio
and eritical fuel concentration. Two extreme blanket designs were con-
s1dered. In the most optimistie case, a spherieal core, equivalent to a 61-in.-
diameter eylinder, was surrounded by a 3-ft spherical blanket. The pessi-
mixtic caleulations assumed a cylindrical core with a diameter of 61 in.,
height equal to 1.5 times the diameter, a 3-ft radial blanket, and no end
blanket. Critical values of fuel concentrations and breeding ratio were
caleulated for four power fractions in the blanket for cach design.

All caleulations were performed for hot, clean conditions with an average
temperature of 900°I. A two-group, multiregion code was used to solve
the diffusion equations, and a 37-group spectral code was used to determine
the two-group nuclear constants.  The results of these calculations are
taubulated in Table 24-1. The breeding ratio is decreased 0.20 to 0.25 by
completely eliminating the end blankets. This is due primarily to the
added neutron leakage out the ends of the core, despite the fact that the
core height 1s Inereased.  Although the eritical mass of fuel in the core is
higher without end blankets, the fuel concentration is somewhat lower
due to the increased core volume.

TasLe 24-1

CriTicaniTy Cancurations rorR Two-I'wuip LMEFR
WITH AND WITHOUT IKND BLANKETS

 

 

 

 

 

 

 

 

Vay/Npi X 10° Ratio of . Blanket
‘ \ Breeding : .
(use ———— | blanket power ratio thickness, (ieometry
-~ Core | Blanket | to total power ' ft
I - 509 152 0.0665 1.053 3.0 Full blanket
IT 530 034 0.205 1.051 3.0 ? ”
1 461 | 1600 0.445 1.039 3.0 7 ”
IV 7 436 2100 0.515 1.033 3.0 7 "
V 403 1050 0.272 0.80 3.0 No end blanket
VI 366 2100 0.425 0.82 3.0 v ”
VII 347 2808 0.492 0.83 3.0 v !
VIIT' 403 | 1050 0.272 — 4.0 v "

 

 

 

The actual core and blanket design is between the two extremes assumed
in these caleulations. The blanket can be extended beyond the end bound-
aries of the core, and a graphite reflector can cover the ends of the core
except for the coolant inlet and outlet. Cooling becomes a serious design
870 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24

0

C ] | L shielding

'

Control Rod/ET ]

Control Rod Drive

]
/L

    

Core Qutlet

       

        

 

 

 

    

  

  
   

Nozzle ///// /// Blanket C])uflef

    
  

T

7

S
R A B R

R R SRR R
R R R R RS
NN

pSSSSE
TR
R R R R R R AR

TN

.

 

 

 

 

 

 

 

A R T

 

]
|

    

Fuef Passage Blanket Inlet

Nozzle
Core Inlet

Nozzle

FiG. 24-4. Two-region, externally cooled liquid metal fuel reactor.

problem, if the end reflector is replaced with blanket material. The design
in Fig. 24-4 is a substantial improvement over no end blanket or reflector.
However, further improvement in breeding ratio could be achieved with
even better end blanket designs.

24-2.4 Power level in the blanket. For a given geometry, coolant-to-
moderator ratio, and thorium concentration in the blanket, specification of
the fraction of total fissions generated in the blanket establishes a unique
set of values for fuel concentration in the blanket, fuel concentration in the
core, and fissions generated in the core. For simplicity, the power generated
in a region is assumed directly proportienal to the fissions in that region.
The data in Table 24-1 indicate that breeding ratio changes very little
with large changes in the fraction of total power generated in the blanket.
This increase in blanket power results in an Increased ratio of resonance
to thermal absorptions, a phenomenum which tends to offset the additional
fast neutron leakage out of the blanket as blanket power increases.
24-21 TWO-FLUID REACTOR DESIGN 871

An economic analysis of the effects of changing the blanket power frac-
tion was performed to determine the optimum core-blanket power split
under equilibrium operating conditions. The parameters affecting this
choice are (1) fission-product poison levels in the blanket, (2) fission-
product poison levels in the core, and (3) chemical processing costs,

Frission-product poisons in the blanket. The chemical processing of the
blanket slurry accomplishes two things:

(1) The removal of bred U?3? from the blanket system at a rate necessary
to maintain the 17232 concentration in the blanket slurry at some equilibrium
value corresponding to the desired blanket power fraction.

(2) The removal of fission products from the blanket slurry.

If the blanket processing cycle 1s determined by the minimum removal
rate of 17238 for steady-state operation, a corresponding poison level in
the blanket 1= automatically set. If the blanket chemical processing cyele
1= determined by the poison level and is less than the cycle determined by
the above criteria, the bred fuel removed from the blanket must be fed
hack into both core and blanket to maintain steady-state fuel concentra-
tions. In this analysis the blanket processing eyele i all eases was assumed
to be based on the minimum removal rate to maintain steady-state U233
concentrations without feeding fuel into the blanket system.

(e mical processing cycle for blanket slwrry. The chemical processing was
a==umed to be performed continuously on the reactor site. Unless other-
wise specified, the fluoride volatility process is utilized as described in
Article 24-3.16. The chemical processing cycle for the blanket may be
caleulated 3] from the equation

_ ZW Mg [V 4 (Z1a/ Z.) (0 a)]

s s — (07

T'p = blanket processing cycle, days,

 

Tg

where

Z, = removal efhciency for uranium = 0.25,
Z13 == removal efficiency for protactinium = 0.04,
M, = mass of fuel in blanket system, kg,
h, a = ratio of Pa®33 to U233 in blanket,
3 = kg of fuel burned per Mwd = 1.05(1 4+ «23),
P, = total power, 825 Mw,
BR = breeding ratio,

Pg = blanket power, Mw,
872 LIQUID METAL FULL REACTOR DESIGN STUDY [cHAP. 24

and

Z

Tg
¥

0% (efh)d™ +
Y13

Qe

where

0% (eff) = an effective absorption cross section to account for resonance
and thermal absorption in 17233

3

o5 = average thermal flux over the blanket system,
Y13 = decay constant for Pa®33,

The poison level in the blanket depends upon T, and Tg 1= a function
of M J, b/a, breeding ratio, and power fraction in the blanket. All these
arviables are interrelated. The ratio b,/a 1s a function of Ty, but Tpis a
slowly varying function of b/a due to the low value of Z,3/Z, (0.16).
Breeding ratio is a slowly varying function of fission-product levels in the
blanket due to the heavy loading of fuel and thorium in that regiow.
The breeding ratio is sensitive to the poison level, and thus to the chemical
processing rate, in the core fuel solution. An iterative calculation procedure
was required to arrive at optimum values of T, fission-product poison
level in the blanket, and the power fraction in the blanket.

For a given chemical processing rate in the blanket, the fission-product
poison level was determined from the data in KAPL 1226 [4]. Relative
poisoning, RP, is defined as the absorptions in fission products per thermal
fission in fuel, while the fission-product poison fraction is the absorptions
in fission products per total absorption in fuel. Xenon and samarium are
treated separately and are not included in the term fission products. The
burnup, F, in a region is defined as the atoms of fuel fissioned per atom
present in the region. The burnup F at time T in the blanket is caleulated
from

. 0.866 T(Pg/1IP)
F= - ;B
Mas

Using this relation, the relative polsoning in the blanket was determined
for ench processing cycle from a graph of R versus I7 [4]. The RP curve
used is based upon high cross sections of all fission products with the excep-
tion of u low value for Zr¥s,

Xenon in the blanket. Xenon is removed from the blanket by the degasser.
Although the removal rate of fission-product gases cannot be determined
until experimental information becomes available, a poison fraction of 0.01
was assumed for Xe!33,
24-2] TWO-FLUID REACTOR DESIGN 873

Samartum n the blanket, The removal rate of samarium by chemical
processing wus neglected. The steady-state ratio of Z50/Z3 using ap-

propriate thermal absorption cross sections, is determined by the relation

8
5

Som= 1.42 X 10716¢ + 0.0126,

where ¢ = average thermal flux in the region of interest.

Fisston-product poisons in the core. The level of fission products, FP,
other than xenon and samarium, in the core 1s determined by the chemical
processing cyele for the core fuel solution. The steady-state value of FP
poisons in the core should be established by an economie balance between
the value of improved breeding ratio and inereased chemical processing
costs, The relationship between the core processing cycle, T, and the rela-
tive poizon, RP, in the core may be expressed as

 

d(RP) _RP
dF ~ F
and
0.866 T.(P./P)
F=——
My,
where
D
d(;;‘}') 1s the slope of the curve RP versus F [4],

Mgy = total mass of U23? in the core system.

The xenon and samarium poisons in the core are determined as described
for the blanket.

[eonomie optimization.  An optimization study was performed to de-

ternune the most economic power split between core and blanket systems
qind fission-product poison level for the core during equilibrium operation.
The tuel cost items which vary with these two parameters are (1) bismuth
Heventory. (29 fuel inventory, (3) fuel burnup, (4) thorium amortization,
5 thortum burnup, and (6) chemical processing.  Nuclear calculations
specitied the fuel concentrations for both core and blanket and breeding
ritin=. These values were then used to determine the chemical processing
cvele for the blanket and the pertinent costs.
874 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP, 24

2800

 

2600 -
2400 —
2200 |
2000 -
1800

1600 |—

Ngq/Np; in Blanket x 10®
5 ~ =
g8 8 8
|

o
o
<O
I

600 |~

400 -

 

200 —

 

 

 

Pg/Py

F1g. 24-5. Fuel concentration in blanket vs. Pg/P; for two-fluid LMFR fully
blanketed sphere.

Nuclear calculations. The values of the parameters investigated were

RP {(core) = 0.03, 0.09, 0.15,

Pp/P; = 0.10-0.50.

Since only a relative comparison was needed, all calculations were made
with a spherical core and complete 3-ft spherical blanket. The xenon poison
fraction was taken as 0.01, and the samarium steady-state value was com-
puted for each region in each case.

The fission-product poison level in the blanket cannot be determined
without first knowing the blanket processing cycle. As a first approach,
the breeding ratio for the hot clean conditions was used to determine the
cyele time from which the RP in the blanket was calculated as described
previously. The relative poison levels determined on this basis were as
follows:

 

Pg/P; RP (blanket)
10% 0.029
25% 0.048

50% 0.155
24-2] TWO-FLUID REACTOR DESIGN 875

 

750 T

 

 

 

T
700 — —
&
<
o
o
(=]
x 600 - —
=
z RP=0.15
[ye)
o™
z
RP=0.09
500 —
RP=0.03
450 I ] | i |
0 10 20 30 40 .50 60

Pg/P,

Fic. 24-6. Fuel concentration in core vs. Pg/P; for two-fluid LMFR fully blan-
keted sphere,

All eriticality calculations were performed using the specifications out-
lined 1 Article 24-2.2. Two-group diffusion theory was employed, and a
two-group. multiregion code was used for solving the diffusion equations.
A~ previously mentioned a 37-group spectral code was used to generate
the two-group coefficients. The critical concentration of fuel in the core
and blinket, breeding ratio, and neutron losses were determined for several
power =plits for each relative poison level in the core. The blanket power
fraction values of 10, 33.3, and 509 were used as reference values for com-
pari=oni, and the important nuclear parameters were determined from a set
of puranetric curves for these precise values. (Cases actually caleulated
corre=ponded very closely to the desired blanket power in most calecu-
lations-,

The nuclear parameters corresponding to these power splits are sum-
marized 11 Table 24-2.  Figures 24-5 and 24-6 show the variation of
Noy Vg oin both the core and blanket as the blanket power fraction
changes. This atom ratio of U2 to bismuth in the blanket ranges from
233 x 1077 to 2420 X 1078 for Pp/P;=0.10 to 0.50. In the core the
Ny N ratio decreases approximately 209 over the same range. The
TABLE 24-2

Resurts oF NUCLEAR CALCULATIONS FOR VARIOUS Power SpLITS

 

 

 

 

Relative | Relative fi:’ oes A}:’ Crage
isc i N2a/Nri X 108 Nag/Npi X 108, ermal | thermal
Case | P/l POISOIL | POISON | pp 1y (ygq |t 23/ BI Mg ke | %8 Mgz, kg flux in flux in
in in (core) (blanket) core blanket
core blanket
system system
I(a)|0.10 0.03 | 0.029 |1.0256| 1.132 620 368.7 255 53.2 77X 1013 | 5.20 X 1013
(b) (.09 1.007 | 1.132 664 395 255 53.2 15 4.97
(e) 015 0.978 | 1.132 732 435 .4 255 53.2 425 467
II(a)[0.3333] 0.03 0.0475 [1.007 | 1.132 554 215.8 1150 317 .13 x 1043 | 2.61 x 1013
(b) 0.09 0.993 | 1.132 599 233.5 1190 328 .40 2.39
(e) 0.15 0.978 | 1.132 667 260 1230 334 71 2.23
IT1 (a)|0.50 0.03 | 0.155 [0.980 | 1.135 494 154 .8 2420 834 62 x 1013 | 1.28 x 1013
(b) 0.09 0.959 | 1.135 542 170 2670 920 72 1.01
(¢) 0.15 0945 | 1.135 590 185 2760 951 .19 0.97

 

 

 

 

 

 

 

 

 

 

 

 

 

9.8

XdNIS8 NDISHd HOILOVHAY TAAd TVLAW dIndrl

$Z 'dVHD]
24-2] TWO-FLUID REACTOR DESIGN 877

2.4

 

22 B

$2rcx10'"]5

20} -

 

 

Average Thermal Neutron Flux in Core,
o
!

 

 

RP== 0% |
RP=.15
12 . - :
¢ 0.1 02 03 0.4 0.5 6.0

Pg/Py

Fic. 24-7. Average thermal flux in core vs. Pg/P; for two-fluid LMFR based on
a fully blanketed sphere at 825 Mw.

values of the average thermal neutron flux in the core and blanket are
graphed in Figs. 24-7 and 24-8, and BR in Fig. 24-9.
Bismuth tnrentory. The primary system volumes for Pp/P;= 0.33 and
0.50 are based on a six-loop capsule design. Each loop contains a bismuth
inventory of 245 ft3. 1f 509 of the power is generated in the blanket,
three loops contain blanket slurry and three contain U-Bi core solution.
It one-third of the power originates in the blanket, two loops are devoted
to the blanket system and four to the core system. If only 10% of the total
power 1= venerated in the blanket, a three-loop design is assumed for the
core sv=rem, and two small loops of 125 ft3 each are used for the blanket.
The reactor holdup has been estimated from the reactor drawing in Fig.
24-1. Fuel inventory volumes are summarized in Table 24-3.
Using the value of $2.25/1b of bismuth, 129 annual fixed charges, and a
densitv of 613.5 Ib/ft? (9.83 g/ce), the annual bismuth inventory charges
are
(S yr) = 165.6 (Ves + Vis),

Wiiere
1'es = inventory volume of core system, ft3,
I'vs = Inventory volume of blanket system, ft3.

Foud toendory, Five days’ holdup of fuel from both blanket and core is
assumed ror the chemieal processing plant. Pa?33 is held up for 135 days to
allow for decav to U233, Approximately 3% of the Pa?33 remains after
135 days and ix discarded with the fission-product waste. This loss, while
878 LIQUID METAL FUEL REACTOR DESIGN STUDY [caAP. 24

 

f |

—

!
il

Average Thermal Neutron Flux in Blanket, 252 B X 10-14

 

 

 

10 —
0.8 —
RP=.03
06 —
RP=.09
RP=15
y L
0 0.1 0.2 0.3 0.4 0.5 0.6

Pp /Py

Fic. 24-8. Average thermal flux in blanket vs. Pg/P; for two-fluid LMFR
based on a fully blanketed sphere at 825 Mw.

 

 

 

 

1.03 T T
1.02 — ]
1.0 |
1.00 |- ]
0
£ 095 |- _
o
on
£
-
@
¢ 098 |- —]
om
0.97 = RP== 03]
0.96 |— ]
095 — —
RP=.09
RP==0.15
0.94 I | ] | l
0 A0 .20 .30 40 50 40

Fic. 24-9. Breeding ratio vs. Pp/P; for two-fluid LMFR fully blanketed sphere.
24-2] TWO-FLUID REACTOR DESIGN 879

TaBLE 24-3

INvENTORY VoLuMES IN Two-¥Fruip LMFR

 

Pp/P;=0.10 P,/P,=0.333 P/P.=0.50

 

 

Core svstem:

Reactor 275 ft3 275 ft3 275 ft3
External system 1640 980 735
Subtotal 1915 1255 1010

Blanket system:

 

Reactor ! 495 495 495
External svstem 250 490 735
Subtotal 745 985 1230

Total 2660 2240 2240

 

 

 

 

 

quite small, has been included with the fuel inventory charges, which may
be expressed as

(o (% vr =626 MG (1 + TO) + ME (1 + _T-S-) + 2 ME
b 1][233 Z13'

a Tg

 

(1 + 13;’,? 3) + 30 4 132,000

This equation assumes a 30-kg inventory of U233 feed material external
to the reactor. The economic assumptions used in this equation are 497
fuel fease charges and a U233 price of $15.65/g.

Foucl boirup. The annual cost of the net U232 fuel burned in an 825-Mw
reactor. a=suming an 809 plant factor, is

(3 ($/yr) = 3.96 X 10° (1 4+ a23)(1 — BR).

Thorinm amortization charges.  Assuming a cost of $42,/kg for thorium
and an annual amortization rate of 159, based on a 20-yr life, the annual
amortization charges for the thorium are

Cy ($/yr) =6.3 Mys.

Thortiwm burnup. The thorium replacement costs due to burnup are cal-
culated according to the equation

05 ($/y1‘) = 10,620 (1 +0523) BR.
880 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24

 

 

 

 

 

3000 r I T I r T l r
B RP =15 7
| Blanket RS = 09 ]
—=——-Core RP = .03

1000 | T .

o [ - ——RP= .]SE

o = =

a [— —]

@ [ —_— ]

£ F e E

= = 4

® |— ]

~ — —

> L

o

o - 4
£

w — —
2
U

) _

& -

100 ]

5 -— 3

[— T ~—RP = 03—

501 | ] I ] I ] I L1 ]

Q.10 20 30 .40 .50 .60
Pg/Py

Fig. 24-10. Chemical processing cycles vs, blanket power, based on a blanketed
sphere with total reactor power of 825 Mw and the removal efficiencies of Z,, = 0.25,
Zy3=0.04, ZB = 0.10, ZEp = 1.00.

 

    
 
   
 
   
     

 

 

 

 

5x]06 T T T 1 T T 1 T T 1 T T T
1
B Total Annual Cost
-
‘“_: |
o
"; 1061 Capital Equipment ]
£ — -
g — i
3 — 7
° - . ]
s 6 — Operating Cost 3
T [— -
o —
£ _
c
L - i
I _
—
| Building Cost
10° ! [ bt P Lt bl !
1 3 6 10 30 60 100 200

Plant Liquid Throughput, Cu Ft Per Day

Fic. 24-11. Annual fluoride volatility processing cost vs. plant throughput for
825-Mw-two-fluid LMFR.
24- 2] TWO-FLUID REACTOR DESIGN 881

 

  
 
    

 

 

 

6
810 T T T T T T T T T T T T
6 |- e
4 rs
A - P -
S i Total Annual Cost /’ 7
D 2 - ]
£ -
gl |
§ Capital investment Return
a 5 -
3 10 N e —
E s | -7 Operating Cost j
a4 L B o
6 -7 —
AX]OS 1 I L I L 1 L IJIllllJlli 1 I L 1 L J 1 llll]lll]l
1 3 6 10 30 60 100

Plant Throughput, kg Therium Per Day

Fie. 24-12. Annual aqueous processing costs vs. plant throughput for 825-Mw-
two-fluid LMFR.

("hemical processing costs. The chemical processing cycle time for the
blanket 1= determined by the Pg/I’ ratio and the breeding ratio, as dis-
cu==ed m previous paragraphs. The processing rate for the core system is
determined by the method also deseribed previously; see Fig. 24-10. The
totul throughput to the fluoride volatility chemical separations plant is
~imply:

Vcs Vbs

T T Ty

 

Throughput (ft3/day) =

The wnnmual processing charges based on fluoride volatility can be read
directly trom Iig. 24-11, a plot of annual charges versus plant through-
put.

A=~ & matter of comparison, the chemical processing charges were also
computed for each case, assuming on-site aqueous processing methods.
The capacity and eost of an aqueous processing plant are determined by
the wnmount of thorium per day which must be processed. The core solu-
tion provessing does not enter into the cost unless the ratio of fuel to thorium
present= criticality problems in the process equipment., This situation is
likely 1o oceur for the higher power levels in the blanket. This analysis did
not take this possibility into account, however, and annual aqueous
processing costs were taken directly from Fig. 24-12. This design plant
capacity 1= 33 kg day of thorium feed.

Results of optimization. 'The bismuth inventory is slightly greater for
the case of I’g I’; = 0.10 than for the other two cases, because of the added
primary system volume. Fuel inventory charges are not very sensitive to
882 LIQUID
.40

METAL FUEL REACTOR DESIGN STUDY

 

36}
32}
28—
24—

20—

Relative Fuel Costs, Mills/KWH

-.04+
-.08

 

| l l l l

 

Thorium Charges

RP=.157]
RP= .09
RP=.037]
Fuel inventory
;
Bi Invemory/
RP = .15

  

 

 

A0 .20 .30 .40 .50 .60

Pg/Py

[cHAP. 24

Fig. 24-13. Relative fuel costs vs. blanket power for two-fluid LMFR based on a

fully blanketed sphere operating at 825-Mw with a plant factor of 809,.

2.00

 

1.801

1.601—

1.40—

1.201—

1.00H—

Chemical Processing costs, Mills/KWH

0.80 |—

0.60—

0.40

 

   
      

Aqueous On-Site

Fluoride
Volatility

RP=0.09

RP=0.15

 

 

0.10 .20 .30 A0 .50 .60

Pg/Py

I'1g. 24-14. Chemical processing costs vs. blanket power. The cycle times are
based on a blanketed spherical reactor with a total heat power of 825-Mw.
24-2] TWO-FLUID REACTOR DESIGN 883

 

    
  
 

 

 

 

2.4
| I l 1 l
2.2 —
20 | —
=
X
"
28 | ]
7
S On-Site Aqueous
Tg 16 Lo Processing B
[V
.qaa RP =.03
I
@ 14— RP =.15.
RP = .09
RP =15
Fluoride RP = .09
12 — Vo|a?i|i.ry RP =.03
Processing
o | 1 1 | 1
0 10 .20 .30 .40 50 .60
PB/Pi

Fi6. 24-15. Relative fuel costs vs. blanket power for a blanketed spherical re-
actor operating at a total power of 825 Mw with a plant factor of 809.

the relative poison level in the core, but they increase sharply with an in-
crease In power level (Fig. 24-13). Thorium charges increase linearly with
blanket svstem slurry volume, and fuel burnup charges increase as Pg/P,
increases, as shown in Fig, 24-13.

Chemical processing costs drop rapidly as the power fraction in the
blanket increases. The increased processing rate required to maintain a
stendy-state fission-product relative poison level in the core of 0.03 results
i u processing cost much higher than required for RP values greater than
0.09. The aqueous processing costs appear to become essentially equal to
fluoride volatility costs at a value of 509, for Pg/P,. Further analysis
would be required to determine the validity of the aqueous processing cost
curve for low throughput and high N23/Np2 ratios encountered in the
cases of high blanket power. The chemical processing costs are tabulated
in Table 24—4 and shown graphically in Fig. 24-14,

The results of the economic comparisons are summarized in Table 24-5
and are graphed in Fig. 24-15. (RP on the graphs refers to the relative
poison level of the fission products in the core.) Figure 24-15 shows that
for all values of RP a minimum fuel cost occurs for a Pg/P; of approxi-
mately 0.33.

24-2.5 Selection of a reference design. The optimization study indi-
cated that the most economic reactor design should produce one-third of
TaABLE 244

Cuemicar, ProcessiNng Costs Two-Fruip LMFR

 

Slurry

 

 

 

 

U-Bi
flow flow to C.hem— Fluoride | Fluoride Th/day
Blanket | rate to : ical . L to Aqueous | Aqueous
i chemi- volatil- | volatil- .
o process | chemi- | RP T, plant |. - __4+ | chemi- | process- | process-
Case | PB/P: ! cal 1ty costs, ity costs,| TB(Z,=1) . :
cycle, cal |(core)| days thru- | 7, . cal ing cost, | ing cost,
plant, $/yrx | mills/ b
days | plant, ft2/day put, 10-6 kwh plant, [$/yrx10" 6mills/kwh
ft‘g/d&y, Vcs/T ’ ft3/day, M()Q/ TB
Vos/ T ‘

I1(a)0.10 16.26 | 45.8 0.03 75. 25.3 71.1 3.03 1.37 65 325 4 .20 1.90
(b) 16.59 | 44.9 | 0.09 | 557 3.44 | 48.3 2.57 1.17 66 320 4.15 1.88
(¢) 17.08 | 43.6 | 0.15 | 1116 1.71 | 45.3 2.51 1.14 68 310 4.10 1.86

Il (a) |0.3333] 122.3 8.05 | 0.03 60 20.9 29.0 2.15 0.974 489 57.1 2.10 0.951
(b) 129 7.64 | 0.09 | 446 2.81 | 10.5 1.65 0.747 516 54 1 2.08 0.937
(c) 136.3 7.23 1 0.15 1 900 1.39 8.62 1.58 0.716 525 53.1 2.06 0.933
11T (a) { 0. 50 444 5 2.77 10.03 56.5| 17.9 20.7 1.95 0.883 1778 19.6 1.38 0.625
(b) 513 2.40 | 0.09 | 432 2.34 1 4.74 1.40 0.634 2052 17.0 1.32 0.598
(¢) 547 2.25 |1 0.15 | 854 1.18 3.43 1.33 0.602 2188 15.9 1.30 0.589

 

 

 

 

 

 

 

 

 

 

 

 

 

 

788

AdALS NOISHd YOLOVAY THAd TVIHW dindIl

¥z "dVHD]
TaprLem 21

D

Revartive IFven Cost vor Two-'Lumn ENMEFR
(WrtH PorsoNs-BLANKETED SprisRy)

 

 

 

 

 

 

 

 

 

Bismuth Fuel Thorium Fluoride | Total costs inel, A ueoUs Total costs incl.
il;\'en— im’eil- Fuel ven- Thorium | volatility fluoride vol. : ;roc - AQUeOUS Proc-
y ' ' burnup burnup, | process- pProc. ' essing
tase | ’g/P tory tory L1 tory, ol
Case Bre C :Ifi)’,, 3l o >(<)Il‘\0’”3 C3x10°3 o 21]}0,3 CsX1073 | ing costs, |~ - o (;HI(A)L'?
e | S S| S/yr | Cpx1078 Cix1073 ) mills R0 00 x1078 ], mills
T T o $ar v “'kwh o $/yr “kwh
T(a);0.10 440 353 —116.5 133 12.3 3031 3852 1.75 4200 5022 2.28
(b) 440 303 —31.4 133 12.1 2570 3477 1.58 4150 5057 2.29
{¢) 440 369 98 .6 133 11.8 2510 3562 1.61 4100 5152 2.33
IT (a) | 0.3333 7 399 —31 .4 176 12.1 2150 3077 1.39 2100 3027 1.37
(h) 71 407 31.4 176 11.9 1650 2647 1.20 2070 3067 1.39
(¢) 371 429 98. 6 176 11.8 1580 2666 1.21 2060 3146 1.44
III (a) | 0.50 371 678 89,8 220 11.8 1950 3311 1.50 1380 2741 1.25
(b) 371 731 184.5 220 11.6 1400 2018 1.32 1320 2838 1.29
(¢) 371 766 247 220 11.4 1330 2945 1.33 1300 2915 1.32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NDISEd MOLOVAY dINTId-OML [Z-%2

G8%
8806 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24

the total power in the blanket system and that the relative poison in the
core due to fission products should be approximately 0.09. However, sev-
eral effects must be considered in relating the optimum reactor to the ae-
tual operating reactor. A geometry more realistic than the fully blanketed
sphere must be considered in establishing new specifications; effects of
higher uranium isotopes, Pa losses, and control rods on breeding ratio must
be taken into account; and a new chemical processing cycle for the blanket,
along with a new fission-product poison level in the blanket, must be cal-
culated based upon the adjusted breeding ratio.

Geometry effects. The inability to wrap a blunket around the ends of
the core requires an adjustment to the parameters for the reference design
hased on the caleulations with a full blanket. The axial leakage out of a
bare ended core and a blanket with a height 1.5 times its diameter was cal-
culated to be 0.18 neutron per absorption in fuel. An extension of the
blanket length and the addition of partiat end graphite reflectors are esti-
mated to reduce the end leakage to one-half this value. The total neutron
leakage, both fast und thermal, out of the partially blanketed reactor is
estimated at 0.17 neutron per absorption in fuel.

The added length of core and blanket will slightly increase the critical
mass, but the required Nas/Npi ratio will decrease slightly. In order to
be conservative in the fuel inventory costs, however, the critical values of
Nas/Nyg; for the fully blanketed sphere are assumed for both core and
blanket.

Breeding ratio. Higher uranium isotopes. The higher uranium isotopes,
primarily U283 U235 and U2, continue to build up in both the core and
blanket fuels throughout reactor life, since they cannot be separated in the
_chemical plant. The relative poison due to these isotopes, however, rises
rapidly at first with the buildup of U2 but increases very slowly there-
after. The return from U235 fissions almost balances for losses to U#3% and
U236 [4]. An average poison fraction of 0.01 for the reactor is used for the
reference design. ‘

Protactinium losses. The equilibrium Pa233 concentration can be com-
puted from the relationship

N =N,
using an effective thermal absorption cross section of Pa**3 based on the
calculated neutron spectrum in the blanket. The relative absorptions of
the Pa233 are very small (0.005), but they are included.
Control rods. The self-regulating properties of an LMFR have not been
established at this time. An allowance of 0.01 in relative absorptions is
included to account for the possibility of using a regulating rod and a small
24-2] TWO-FLUID REACTOR DESIGN

REFERENCE DESIGN SPECIFICATIONS

SPECIFICATIONS FOR HQUILIBRIUM OPERATION

Core:

Thermal power

Flectric power

Diameter, inches

Height, inches

Fuel

pi/Ve

Noy/Npi

Mass of U233 in system, kg

Total volume of fuel, ft3

Breeding ratio, over-all

Chemical processing cycle, days

Volume flow rate through chemical plant, ft3/day

Mas= flow rate through chemical plant, g 17233 /day

Average thermal flux in active core

Average thermal flux in core system
Blanlt:

Thermal power

Flectrie power

Thickness, ft
[ I'C
.\hmy content:
Thorinm (as ThsBis) 10% wt
Bi~muth 90% wt
Noy N (atom I‘thiO) 1190 x 10~¢

Miss of U299 1n system, kg

Mis= of thorinm in system, kg

Total volume of fuel, ft?

Chemieal processing eyele, days

Volume flow rate through chemical plant, {t%/day
Mas= flow rate through chemical plant, kg of Th/day

887

550 Mw
210,000 kw
61

91.5

233

1.22

600 X 1076
234

1255

0.86

446

2.81

525

1.6 X 1019
6.4 X 1018

275 Mw
105,000 kw
3

0.5

328
27,900
985
200
4.91
140

amount of =shim control for normal operation. Safety rods are included in

the reference design but do not affect neutron economy.

Fission-praduct potsons. The adjustment of breeding ratio to correspond
to the effect= outlined above changes the required chemical processing
cyele for the blanket system. This change in T'g also changes the equilibrium
88K LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24

value of fission products in the blanket. Proper adjustments result in a
blanket processing cycles of 200 days (assuming Z, = 0.25) and a fission-
product poison fraction in the blanket of 0.039 (RP in blanket = 0.15).

Neutron balance. The neutron losses proportional to one absorption
in U233 are listed below:

Absorptions in: U233 1.000
Th 0.860

C 0.025

Bi 0.050
Xeld 0.010
Sm 149 0.017
Figsion products 0.073
Higher isotopes 0.010
Control rod 0.010
Pa233 0.005
Leakage 0.170
Total 2.230

214-3. SysTEMSs DESIGN

24~-3.1 General. Systems design covers all of the reactor plant external
to the reactor, except for chemical processing. The reactor plant includes
the steam generator, but not the steam system or its auxiliaries. The
principal purpose of the systems is to transport heat from the reactor and
generate steam. They also provide supporting functions, such as shield
cooling, uranium addition, ete.

The primary system consists of six heat transport loops, each consisting
of a pump, a heat exchanger, check valve, and interconnecting piping. The
hot-leg temperature is 1050°F; the cold-leg temperature 750°F. In each
of the intermediate heat exchangers, heat 1s transferred from the bismuth
to the intermediate fluid, sodium. There are six intermediate heat transport
loops, each containing a pump, steam generator, and interconnecting
piping. The hot-leg temperature is 1010°F; the cold-leg temperature
680°F. Steam is produced at 2100 psia, 1000°F,

Selection of the above parameters was a problem involving consideration
of the steam plant as well as the reactor plant. The primary system
temperatures were first fixed by using the largest AT considered likely
to prove practical.

The temperature approach of the intermediate heat exchanger was set at
40°F, resulting in a sodium hot-leg temperature of 1010°F. To provide the
close approach necessary for steam temperature stability, the steam
temperature was set at 1000°F. A steam pressure of 2100 psig was picked
to correspond with 1000°F.
24-3] SYSTEMS DESIGN 889

Shifting the sodium cold-leg temperature redistributes heat-transfer
surface between the intermediate heat exchanger and the steam generator.
However, it seems desirable to favor making the intermediate heat ex-
changer small to cut down on fuel inventory. For this reason, the sodium
cold-leg temperature was established at 680°F,

24~-3.2 Plant arrangement. Plant arrangement starts with positioning
the primary system relative to the reactor, and this is determined by seven
principal considerations: (1) reactor design, (2) plant operation, (3) main-
tenance, (4) operational limitations of major components, (5) structural
integrity of piping, (6) economies, and (7) safety.

A preliminary analysis of the two reactor concepts, single-fluid and
two-fluid, resulted in the decision to use three external loops for the single-
fluid and six for the two-fluid reactor. For both these alternates the main-
tenance philosophy selected was that of removal and replacement by hori-
zontal transfer of a complete primary loop upon failure of any major
component in the loop [5]. Thus, for arrangement purposes, the primary
loop= assume the shape of a rail-mounted horizontal containment vessel,
or capsule, sized to contain all loop components. The height of the capsules
relative to the reactor is dictated by an economic balance between height
or elevation costs and pump net positive suction head.

The arrangement for the two-fluid reactor with six primary loops is
shown in Figs, 24-16 and 24-17.

[ plan, the primary loops were located radially around the reactor,
Fig. 24-16. A minimum length of interconnecting pipe between the reactor
and the loops was used because of high fuel inventory costs. This latter
con=ideration ruled out shielding of any appreciable thickness between the
reactor and the loops.  Maintenance access doors and other shielding
aronnd the outside of the loops was sized for source conditions 6 to 8 hr
after shutdown of the reactor to permit access by maintenance personnel
at that time into the annular area.

With the primary loop arrangement established, the next problem was
locution of the intermediate system. Since this system is the connecting
link between the primary systems and the steam turbines, it must be
locuted between them. The turbine is above ground level for gravity drain-
aue of condenser cooling water, and the primary loops are below ground
level for economy of shield costs. The path taken by the intermediate
svstemn can be either a high-level path, immediately up from the primary
svstem, or a low-level path, immediately down from the primary system,
and then horizontally to an area outside the primary system area.

The intermediate system in this arrangement follows the high-level
route to the steam plant. Sodium lines are brought straight up to an
annular area around the reactor maintenance chamber. Since access to
 

 

     

 

890 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24
! Turbo-Generator Building
Sodium Cold
Dump Component
__Tanks 25200 Sterage ;
N, - } l bhl,Jp!flq i
. s | i
. Foacs \ \.) \ | and
: -M G,Q o achine | | |Receiving]
3
e LN \ :\T"__“_'" : Pool
\ ) '/Fi ‘ ©
7 b i
r‘\m\ ! I—_ Platform
apsule: I\ L
o L. Decontamination
% ) J‘.LU___.__ Area
\T:,Zj,: - jP|utform
‘@ Blowers
@ and
/‘ Venhlclhng
Stack  system

   
 

Pyroprocessing and Aqueous

Capsule Removal
and Equipment Storage

First Floor Plan

Fra. 24-16. LMFR-6: Capsulate loop conceptual plant layout.

this chamber will not be permitted during reactor operation, a heavy shield
wall i1s not required around the chamber.

Within this annulus are the sodium pumps and the steam generators.
Final layout of this equipment will require considerable ingenuity, but it
is feasible. Steam lines will cross the roof of the reactor building to the
turbine building,

Because the primary loop hot maintenance shop for this concept serves
such specialized functions, its usefulness for maintenance of chemical
processing equipment is doubtful. Accordingly, the chemical processing
facilities for this two-fluid six-loop plant, together with its supporting hot
and conventional laboratories, fuel addition and other systems, are located
in a separate building.

The turbine building 1s of conventional construction and will be in all
essential respects wdentical for both plants.

Startup heating switch gear, gas heating and cooling systems for the
reactor and dump tanks, inert gas storage systems, control rooms, and
other auxiliaries are located relative to the above systems as logically as
possible in the light of their functional requirements.

With respect to contamination control the basic philosophy 1s (1) con-
trolled access to areas having different order of magnitude activity levels
and (2) controlled circulation of ventilating air to assure flow from low-
24-3] SYSTEMS DESIGN 891

 

 

   
 
  
 
    
  

 

 

 

 

 

Reactor
Service 5 Ton Crane
Capsule  Area
Sodium Acc:ss - Sodium
rea —k
Systems S Systems Shielding Windows
Mechanical Arm Rails
<L 4/Crune Rail
Transfer Area4— R . o £ B W =TT 0
<A e

 

Sodium Capsule
Dump

Tanks

 

“~Wasta~ Dry Hot

Tank Pit M ransfer Car Magnesium Sump Storage
Fuel Dump Room Wash  Tanks

   

Fig. 24-17. LMFR-6: Capsulate loop conceptual plant elevation.

to high-level activity areas. IFor guidance in achieving these objectives
a rough scale of activity levels has been proposed, as follows:

Class = l—conventional steam turbine plant, personnel monitored.

(lass = 2—uncontaminated areas of nuclear plant, personnel monitored.

(lass = 3—potentially contaminated areas, personnel closely monitored;
e.g., shield cooling, reactor and dump tank heating and
cooling, hot shop operating area.

(lass 24— low activity, accessible by closely monitored personnel only
under favorable conditions; e.g., exhaust blower room, hot
chemical laboratory.

Class #5—medium to high activity, accessible by closely monitored
personnel only after executing standard decontamination
procedures; e.g., hot maintenance shop.

Class Z6—high activity, no access during life of plant except after
extended shutdown and special decontamination; e.g., chem-
ical processing and chemical hot cell.

(Class =7—very high activity, no access by personnel during or after
life of the plant; e.g., primary loop and reactor areas.

 

24-3.3 Primary system. The LMFR primary system is designed to re-
move up to 825 Mw of heat from the reactor. The primary system consists
of =ix =eparate heat transport loops.

The fuel stream enters the bottom portion of the reactor vessel at a
minimum bulk temperature of 750°F, and flows upward through the core,
where tisslons within the fuel cause the fluid to undergo a temperature rise
of 300°F, resulting in a maximum fuel temperature of 1050°F. Upon
leaving the core, the fluid passes upward to a degassing area, where volatile
ti=<lon products are removed from the fuel stream. The reactor discharge
consist= of a header which splits the fuel flow into the primary heat-
transport loops.
892 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24

The primary loop piping, 20 in. in diameter, s sized to obtain & maximum
fuel velocity of 10 fps.

From the degassing area discharge, each fuel stream flows to the suction
of a variable speed centrifugal sump type pump. Each pump is de-
signed to deliver about 9000 gpm at 20-ft head of pumped fluid. To obtain
a reasonable pump speed, the net positive suction head requirement is
11.5 ft. A gas pressure (helium) is maintained over the pump sump to
prevent flooding of upper parts (motor windings, cooling system, ete.) of
the pump.

From pump discharge the fuel stream flows to the tube side of a U-tube
U—shell intermediate heat exchanger in which the fuel stream gives up heat
to the intermediate fluid, sodium. TUpon discharge from the intermediate
heat exchanger, the fuel solution flows into the reactor.

To meet safety requirements, the reactor and the major components of
the primary loops are enclosed within econtainment vessels. The contain-
ment vessel which houses the pump and heat exchanger of each primary
loop 1s a cylindrical capsule, 20 ft in diameter by 30 ft long, including 2:1
elliptical heads. The capsule is equipped with access holes such that cer-
tain maintenance operations may be performed [5]. The reactor con-
tainment vessel is a right circular eylinder 30 {t in diameter, with a hemi-
spherical top head and a flat bottom head. Access holes are provided in the
vessel for maintenance operations.

Each heat-transport loop is provided with four dump tanks which re-
ceive the loop and a portion of the reactor volumes. The tanks are sized and
arranged to prevent a fast chain reaction. The primary loops are filled from
the dump tanks by means of small electromagnetic pumps. These pumps
also promise a means for agitation of the fuel.

Two dump lines, each with a maintainable valve, connect each loop with
the dump tanks.

The bismuth charge system consists of a bismuth melt tank, filter,
valves, and piping to the dump tanks.

The proposed material of construction exposed to primary fluid is 21%
r-19% Mo steel.

24-3.4 Intermediate system. The intermediate system, which also con-
ststs of six separate heat-transfer loops, utilizes sodium as the heat-transfer
medium. All material of construction of the intermediate system, except
the steam generator, is 23% Cr-1% Mo steel. The steam generator, which
is designed for high-pressure, high-temperature service, is constructed of
type-304 stainless steel. The intermediate piping (12-in. schedule-30) is
sized for a maximum sodium velocity of 17 fps.

Sodium flowing at 11,000 gpm enters the shell side of the intermediate
heat exchanger (which is a U-tube, U—shell unit containing 2400 5/8-in.-OD
24-3] SYSTEMS DESIGN 893

tubes with an average length of 21 ft) at 680°F, flows countercurrent to the
fuel stream, and exits from the heat exchanger. Sodium flows to the suction
of a variable speed centrifugal sump type pump. Each intermediate pump is
designed to deliver 11,000 gpm at 180-ft head.

From pump discharge sodium flows to the shell side of the steam gen-
erator. The steam generator 1s a U-tube, U-shell "once-through” type
unit which 13 constructed of type—304 stainless steel. The unit consists of
030 1,2-1n.-0D tubes with an average length of 65 ft. The shell OD is
29 in. and the over-all length is 68 ft.

Sodium flows countercurrent to superheated steam, boiling water, and
feedwater 1n the steam generator and gives up heat which produces
1,100,000 b /hr of superheated steam at 2250 psig and 1000°F.

From the steam generator sodium flows to the intermediate heat ex-
changer inlet to complete the cycle.

[n addition to the components listed above, auxiliary components are
necessary to obtain proper function of the intermediate system. An ex-
pansion tank 13 located at the highest point of each intermediate loop. This
tank =erves as a cushion for pressure surges, a surge vessel for thermal ex-
pansion of sodium, and suction head for the pumps. The lowest point of
each intermediate loop is connected by pipe and dump valves to a sodium
dump tank which receives the inventory of the respective loop. Sodium is
repliced in the loop by a small electromagnetic pump which takes suction
from the bottom of the dump tank. A plugging indicator and a cold trap
are provided to determine sodium oxide concentration and to maintain
the oxide concentration at low levels.

[n the event fission-product “hangup” occurs in the intermediate heat
exchanger, fission product activity will generate heat within the metal.
To prevent excessive metal temperatures, cooling must be provided when
the unit ix drained. This cooling is accomplished by providing removable
section= of insulation which, when removed, will permit heat to be dissi-
pated by radiation, conduction, and convection heat transfer. Flow control
of the intermediate system will be by the variable speed pump drives. This
method of control should provide a reasonably constant steam temperature
and pressure.

24-3.5 Reactor heating and cooling system. The reactor must be pre-
heated prior to operation and for outgassing purposes. The required tem-
perature for outgassing the graphite is 1000°F. To achieve preheating,
hot helium gas will be circulated through the close-fitting jacket or double
containment which creates an annulus surrounding the reactor vessel.
During the preheating phase, helium gas will be pumped from one of two
blowers, pass through an electric resistance heater, be introduced at the
bottom of the annulus, pass up around the reactor vessel giving up its
894 LIQUID METAL FUEL REACTOR DESIGN STUDY [crap. 24

transported heat to the cooler surface, and return by ducting from the
upper end of the containment to the blower suction.

When, for any reason, it becomes desirable to shut down the reactor and
dump the primary system, reactor cooling must he provided to remove
decay heat generated by fission-product hangup within the reactor after
dump. This is necessary to prevent internal temperatures from exceeding
design limits. The system as just described provides cooling by opening a
valve to bring a finned tube helium-to-water heat exchanger into the cyele
and by elosing a stop valve to remove the gas heater from the gas flow
path.

Helium system design pressure and temperature will be 5 psig and 1050°F.
The entire loop is of all-welded construction to minimize helium leakage
and leakage of volatile fission products should a rupture of the reactor
vessel or piping give volatile fission products access to this loop.

24-3.6 Dump tank heating and cooling. When fuel ix drained from the
primary system into the dump tanks, fission-product decay produces heat
within the fuel which must be removed to prevent dump tank metal
temperatures from exceeding design limits,

Cooling is accomplished by circulating helium at 140 psig through a nar-
row annulus around cach dump tank. Helium which has removed heat from
the dump tanks passes through a finned-tube heat exchanger and gives up
heat to river water. Circulation of helium is provided by six 14,000 ¢fm
blowers, each rated to provide a head of 18in. water. Three standby
blowers are also provided. Helium piping is arranged such that four dump
tanks are serviced by one hlower.

To preheat the dump tunks and to maintain their temperature at a level
such that fuel precipitation does not occur, electric heaters are paralleled
with the heat exchanger such that the same piping svstem serves for heating
or cooling. The heaters or heat exchangers muy be brought into or taken
off the evele by valving,

24-3.7 Startup heating system. Prior to power operation, the LMFEFR
heat-transport system must be preheated to about 800°F. The reqactor and
the prinuiry dump tanks are preheated by electrie farnaces and cireulating
hetium.  The remainder of the heat-transport systems, i.e., primury pipe,
mtermediate heat exchanger, intermediate piping, dump tanks, expansion
tanks, steam generator, and the steam svstem pipe and components, are
preheated by induction heaters.

Since 25 Croloy and stainless steel are nonmagnetic, a thin sheet of car-
bon steel will be required under arveas where induction heaters are appled.
24--3] SYSTEMS DESIGN 895

24-3.8 Primary inert gas system. Inert gasis used in the LMFR primary
svstem to cover all free hiquid metal surfaces and to provide a gas seal
within the pumps.

Helium, by virtue of 1ts very low activation cross section and inertness,
i= utilized as the cover and seal gas for the primary system, It 1g stored
at 200 psig in a storage tank and 18 piped via pressure-reducing valves to
the pump, dump tanks, and reactor. Since relatively small quantities of
heliuni witl be used, 1t 18 expected that waste helium will be discharged via
the off-gas system to the stack.

Sinee commercial helium 1y sufficiently pure for use in an LMIEFR, no
purification will be required.

24-3.9 Intermediate inert gas system. Nitrogen 1s used in the LMFR
intermediate system to cover all free sodium surfaces and to provide a gas
sealin the pump. It 1g stored at 200 psig in a storage tank and 1s piped via
pressure-reducing valves to the pumps, expansion tanks, and the dump
tanks. Used nitrogen 1s discharged to the stack.

Commercial nitrogen must be purified prior to use in the intermediate
svetem. Purification is accomplished by bubbling nitrogen through several
tanks contaming Nalx.

24-3.10 Shield cooling. "The concrete surrounding the primary cells
serves as a shield from the neutrons and gammas leaving the primary fluid.
In the absorption of these neutrons and gammas, considerable heat 1s
generated within the conerete. To hold temperatures and thermal gradients
within the conerete to reasonable limits, a cooling system must be utilized.
This cooling system consists of panel coils embedded about 6 m, within the
concrete <hield. High-purity water, flowing inside the panel coils, removes
heat from the conerete and prevents temperature damage to the concrete.

The closed, high-purity loop which rejects heat to river water iz designed
for o maximum heat load of 6 Mw. One pump of 900-gpm capuacity pro-
vides cireulation for the closed water loop. Flow control valves proportion
the flow to the various panels such that panel coil outlet temperatures are
el

A dump tank for the closed loop (about 300 ft?) is located beneath the
panel coils, o that the coils may be gravity draimed. Water 1s returned to
the closed loop by means of gas pressure. In the event it is necessary to
dispose of the water in the closed loop, it may be drained from the dump
tunk to the rmdioactive waste disposal system.

24-3.11 Reactor cell cooling. Instruments located within the reactor
cont:inment vessel must be kept at a relatively low ambient temperature.
1o medntain the ambient temperature, a “fin fan’ cooling unit is attached
896 LIQUID METAL FUEL REACTOR DESIGN STUDY [cuaPp. 24

to the containment vessel. Helium, which fills the containment vessel, is
circulated by a blower located within the cooling unit. The circulating
helium removes heat from the containment vessel and transports it to the
finned coil, where it is transferred to water which is taken from and returned
to the closed shield cooling circuit.

24-3.12 Capsule and reactor room cooling. The containment capsules
and the reactor are located in a large room. The ventilation requirements
of this area are dependent upon heat losses from the primary loop contain-
ment capsules.

Ventilation is provided by locating air intake louvers at several points
around the room. An air fan provides circulation of air around the capsules
and removes heat, which is discharged to the stack. A radiation monitor
continuously monitors the air discharge. In the event radiation tolerance
levels are exceeded by the air discharge, the cooling air will be recirculated
to the reactor and capsule room until the source of rudiation is determined.

24-3.13 Raw water system. The raw water system 1s the final waste
heat sink for the entire plant. River water, which 1s screened and treated,
is piped beneath the turbine-generator building. The systems which require
river water, ie., turbine condenser, shield cooling, reactor cooling, dump
tank and pump cooling, take suction from this pipe and discharge to a
similar one which returns the heated water to the river. Where possible,
river water flows tube side in heat exchangers, to facilitate cleaning.

24~3.14 Instrumentation and control. The purpose of the control system
in this plant 1s to provide safe and stable operation while following the
loads imposed by the utility system. The plant follows the turbo generator.
Loads on the turbo generator are set by the utility.

A load change will appear in the steam system as o change in throttle
alve position and, therefore, a change in steam flow and pressure. The
feedwater controllers at the inlet to the steam generators will sense these
changes and operate to maintain steam pressure constant. The steam flow
could also provide an anticipatory sighal to the primary and intermediate
system pumps to change their speed to suit the load.

The reactor will have a negative temperature coefficient of reactivity.
Thus, it will try to maintain its average temperature constant during load
changes. The temperature will change from time to time as reactivity
changes. To take advantage of the negative temperature coetflicient, the
average temperature of the reactor will be set at a constant value.

Programming of flow rate in the primary and intermediate loop is un-
certain. Cost estimates for pumps and control equipment were hased on
the premise that speed of the pumps would be varied. This might be neces-
sary to avoid thermal stresses during transients.
24-3] SYSTEMS DESIGN 897

 
   
 

Blanket

Tquid Pa Decay Tanks

 

 

Core . rq—i
Liquid "1 Hydro

‘n’ Salt
Fluor'n'r

 

 

 

 

 

 

To NFPN ¥

Salt
Process =

 

 

 

 

 

   
  

Waste
From NFPN E
Process 2
Flyor'n'r
Blanket Yo UF
- &
Liquid Return H2 o Reduct'n
Mg, Zr, U

 

 

 

 

Salt 4 UF 4

> Electr'y'c
Core Reduct'n
Liquid Return

 

 

 

 

 

 

F1a. 24-18. Fluoride volatility processing of core and blanket.

24-3.15 Maintenance. The maintenance of the reactor and primary
system components will be completely remote, because of the high levels of
radioactivity of the circulating fuel. The entire plant and reactor system
are arranged for remote maintenance [5}].

24-3.16 Chemical processing. The pyro process chosen for this economie
study 1s the fluoride volatility method applied to a two-region reactor.
Work of adapting this process to bismuth fuel processing is presently under
way at Argonne National Laboratory. Figure 24-18 presents the main
steps 1n this process. As shown, the process may be used for either blanket
or core liquid. When the plant i1s processing core liquid the basic steps in
this process are (1) hydrofluorination to oxidize uranium and some fission
products, (2) transfer of the oxidized material to a fused salt phase, (3)
Auorination of the salt carrying the uranium and fission products for sepa-
ration of uranium as volatile UFg, (4) reduction of the Uly to UF4 by
H. in o fused salt phase, and (5) reduction of UF4 to uranium metal and
transfer into the metal phase (bismuth).

The volatility method can be conveniently used to process a thorium
bizmuthide blanket. The process must be preceded by a phase separation
~tep which separates the thorium bismuthide solids from the liquid carrier
hi=muth (Fig. 24-19). The modification of the core liquid process flowsheet
v ax follows: (1) salt effluent from the hydrofluorination step must be
stored in order to achieve Pa decay to uranium, and (2) the bismuth liquad
15 returned to the blanket head end process without the addition of uranium.
898 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Stripped Bi
l From Volatility Plant
Th
Slurry
Mixer
Feed Sluery
1 Recycle
L___>. Heat
Pulser
Crystallizer
b Phcse
Separator
Slurry Return
Y -
To Blanket
U Rich
Bi Liquid

F1g. 24-19. Head end processing, bismuthide slurry.

Certain of the fission products are not removed by volatility processing.
These may be removed by zine precipitation (Fig. 24-20). This process
requires that the bismuth feed be free of uranium, and the volatility plant
provides such a bismuth feed stream.

The head end process transfers bred uranium, protactinium, and fission
products out of the solid phase portion of the slurry and into the liquid
phase. After this step the two phases are partially separated. A liquid
portion transferred to the volatility plant carries bred uranium, protac-
tinium, and fission products with it for stripping with HF. The stripped
liquid bismuth is returned to the head end plant for mixing with fresh slurry
feed. The head end process is not 1009% efficient; i.e., the uranium and
protactinium are not completely removed from the slurry before reconsti-
tution and return to the blanket region. This problem has been examined
in some detail and was taken into account in determining economics.

24-3.17 Turbine generator plant. A flow of 3,330,000 1b/hr of super-
heated steam at 2100 psi and 1000°F is delivered to the turbine. The
generator has a gross output of 333,000 electrical kw, and the condenser
removes 1.677 X 10° Btu/hr at 1.5 in. of mercury absolute, thus giving a
gross heat rate of 8450 Btu/kwh. About 18,000 kw of electrical power is
used for the various pumps and auxiliary systems in the plant, making
the net output 315,000 kw. Therefore, the net heat .ate is 8940 Btu/kwh,
which corresponds to an efliciency of 38.2%.
24-3] SYSTEMS DESIGN 899

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

From Zn Rich, NFPN
Volatility ‘ TPlus Bi
Plant
Zinc
Concen— .
_Bi Rich | trator Zinc
Plus Zn Waste
Zn
In
Zinc Zn
Crystal -
lizer
| Still
Bi Return To

Volatility Plant

Fia. 24-20., NFPN fission-product removal.

At full load there are 1,825,500 Ib/hr of steam leaving the turbine and
being condensed in the condenser. Also, 113,800 Ib/hr of water from
various parts of the cyele are being cooled by the condenser. The total
head load on the eondenser is 1.677 X 10 Btu/hr. The condenser cooling
water enters one water box at 70°F and leaves the other at 80°F.

24-3.18 Off-gas system. The actual design and efficiency of any con-
ceptual degasser are as yet unknown quantities, and a knowledge of these
important details will have to wait until in-pile loops have provided suffi-
cient data.

The off-gas system will consist of a cooler followed by a series of storage
hottles. Gaseous fission products that have been separated from the Liquid
bizmuth in the degasser are first sent through a cooler which offers a resi-
denee time of about a day, or enough for most of the short-lived isotopes
to decay. From the cooler, the gasses are compressed into storage bottles,
each capable of holding 30 days’ accumulation. The storage bottles will
cach be .25 {t3 in volume, and at 212°F and 60 psia at the time of dis-
cotnection from the compressor.

~ome sweep gas may be included in the above gas stream, but the present
dexign philosophy indicates that no extra sweep gas should be required;
however, if some sweep gas 1s required for efficient degasser operation, this
gas could be obtained by a recycle of previously removed gas. This recycle
~weep stream would most probably be taken from the storage bottles after
~ufficient cooling.
900 LIQUID METAL FUEL REACTOR DESIGN STUDY [cuap. 24

The gas in the storage bottles may be vented to the atmosphere after
90 days of storage, since then only the Kr35 activity is still present in ap-
preciable amounts, and this can be released provided there is sufficient di-
lution. However, the most probable course of action will be to process the
stored off-gas through a gas separation system, where the valuable Kr8?
will be recovered.

24-4. SineLE-Fruip ReacTor DESIGN

24—4.1 General description. The single-fluid LMFR concept has been
investigated to determine the characteristies and economie attractiveness
of this design. In general, the core consists of a large array of solid modera-
tor blocks stacked to provide the desirable geometry of a cylinder. Vertical
cylindrical channels are drilled through the moderator to allow circulation
of the liquid metal slurry containing both the fuel and fertile material.
The fission heat generated in the fuel-coolant stream is transported by
forced convection to heat exchangers external to the reactor vessel. The
unique feature of this concept is that only one coolant, the slurry, is used
for removing heat from all parts of the reactor. The desired slurry-to-
moderator ratio is achieved by selecting the appropriate combination of
channel size and spacing.

24—4.2 General specifications. The general specifications for the system
affecting reactor design are tabulated below:
Power 825 Mw (thermal)

315 Mw (electrical)
Slurry temperature:

T; 750°F

Tout 1050°F
Maximum slurry velocity 10 fps
Fuel J235 op {7233
Fertile material Thorium
Moderator material (iraphite or BeO
Slurry carrier Bismuth or lead
Slurry-to-moderator ratio Variable
Fertile material content in slurry Variable
Core geometry Cylinder

Core size Variable
24-4] SINGLE-FLUID REACTOR DESIGN 901

24-4.3 Parametric study. A parametric study was performed to deter-
mine the optimum nuclear parameters for a single-fluid concept. The
variable parameters investigated and their range of values are:

Slurry-to-graphite ratio, V,/V. = 0.05 to 1.0,
Fertile material content, g/kg of Bi= 0 to 80,

Equivalent bare reactor diameter, D, ft = 10 to 20.

The choice of fuel for the first full-scale LMFR will depend upon the
availability of U238, which is much more attractive than U2*% because of
hetter neutron economy, and a sufficient quantity for fueling an LMFR
may be available in 10 to 15 yr. In the carly stages of this study, how-
over, U2% was arbitrarily chosen as the fuel for the parametric study.
The =election of the reference design should be valid for either fuel.

In each case the critical concentration and conversion ratio were de-
termined by multigroup diffusion theory, using 37 neutron energy groups.
To handle the large number of caleulations, a digital computer was used
once the range of values for the parameters was established by a series of
criticality calculations by hand.

The use of BeO as a moderator has the advantage of reducing the core
<ize because of improved slowing-down power compared with graphite.
Critical size, fuel concentration, and breeding ratio were determined for
one case, using BeO as moderator.

Rince the cost of bismuth as a primary coolant is between $:3,000,000
and $4,000,000, the inventory charges are a significant fruction of the
total fuel costs. One case was caleulated using lead as a coolant in order to
compare the inerease in inventory charges due to the use of bismuth with
the loss in conversion ratio due to the absorptions in the lead.

Basis of nuclear caleulations. To obtain comparative results, the follow-
ing specifieations were assumed for all cases:

Average temperature 862°1°

(Giraphite density 1.80 g/cc
Bismuth density 9.83 g/ce
Geometry Cylinder (H = D)

For consistency and ease of comparison, all caleulations used equivalent
hare reactor dimensions, except the caleulation of reflector savings as a
function of reflector thickness.
TasLe 21-6

SUMMARY OF SINGLE-FLUID NUCLEAR CALCULATIONS

 

 

11144
11154
11164
11234
11232
11244
11254
11324
11325
11326
11334
11335
11336
11342
11344
11345
11346
11424 i
11425
11426
11434
11435
11436
11444
11445
11446
11514
11524
11525
11526
11534
11535
11536
11544
11545
11546
11435*
11431% |
114331 E

11344% |
i

 

 

] Bare equiv.

 

 

 

 

30 ; 0.7

 

 

 

 

 

 

Thorium, Initial c

g/kg Bi, Vo/V, core size, conversion Nazs/ ‘J\‘GBi’
Wos ‘ D=H, it ratio x10
0o | 05 l 14 0.00 | 153
0 i 0.7 \ 14 0.00 1 134
0 1o 14 0.00 | 120
15 0.3 14 0.53 : 458
5 | 03 10 0.43 \ 621
15 b 0.5 14 0. 608 i 451
15 0.7 14 0. 65 481
30 0.2 14 0 625 774
30 0.2 17 0666 | 706
30 0.2 20 0.695 | 666
30 0.3 14 0.692 772
30 0.3 17 0733 708
30 0.3 20 0.760 671
30 0.5 10 0.63 1181
30 0.5 14 0.746 870
30 0.5 17 0.788 704
30 0.5 20 0.814 751
50 0.2 14 0.735 1199
50 0.2 17 0780 1099
50 0.2 20 0.801 1054
50 0.3 14 0.788 1304
50 1 0.3 17 0.817 1203
50 0.3 20 0.843 1144
50 0.5 14 0.787 1757
50 0.5 17 0.827 1508
50 0.5 20 0.854 1504
80 0.05 | 14 0.565 2099
s o2 0w 0,804 1990
80 02 17 0.840 | 1833
g | 02 20 | o865 1 1769
80 0.3 14 L0816 | 2452
80 0.3 17 0849 2274
80 | 0.3 20 0.875 | 2160
K0 i 0.5 14 0.747 | 4471
80 0.5 17 0,788 3939
80 0.5 20 0.851 3684
50 0.3 17 0.720 \ 1531
50 0.3 8 0.680 | 1223
50 0.3 12 0765 1032
5 14 0880 a 678

 

*T.ead coolant

TBeO moderator

10283 fhel

 
24-4] SINGLE-FLUID REACTOR DESIGN 903

The resonance integral of the fertile material is a function of the scatter-
ing per atom, size of fuel channel, and lattice spacing. The channel size
and lattice spacing, however, are not specified; therefore, the lattice
resonance parameters are not known. A maximum value of the effective
resonance integral is the homogencous value based on the scattering in
the core mixture per atom of fertile material. A minimum value of the
resonance integral 18 the homogeneous value based on scattering in the
slurry per fertile atom. For the cases using thorium, a value of Ry2 (ef-
feetive resonance integral) was chosen between the maximum and mini-
mum values, and the calculated uncertainties are 4 2007 in the Ng5/Np;
ratio and 4 3.39; in the conversion ratio.

Resulls of nuclear calculations. The results of the parametric study are
summarized in Table 24-6 for all cases. The eritical concentrations and
conversion ratios for the cases using thorium as the fertile material are
araphed in Figs. 24-21 through 24-25.

The notation used on all graphs have the following definitions:

N5 Np = atom ratio of U235  to bismuth.

Woso = thorlum concentration in grams of Th?32/gBi.
'e V. = volume ratio of slurry to graphite in core.
D = the equivalent bare core diameter in feet.

In all cases, the fuel concentration inereases with an increase in fertile
muaterial, Woe (Fig. 24-24). An increase in V,/V, increases the thorium
content, reduces the slowing-down power, increases the average encrgy of
the neutron spectrum in the core, and increases the thorium absorptions.
A= u result of these effects, the critical fuel concentration in the fluid fuel,
Nos Np; ratio, increases as V,/V, increases (INg. 24-25),

 

 

 

 

 

0.9
! _p=20"
—D=20
—D=17
~~D=14
—D=20'
08 ~D=14'
o [~D=17
z ’/DZM
o
T
% o7 o
o
z D=4/
S — Vs/Ve=0.2
/ . Vs/Vc=03
' —— == Vs/Ve==0.5
26— / S W /Ve=07 7
s
.’/'
05 ! | | | i l
20 30 40 50 60 70 80

W . Grams Thorivm/kg 8i

T'16. 24-21. Conversion ratio vs. thorium concentration for a single-fluid LMFR.
904

0.9

o
©

—---— Wp2=15

— W(p2=30

—-— Wqp=50

------- Wny=80

’f/,,

0.6 t— /—”///D" 14’ —
/
o | | | 1
0.2 03 04 0.5 0.6
Vs/Vc

Conversion Ratio
o
N

LIQUID METAL FUEL REACTOR DESIGN STUDY

 

 

   

 

 

0.7

[cHAP. 24

Fra. 24-22. Conversion ratio vs. slurry-to-graphite volume ratio for a single-

fluid LMFR.

0.90

Conversion Ratio

e
N
o

0.65

0.60

 

 

Wgy =180
Wgp =80
Wp9o =50
Wop =50

4 Woy=80

Wpp=230

W= 50

   

Vs/Ve=0.3
————Vs/Vc=0.5
— - Vs/V¢=0.2

 

02=30

Wno=30

 

| I l |

 

16 17 18
Bare Diameter, Feet

19

20

Fic. 24-23. Conversion ratio vs. diameter for a bare single-fluid LMFR.
24-4] SINGLE-FLUID REACTOL DESIGN 905

 

 

 

 

 

 

4400 T I T ] D=14
_____ =0.2 .
 sonl v
x S 0. D =20’
& 3600+— Vs/Ve=0.5
z O Vs/Ve=07
h 3200(— —
z\
2 2800} —
= D=14
£ 2400 ~dp=17
2 7 AD=20
< 2000(— ~p=14
£ Z-1p=17
E 1600 —D=20
2
T 800|— —
.
400
10 20 30 40 50 60 70 80

Wy Grams Thorium/kg Bismuth

Fic. 24-24. Critical fuel concentration vs. thorium concentration for single-fluid
LMFR.

5000

 

Bi ¥ 106

 

1000

|
|
\.
\

[

Fuel to Bismuth Atom Rotio, N25/N

[

 

 

100 i . | . ] | | ] ! ; l ] J
0.1 0.2 0.3 0.4 0.5 0.6 0.7

Vs/Vc

Fic. 24-25, Critical fuel concentration vs. slurry-to-graphite volume ratio for a
single-fluid LMFR.

 
906 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHap. 24

The conversion ratio 18 highly dependent upon the Ng»/Noj rutio, the
average energy of the neutron spectrum, and the reactor size. igure 24-21
shows that for larger values of 17,/V,, the conversion ratio passes through a
maximum as thorium concentration inereases; however, for smaller values
of V,/ V., the conversion ratio increases continuously as V,/V. increases
over the range of interest; ie,, the maximum value of the conversion
ratio shifts to higher values of W2 as the V,/V, ratio decreases. Likewise,
the curves of conversion ratio versus V,./1, go through a maximum, with
the maximum value occurring at mereasingly higher values of V,/V, as
Woe increases (INg, 24-22),

An inerease in core diameter simply reduces the neutron leakage. As a
result, the conversion ratio mereases as the diameter mereases.  An in-
erease in D from 14 to 20 ft increases the CR approximately 0.09 (Fig.
24-23).

Case 11435 wus recaleulated using lead mstead of bismuth as the coolant
fluid. The conversion ratio decreased by 0.10, and the eritical Vo5 /Ny
ratio increazed from 1203 X 107% to 1531 X 1079,

Beryllium oxide, BeO, was used as moderator in another variation of
Case 11435, This caleulation, case 11433, for a diameter of 12 ft, requires
an No;/ Ny ratio of 1032 X 107Y and yields the slightly lower conversion
ratio of 0.77.

The worth of a pure graphite reflector was caleulated for Case 11435,
The reflector savings as a funection of reflector thickness are shown in
Iig. 24-26. The reflector savings are approximately equal to the reflector
thicknesses for reflectors less than 2 tt thick.

The values of conversion ratio and Nos/Np; ratio caleulated in this
parametric study are for hot, clean reactor conditions, and they are used
for comparative purposes only. The eftfects of fission-product poisons,
control rods, and Pa??? losses have not been included.

24-4.4 Economic optimization. The selection of parameters for a refer-
ence design must be based upon economics. An economic optimization
was accomplished by computing relative energy costs based on those
variable costs which depend upon the parameters selected. The costs
which are dependent upon the nuclear parameters are (1) bismuth in-
ventory, (2) fuel inventory, (3) fuel burnup, (4) thorium inventory,
(5) thorium burnup, (6) reactor core and vessel, and (7) chemical process-
ihg costs.

Reactor cost. Since the range of reactor sizes varies from 10 to 20 ft,
reactor cost is an important variable. Reactor vessel, graphite, and erec-
tion costs have been estimated for several sizes; to these 13 added $167,000
for three control rods and miscellaneous hardware. Contingency and en-
gineering of 4497, were also assumed. A breakdown of these costs 13 listed
in Table 24-7.
24-4] SINGLE-FLUID REACTOR DESIGN 907

 

 

 

 

80
| | l ! ! | |
£
£ 60 —
3
v
B
£ w0 .
>
A
8
S 20— —
Q
o
| i | | | | | I
0 10 20 30 40 50 60 70 80 90

Reflector Thickness, cm

Fig. 24-26. Reflector savings vs. reflector thickness for a single-fluid LMFR.
These data are obtained from case 11435, where V,V,.==0.3, Wz =50 g/kg, and
Dp=17.

TasLe 24-7

EstiMaTED SiNGLE-FLuip Reacror Cost

 

i : : Total
201 Reactor Graphite | Misc. | Irection| Total c(())si $/yr

 

 

10 | 160,000 | 350,000 | 167,000 | 24,000 | 701,000 | 1,000,440 | 151,400
14 | 380,000 | 970,000 | 167,000 | 30,000 | 1,547,000 | 2,227,680 | 334,152
171 570,000 | 1,700,000 | 167,000 | 35,000 | 2,472,000 | 3,559,680 | 533,052
20| 900,000 | 2,800,000 | 167,000 | 40,000 | 3,907,000 | 5,627,080 | 843,912

 

 

 

 

 

 

 

 

 

Bismulh inventory charges. The bismuth inventory is determined by the
primary system volume external to the reactor vessel, the volume of bis-
muth in the core, the volume of bismuth external to the core but inside the
reactor vessel, and the holdup external to the reactor system. The primary
svstem external to the reactor vessel is made up of three heat-exchanger
loops containing a total volume of 1640 ft3. The volume of bismuth in
the core 1s
V/Ve

, ey wh 7= ol
VT, where § core volume

Va=V
The volume of bismuth external to the core and inside the reactor vessel
i~ tiuhulated in Table 24-8.

No additional holdup is included to account for temperature expansion
during startup, fuel feed system, and other sources of bismuth inventory.
The assumption used throughout this study that the volume of bismuth is
cqual to the volume of slurry accounts for an additional 3 to 10% excess
bismuth due to the ThO2 content of the slurry.
908 LIQUID METAL FUEL REACTOR DESIGN STUDY fcaap. 24

TaBLE 24-8

BismurH INVvENTORY IN REACTOR VESSEL
ExTERNAL TO CORE

 

 

 

Core diameter, ft Bi inventory, ft3
10 550
14 600
17 650
20 700

 

 

 

The density of bismuth is taken as 9.83 g/ce, and the price is assumed
to be $2.25/lb. Bismuth is a nondepreciating capital investment with a
12% annual amortization rate. The annual bismuth inventory charges
may be represented by the equation

. -~ V/Ve
Cl(fb/yr) = 012(220) V., 'H__V/I/T + Vp PRBi,
where
V» = total primary system volume except core, ft3,

pri = density of bismuth, Ib/ft3.

Fuel inventory charges. The annual lease charges on the UZ23% are as-
sumed to be 4%. Treating Pa23? as fuel, the annual fuel inventory charges
can be expressed as

C2($/yr) = 0.04VosM o5+ VasMaz + VisMys,

where
Vaz = Viz = value of U??? as fuel,

Vo5 = value of U%3% as fuel, $17,760/kg,

M ; = average mass of element j in entire reactor system
during life of plant.

To simplify the work in the absence of information concerning average
values of fuel mass, the total mass of fuel was considered to be the hot, clean
critical loading at startup. The value of M2s is taken as the initial value
with M3 and M3 taken as zero.
24-4] SINGLE-FLUID REACTOR DESIGN 909

Fuel burnup costs. Using U235 as fuel, the yearly burnup costs are

C3($/yr) = 17.76(202)P3(1 — CR),

where
P = power, 825 My,

B = grams of fuel burned per MwD), 1.25,

CR = average conversion ratio.

The initial value of the conversion ratio is used, since only relative costs
are needed.

Thortwm burnup costs. Thorium is periodically replenished in the reactor
to maintain the desired concentration in the slurry. The thorium burnup
costz may be expressed as

C4($/yr) = V2 PBCR(292),

where Vo2 = value of thorium, $12/kg. These costs are very small, ap-
proximately $10,000/yr, and are neglected.

(‘hemical processing costs. The chemical processing is assumed to use
solvent extraction aqueous chemistry in a central processing plant. The
irradiated fuel is removed from the reactor on a batch processing cycle.
The processing costs are represented by

Cp = 292 [95.875 “5“— 14705 MQ;}ET) 4 w 4 596] :

 

 

where

My2 = total thorium inventory kg,
M5(T) = Moz + Mz at time T after loading of fuel charge, kg,

T = chemical processing cycle time, days.

Results of economic optimization. Since chemical processing costs are
very sensitive to the chemical processing cycle time and the optimum cycle
time may vary with reactor design, the relative energy cost of each reactor
design was determined neglecting the chemical processing costs. The results
of this study are tabulated in Table 24-9 and are shown graphically in
Figs, 24-27 and 21-28.

The pure burner, Wo2 = (0, shows costs more than twice as high as several
of the more attractive concepts (Fig. 24-28). In general, the minimum
 

 

 

  

 

 

 

 

 

910 LIQUID METAL FUEL REACTOR DESIGN STUDY [cHAP. 24
28 ' ‘ ! T ' Vs/Ve=0.5
5 ¢—=\U.
26F T D=20
- ———D=17 /
24, ——— b=l4 Vs/Ve=0.5
\ e o VsIVe=0.5
2.2 p
@ :
viof 7
2 \ Vs/Ve=0.3
18| :
- \\ Vs/Ve=0.2
et NN ; Vs/Ve=03
' h : -Vs/Ve=0.3
5 ST~ <\\_,, T - -'_/_:// Vs/Ve=0.2
1.4 T ;;:"_2*}/‘5’—1?* o Ys/Ve=0.2
12 F
1.0 1 ] 1 . | ! | |
0 10 20 30 40 50 60 70 80 90
Wg2., Grams Thorium/KG Bismuth
Fic. 24-27. Relative cost vs. thorium concentration for a single-fluid LMFR.
3.0 T | T r T I T ! T 1’ T [ l !
55 |- _—— — — Wp2=0|
26~ -—- D=20 / ]
D =17 s
24 - D =14 /W2 =80 —
% D=1 s Wg2 =80
822 ’ RO / ]
o . /«/./
@ // S
> 20 +— . s ]
3 \ e ’
18 e /"4/ W 0 W =30 ™
\ o =157 02 =38~ "027
16 - el /\‘xoz“ 0]
\ = - ,,,:/,\_4';'_"'7 Ry 02 =50
14 TR _ <<W02 30
et i‘:'if,i:i——i_ - Wp2=30 Wy =15]
1.2 |- T Wo2=30 —]
10 | | | | : i | J‘ ; | L | | | .
0 10 20 .30 40 .50 .60 .70 .80
Vs/Vc

Fig. 24-28. Relative cost vs. slurry-to-graphite volume
LMFR.

ratio for a single-fluid

costs are achieved with thorium loadings corresponding to Woe =20 to

50 g/kg.
conversion ratio.

The most attractive designs do not have the highest values of

In many cases the additional fuel inventory charges and reactor vessel
costs corresponding to higher conversion ratios more than offset the redue-

tion in fuel burnup costs.

The cconomically optimum reactor is neither a

hurner nor a converter with maximum conversion ratio, but somewhat

between these extremes.

Using a cost of 18¢/Ib for lead as a coolant, comparisons of lead versus

Bi as a coolant were made for Case 11435.

The annual fixed charges on

lead were only $44,000 compared with $478,000 for bismuth in this case;
24-4] SINGLE-FLUID REACTOR DESIGN 911

however, the increased fuel burnup and inventory charges associated with
the lead coolant resulted 1 a net inerease of $288,000/yr or 0.14 mills/kwh
in the fuel cost. BeO) is not feasible as a moderator material for this concept
because of 1ts high cost.  Fixed charges on the BeO alone add almost a
mill kwh to the fuel cost.

The six most attractive caxes were selected and the chemical processing
costs computed for several processing eyele times. The total eosts tabulated
in Table 24-10 are based on a 3000-day eyele time, and other costs are from
Table 24-9. Since the aqueous processing costs are dependent upon the
total thortum inventory to be processed, chemical processing costs penalize
the designs with heavy thortum loadings. In Tables 24-9 and 24-10, the
total costs are reduced to mills/kwh by using an eleetrieal power output
of 315 Mw with an 80%: plant factor.

24-4.5 Selection of a reference design. Using the data presented in
Tuble 24210, a design was selected for further study. Tt s important to
re;uize that when chemieal proeessing costs are included 1 the comparison
ol energy costs, there is little difference in the cheapest four or five cases.
The relative attractiveness of these cases depends very heavily on the
ceonomie ground rules. Fven o change in chemical processing evele may
chinge the relative order of the cases. With the wide range of freedom for
cholee of nuelear parameters n this concept, the economic optimum can
be chosen to correspond to any set of basie assumptions on economics.
For example, an inerease in fuel price would emphasize higher conversion
ratins. The design selected for further study was Case 11344

Time study. The nuelear performance of the reference design, Case 11344,
wis determined usimg a thorium hfetime program written for the digital
cotnputer. These caleulations provided mformation concerning the varia-
o~ of fission-product poisong, breeding ratio, and critical fuel mass as
functions of reactor operating time. Fhis Information then made possible
the cholee of an optimum fuel processing cyele and the determination of
over-ull fuel cost for the operating reactor,

Busis of time study. The reference design caleulations used U2 as fuel
tor both the initial charge and feed muaterial. Sinee the contemplated con-
~trnetion date for an LMER is 10 yr in the future, the assumption that
(5 el will be available seems reasonable, and data based on U223 allows
coteparison with previous work [2].

The reference design on which the time studies were based has a graphite
il reflector 15 ft thick, an active core diameter of 11 ft, and a core
betohit of 14 £t The average core temperature is 900°LF. The nuclear con-
~tunt= u=ed n the two-group eriticality and isotope buildup caleulations
were determined by using a 40-group spectral code.
Rerative Exercy Costs For SiNGLE Fruip LMFR

TABLE 24-9

 

 

 

 

 

 

.. Bismuth ) . Thorium | Reactor
Initial , inven- Fuel in- Fuel inven- | core and
Clage | CORVEr- ‘\“'25/1\iBi, Wos Vi/V. D(bare), tory ventory, burnup,r fory vessel Crx 1073, Cr,
; IS;E:E) X 1086 s/ Ve it ) % 1"0’,3, (725/1'()‘3, C3 glvo_d’ 0 x 10{_3, Cox 10 13, $/vr mills/kwh
s/ye | T PR s s

11144 0 153 0 0.5 14 444 91 5348 0 334 6217 2.82
11154 0 134 0 0.7 14 461 83 5348 0 334 6226 2. 82
11164 0 120 010 14 480 77 5348 0 334 6240 2.83
11232 | 0.432 621 151 0.3 10 377 310 3039 60 151 3937 1.78
11234 | 0.530 458 15 1 0.3 14 421 256 2516 67 334 3594 1.63
11244} 0.609 451 15 | 0.5 14 444 265 2093 71 334 3207 1.45
11254 | 0.647 481 15 1 0.7 14 461 293 1890 73 334 3052 1.38
11324 | 0.625 774 30 | 0.2 14 407 411 2007 130 334 3289 1.49
113251 0.666 706 30 | 0.2 17 451 415 1784 143 534 3327 1.51
11326 | 0.695 666 30 | 0.2 20 512 445 1632 163 844 3596 1.63
11334 | 0.692 772 30 | 0.3 14 421 424 1645 134 334 3958 1.34
11335 0.733 708 30 | 0.3 17 478 442 1427 152 H34 3033 1.37
11336 | 0.760 671 30 | 0.3 20 560 490 1284 178 844 3356 1.52
11342 | 0.628 1181 30 | 0.5 10 421 256 2516 67 151 3411 1.55
11344 0.746 870 30 | 0.5 14 444 504 1358 141 334 2780 1.26
113451 (.788 794 30 | 0.5 17 523 541 1136 166 534 2901 1.31
11346 | 0.814 751 30 | 0.5 20 478 736 977 254 844 3303 1.50
11424 | 0.735 1199 50 | 0.2 14 407 624 1417 216 334 2998 1.36
11425| 0.780 1099 50 | 0.2 17 451 633 1179 239 534 3036 1.38

 

 

 

 

 

 

 

 

 

 

 

 

 

G616

AGNIS NDISHE HOLOVHIY Tiad TVIANW d1adIll

¥3 dVHO]
 

11426
L1454
11435
11436
11444
11445
11446
11514
11524
11525
11526
11534
11535
11536
11544
11545
11546

 

o OO

L8001
kS
81T
843
JTST
827
804
.506H
804
.840
. 865
.816
850
875
747
(788
815

 

1051
1304
12023
1114
1757
1598
1504
2099
1990
1853
1769
2452
2274
2160
4471
3959
3684

 

S0
o0
ol
ol
H
H
50
80
30
80
80
80
80
80
80
80
80

 

Tl TroLloLooLiotg

TJIICODJ

'

v Ot

 

2()
11
i
20
14
17
20
14
14
17
20
14
17
20
14
17
20

 

H12
421
478
H60
444
H23
(637
381
407
451
512
421
478
a60
444
523
637

 

6O
702
736
820)
097
1068
1226
993
1007
1038
1126
1284
1352
1504
2462
2569
2912

 

1014
1187
077
839
1139
927
783
2325
1051
837
724
986
803
671
1356
1132
990

 

272
223
254
297
235
277
338
323
345
382
434
307
406
475
376
444
541

 

 

3381
2868
2979
3359
3149
3330
3828
4256
3144
3263
3640
3382
3074
4053
4972
5201
5024

 

B DD BY = b kel e e e e e e e e e

.ol
.73
.93
.42
.48
.65
.83
.62
.84
.25
.36
.68

 

 

(¥-¥2

NOISHd HOLOVIYH dINTA-ATONIS

&6
TaBLe 24-10

ReLaTivE FueL Costs IncLuping CHEMICAL PROCESSING
SinerLe Frvip LMFER

Assumed processing cycle = 3000 days

 

 

 

 

 

 

 

Chemical

Initial i} D(bare) processing Other cost Total costs, | Total costs

. T T I ’ N ! ! !

Case No. | Woz CR Va/ Ve ft Moz kg | M2, ke cost, Crx10~3/yr 1073/yr mills/kwh

Cpx1073/yr

11254 15 0.65 0.7 14 11,640 413 504 3052 3556 1.61
11334 30 0.69 0.3 14 21,268 597 681 2058 3639 1.65
11344 30 0.75 0.5 14 22 407 709 745 2780 3525 1.60
11345 30 0.79 0.5 17 26,407 762 808 2901 3709 1.68
11434 50 0.78 0.3 14 35,447 989 1000 2868 3868 1.75
11435 50 0.82 0.3 17 40,275 1036 1068 2979 4047 1.83

 

 

 

 

 

 

 

 

 

 

 

¥i6

AdN18 NOISHd HOLOVAY TANA TVIAN qiadIl

¥Z "dVHO]
24-4] SINGLE-FLUID REACTOR DESIGN 915

The neutron poisons due to fission products and higher uranium iso-
topes were calculated using the data by W. L. Robba et al. [4]. The xenon
poisoning {(absorptions in Xe!'3% to absorptions i fuel) was held at 0.01
throughout life, and Sm'#? was allowed to reach steady state. The other
fission-product poisoning corresponds to poison data labeled “'less Xe and
Sm with high cross sections except low Zr%.” Due to lack of information,
no resonance absorption by the fission products was considered. The
neutron flux averaged over the entire primary system volume was used in
all 1sotope and poison buildup computation, since this is a circulating fuel
reactor.

FFuel was added at frequent time intervals to maintain kes = 1.01
(assuming 197, rod holddown). Thorium was added to the core with the
fuel to maintain a constant thorium loading.

Results of time study. The study was carried to 2000 days of full-power
operation. The mass of U233 fuel and the buildup of PPa?3 are shown in
Fig. 24-29, and the buildup of fission product poisons (other than Xe!3?
and =m!'**) along with breeding ratio are graphed in I'ig. 24-30. The fission-
product polsons vary in an almost Imear manner for burnups corresponding
to 2000 to 6000 days. Other calculations have indicated that extrapolations
(represented by dashed lines on Figs. 24-29 and 24-30) to 5840 days, the
expected life of the plant, are reasonable.

The quantities necessary to evaluate the chemical processing costs for
viarious processing eycles are average values of fuel mass and breeding ratio
(M 2y, M3, and BR). The average value of Mis s approximately the
steadv-state value; Moz and BR are shown in Figs. 24-29 and 24-30.

~Neleetion of chemical processing eycle. The fuel costs which are dependent
upon the chemical processing eycle are fuel inventory, fuel burnup, and
processing charges. These charges were computed using formulas similar
to those deseribed in Article 24-4.4 but using data appropriate to U233 fuel.
[Zquations giving costs in dollars per full power day are

Fuel ventory: Co ($/day) = 2,143 (Lfigfi Mis)
Iuel burnup: Cy ($/day) = 15,250 (1 — BR)

 

 

 

Chemieal processing:
Moz
T

M3s(1)

) | 250,000
:‘Fl‘!

+ T + 596,

 

Cp (8 /day) = 95.875 + 4795

The results of these caleculations are tabulated in Table 24-11 and
craphed i g, 24-31. This analysis indicates an economic optimum
processing cyele of approximately 4000 full-power days. However, only a
<inall penalty of slightly more than $200/day (less than 0.03 mills/kwh) 1s
meirred by operating the reactor for its complete life (5840 full-power days)
before <ending the fuel to a chemical separations plant.
916

LIQUID METAL FUEL REACTOR DESIGN STUDY

[cuaP. 24

 

Fig. 24-29. Mass of U233
vs. reactor operating time
for a single-fluid reactor

 

 

 

operating at 825 Mw.

Reactor Operating Time, In Full Power Days x 103

Fia. 24-30. Neutron
losses vs. reactor operating
time for a single-fluid re-
actor operating at 825 Mw.

Relative Neutron Losses

0.9

G.8

e
N

o
w

o
o

0.1

 

”
- Fission Product Poison

 

 

l A l ]

 

1 2 3 4 5

Reactor Operating Time, In Full Power Days x 10-3

 

 

 

 

 

11000 , 1 , , 1.4
. 10000 |-
O
0O
~
o
g o000 |
@ T
8 S
T ~. .
S 8000 |- £  Fre. 24-31. Varia-
2 2 ble fuel cost for
- 825-Mw single fluid
® 7000 |- LMFR vs. chemi-
—10.9 cal processing cycle
(aqueous batch proc-
6000 L ' ' | | ess).
0 1000 2000 3000 4000 5000 6000

Cycle Time, T, Full Power Days
TaBLE 24-11

VARIABLE FUeL CosT FOR AN 825-Mw LMFR

For various processing cycles

 

 

 

 

Chem. proce. i = 25 Sy Fuel inven- Fuel burnup | Chem. proc. Total vari-
eycle, days Moa(T) | Mis(T) | M2s(T) | Mya(T) | M*25(T) | BR(T) |tory charges, costs, $/day | costs, $/da able cost,

P $/day ’ ) DIGRY $/day
6000 780 30 730 30 760 0.722 1690 4240 1,627 7,557
4000 725 30 675 30 705 0.755 1510 3738 2,101 7,349
2000 671 30 623 30 653 0.787 1400 3230 3,575 8,105
1000 635 30 595 30 625 0.807 1340 2943 6,184 10,470
500 603 30 578 30 608 0.820 1303 2746 16,467 15,500

 

 

 

 

 

 

 

 

 

 

 

[F-¥2

NDISHA HOLOVHY JdINTA-TONIS

 

L6
[cHAP. 24

LIQUID METAL FUEL REACTOR DESIGN STUDY

918

Control Rod Drive

T/Shidding

 

 

S
- SR ///////////////// SRR
y/////// //////, N ///////////////////
////// /////////////////// S
AR // T // e
///// ////////////////////////////////// //
aa A
//////////////////////////////////////////////////////////////
%/////////////////////////////////////////////////////////// .
A R /////////
/////////////////////////////////////////////////////////////
r L ‘ Sl //////////////////////////////%
S ///////////////////////////////////////// —
//////////////// /// ////////// //////%
SV /////////// N
o ///////// ///// //////////////////////////
i /////////// R
//////////////////// AR //////
° SSINEETERIIING /w I
O

’

Fuel Passage

Fuel In

Fic. 24-32. Single-region, externally cooled liquid metal fuel reactor.

The single-region reactor design 1s 1l-

The core is constructed of large blocks of high

density reimpregnated graphite, with 1.5- to 2.0-in.-diameter axial holes

Specifications of reference design.

lustrated in I'ig. 24-32.

The graphite is supported by a number of

compensated molybdenum rods and a bottom support plate.

for the passage of fuel slurry.

Provision 1s

made for three or four liquid metal control rods, if experience indicates

they are necessary.

The reactor vessel is constructed of 219, Cr-19, Mo stecl, 2% inches

thick, designed for a temperature of 1150°F and maximum pressure of

120 psi. Three 28-in.-diameter pipes carry the fluid into the reactor at the
bottom and leave at the top. The entire reactor vessel is doubly contained

I'he reference core

r

A drain line to the fuel

dump tanks is also provided. The free space above the reactor core is used
as the degasser to remove volatile fission products.

by a relatively thin-walled containment vessel.
design has the following specifications:
24-4] SINGLE-FLUID REACTOR DESIGN

Power:

Thermal power
Net electrical power
Station efficiency

Materials:

Fuel

Fertile material
Moderator
Reflector
Coolant

Coolunt-to-moderator ratio, V,/V,

Thorium concentration, Wos

Geometry:
C'ore radius
C'ore height
weflector thickness

Number of primary coolant loops

Fuel-<lurry volume:

Coolant loops 1640 ft?
leactor core 443
Reactor vessel ;600

 

Total

Clenoeal processing cyele

Nuelear data:

Maxs U299

Mass Pa?3s

Mass 233

Initial average core thermal flux
Breeding ratio

Poison fraction

Mass of bismuth

Mass of thorlum

Startup
046 kg

0
546

0.87
0

1,646,000 1b

22,400 kg

825 Mw

515,000 kw

38.27%,

U233

Thorium

919

High-density graphite
High-density graphite
U0O2-ThO2-Bi Slurry

0.5

H.5 ft
14.0 £t
1.5 ft

)

2683 ft3

4000 days

4000 days
725 kg
30
735

0.725
0.216

Average

675 kg
30

705

3 X 101

0.75
920 LIQUID METAL FUEL REACTOR DESIGN STUDY [crAP. 24

24-5. FEcoxowics

Economie considerations were essential to the optimization studies re-
quired to establish the reference designsg presented in Sections 24-2 and
24-4. An important objective of this study is the economic comparison
of energy costs for the single-fluid and the two-fluid externally cooled
LMFR. A brief summary of energy costs for the optimum design in cach
concept 1s presented in Table 24-12.

TABLE 24-12

ExeErGY Cost (809, PraNT FACTOR)

 

 

 

 

 

 

 

Mills/kwh
Single-fluid Two-fluid
LMFR LMFR
Fixed charges on total capital investment 4.09 4.31
Nuclear fuel and inventory costs 1.24 1.41
Maintenance 1.18 1.05
Operation 0.38 0.38
Interest on working capital 0.04 0.04
Total energy cost, mills/kwh 6.9 7.2

 

 

 

24-5.1 Fixed charges on capital investment. Direct construction costs.
The estimated costs of equipment, installation of equipment, and con-
struction are based on the plant layouts for the two reference designs eval-
uated in this study. Construction and erection costs of all items, as well as
direct materials costs for those components manufactured by the Babcock
& Wilcox Company, were developed by B&W estimators. Delivered costs
of equipment supplied by manufacturers other than B&W were taken from
vendors’ quotations.

A summary of direct construction costs for each reference design is
tabulated in Tables 24-13 and 24-14.

Total capital tnvestment. The total capital investments are summarized
according to account numbers in Tables 24-15 and 21-16.

24-5.2 Maintenance and operation. In computing energy costs, the
fixed charges on maintenance equipment and spare parts are included in
the maintenance costs, while fixed charges on buildings used for main-
tenance are included in fixed charges on capital investment.
24-5] ECONOMICS 0921

24-5.3 Fuel costs. The fuel costs as presented in this report include
(1) bismuth inventory, (2) fuel inventory, (3) fuel burnup, (4) thorium
inventory, (5) thorium burnup, and (6) chemical processing. Sodium in-
ventory is not included, since it is used as coolant fluid for the inter-
mediate system and does not contain fuel. Fuel costs are summarized in
Table 24--17.

24~5.4 Summary of energy costs. The energy costs in mills/kwh, based
upon an electric output of 315,000 kw and a plant factor of 80%, are
tabulated in Table 24-18 for various categories.
TasLE 24-13

Sumvary or DirecT CoxsTrUucTION Costs FOR SINGLE-FLuip LMFR

 

 

Account
no.

310

311

312

Land and land rights

 

 

 

 

Direct site
Structures and improvements

Nuclear steam generator building
Accessory buildings
Total

Nuclear Steam Generator and Chemical
Plant Equipment

Reactor
Primary and blanket system
Blanket system
Intermediate system
Steam system
Primary inert gas system (He)
Intermediate gas system (N2)
Reactor heating and cooling system
Dump tank heating and cooling systems

 

 

 

Direct Direct Dlr'egt.

labor materials construction
cost

$500,000 $500,000
Improvements and miscellaneous structures 78,000 126,000 204,000
2,981,000 3,535,000 6,516,000
51,000 84,000 135,000
3,110,000 3,745,000 6,855,000
60,000 2,028,000 2,088,000
309,000 1,470,000 1,779,000
279,000 2,778,000 3,057,000
125,000 1,629,000 1,754,000
5,000 22,000 27,000
6,000 24,000 30,000
21,000 84,000 105,000
100,000 149,000 249,000
Primary system capsule area ventilating system 1,600 2,000 3,000
Primary system reactor cell cooling system 3,000 3,000 6,000
18,000 21,000 39,000

Shield cooling system

 

ah

AALS NOISAAd HOILOVIY TANd TVYLAW dIadll

¥Z dVHD]
 

342-343

 

Water cooling system

Offons system

Feedwater heating svstem

Instrumentation and controls

Spare parts

Miscellaneous equipment

Inventories

Chemical plant equipment
Total

Turbine Generator Equipment
Turbine and condensing

Accessory electrical equipment
Miscellaneous power plant equipment
Transmission structures
Maintenance equipment
Total

Station equipment

Total Direct Construction Cost

 

3,000

20,000
286,000
509,000
110,000
315,000

156,000
2 416,000
828,000
400,000
50,000
172,000
222 000

189,000

 

7,000
50,000
2,541,000
1,236,000
2,666,000
538,000
4,680,000
403,000
20,381,000

13,382,000
2,484,000
130,000
9,921,000
3,051,000

1,246,000

 

10,000
70,000
2,827,000
1,835,000
2,776,000
903,000
4,680,000
559,000
22,797,000

14,210,000
2,884 000
180,000
3,093,000
3,273,000

1,435,000

$51,954,000

 

 

[c-¥g

SOITNONODH

€e6
TaBLE 24-14

SumMarY oF DirecT ConstrUcCTION CosTs ForR Two-Fruip LMFR

 

 

 

 

Account Direct Direct Direct
no. labor materials cost .
construetion
310 Land and land rights
Direct site $500,000 $500,000
311 Structures and improvements
Improvements and miscellaneous structures $78,000 126,000 204,000
Nuclear steam generator building 3,184,000 4,035,000 7,219,000
Accessory buildings 51,000 84,000 135,000
Total 3,313,000 4,245,000 7,558,000
312 Steam generalor and chemical plant equipment
Reactor 50,000 1,531,000 1,581,000
Primary and blanket system 436,000 1,788,000 2,224,000
Intermediate system 295,000 3,365,000 3,660,000
Steam system 250,000 1,949,000 2,199,000
Primary inert gas system (He) 8,000 34,000 42,000
Intermediate gas system (INg) 8,000 40,000 48,000
Reactor heating and cooling system 21,000 84,000 105,000
Dump tank heating and cooling systems 100,000 149,000 249,000
Primary system capsule area ventilating system 1,000 2,000 3,000
Primary system reactor cell cooling system 3,000 3,000 6,000
Shield cooling system 18,000 21,000 39,000

 

 

 

 

 

1743

AdNas NDISId HYOLOVAY THANd TVIAN aI1ndIl

¥g "dVHD]
 

314

315

316

342-343

 

Water cooling system
Offgas systen
Feedwater heating svstem
Instrumentation and controls
Spare parts
Miscellaneous equipment and inventories
Inventories
Chemieal plant equipment
Total

Turbine generator equipment
Turbine and condensers

Accessory electrical equipment

Mauscellaneous power plant equipment
Transmission structure
Maintenance equipment

Total

Station equipment

Total Direct Construction Cost

 

 

 

3,000 7,000 10,000
20,000 50,000 70,000
200,000 2,546,000 2,856,000
765,000 1,547,000 2,312,000
72,000 2,000,000 2,072,000
273,000 564,000 837,000
4,289,000 4,289,000

540,000 1,461,000 2,001,000
3,153,000 | 21,430,000 | 24,583,000
828,000 | 13,382,000 | 14,210,000
401,000 2,503,000 2,904,000
50,000 130,000 180,000
173,000 2,274,000 2,447,000
223,000 2,404,000 2,627,000
189,000 1,246,000 1,435,000
$8,107,000 | $45,710,000 | $53,817,000

 

 

 

 

R

SOITWONODHA

Ga6
TABLE 24-15

SUMMARY OF CAPITAL INVESTMENT FOR SiNGLeE-Fruip LMFEFR

 

 

 

 

 

 

 

 

 

 

 

 

 

Direct . .
Account construction Contingeney Mw(:ellanei)us Total capital
no. e - charges mvestment
costs
310 Land and land rights $500,000 &0 $0 $500,000
311 Structures and inmiprovements
Nuclear and turbogenerator plant 6,697,000 1,479,000 3,158,000 11,334,000
Chemical plant 158,000 34,000 71,000 263,000
312 Nuclear steam generating and chemuical
plant equipment
Nuclear plant equipment 14,782,000 2,753,000 3,436,000 20,951,000
Chemical plant equipment 559,000 112,000 188,000 859,000
Spare parts 2,776,000 479,000 418,000 3,673,000
Inventories 4,680,000 0 0 4,680,000
314 Turbine generator equipment 14,210,000 2,483,000 2,346,000 19,039,000
315 Accessory electrical equipment 2,884,000 531,000 655,000 4,070,000
316 Miscellaneous power plant equipment
Transmission structures 180,000 36,000 61,000 277,000
Maintenance eguipment 3,093,000 539,000 505,000 4,137,000
342-343 Station equipment 1,435,000 263,000 318,000 2,016,000
Total $51,954,000 %8,689,000 11,156,000 $71,799,000

 

 

 

*Includes indirect construction cost, interest, and engineering charges.

9¢6

AALS N9ISTd HOLOVIM TEAL TVIHN dI1adI1

F¢ "dVHD]
 

TapLe 21 16

Suvvanry or Carrran Investvaest ror Two-From LMFR

 

 

 

 

 

Direct : al eanits
Account construction Contingency Mlblaie‘llfme‘ous Tot:l,l-«é dpltgl
no. costs charges investimen
310 Land and land rights $500,000 $0 %0 $500,000
311 Structures and improvements
Nuclear and turbogenerator plant 7,010,000 1,538,000 3,223,000 11,771,000
Chemical plant 548,000 119,000 247,000 914,000
312 Nuclear steam generating and chemical
plant equipmient
Nuclear plant equipment 16,221,000 2,930,000 3,915,000 23,066,000
Chemical plant equipment 2,001,000 401,000 658,000 3,060,000
Spare parts 2,072,000 360,000 301,000 2,733,000
[nventories 4,289,000 4,289,000
314 Turbine generator equipment 14,210,000 2,483,000 2,346,000 19,039,000
315 Accessory electrical equipment 2,904,000 534,000 659,000 4,097,000
316 Miscellancous power plant equapment
Transmission structures 180,000 36,000 61,000 277,000
Maintenance equipment 2,447,000 432,000 427,000 3,306,000
342-343 Station equipment 1,435,000 263,000 318,000 2,016,000
Total $53,817,000 $9,096,000 $12,155,000 $75,068,000

 

 

 

 

 

 

 

 

 

 

 

 

[e-¥2

SOINONODA
TABLE 24-17

SUMMARY oF FUEL CosTs

315 Mw (elec.); plant factor = 809

 

Capital mvestment

Annual cost, $/yr

Energy cost, mills/kwh

 

 

 

 

 

 

 

 

 

 

 

Item
Single-fluid Two-fluid Single-fluid Two-fluid single-fluid Two-fluid

Bismuth inventory $3,704,000 $:3,000,000 $444,000 $371,000 (.201 0.168
IFuel inventory 455,000 396,000 (. 206 0.179
Fuel burnup 1,001,000 627,000 0.494 0.284
Thorium inventory 941,000 1,171,000 141,000 176,000 0.064 0.080
Thorium burnup 9,000 10,000 0.004 0.005
Chemical processing:

Offsite processing 144,000 0.063

Buildings 263,000 941,000 35,000 127,000 0.016 0.058

LFquipment 859,000 3,060,000 116,000 744,000 0.053 0.337

Operating costs 18,000 530,000 0.008 0.240

Shipping charges 65,000 0.002

Thorium inventory 13,000 0.006

Fuel inventory 44,000 0.020

FPuel depreciation 222,000 124,000 0.101 0.056

Total $5,767,000 58,262,000 $2,744,000 3,105,000 1.24 1.41

 

 

 

 

8GH

AJALS NDISEJ ¥0LOVHY TIad TVIHN dindrl

¥¢ 'dvHO]
24-5] ECONOMICS 929

TaBLE 24-18

Unit ExErcy CosTs

 

 

 

 

 

 

 

r Cost, mills/kwh
Item
Single-fluid Two-fluid
Land and land rights 0.03 0.03
Structures and improvements (less chemical
processing facilities) 0.69 0.72
Equipment (less maintenance equipment and
spares):
Reactor vessel and internals 0.19 0.14
Primary and blanket system 0.21 0.27
Intermediate system 0.36 0.44
: Feedwater heating system 0.26 0.27
: Instrumentation and controls 0.29 0.36
! Aiscellaneous equipment and Na inventory 0.11 0.10
| Auxiliary systems (.22 0.27
| Station equipment 0.14 0.14
“ Accessory electric equipment 0.28 0.28
- Turbine generator equipment 1.29 1.29
" \liscellaneous power plant equipment 0.02 0.02
Fuel costs (includes chemical processing fa-
cilities} 1.24 1.41
. Plant operation 0.38 0.38
Maintenance (includes maintenance equip-
ment and spares) 1.18 1.05
Interest on working capital (.04 0.04
Total 6.93 7.21

 

 

 

REFERENCES

1. Bascock & Wincox Co., 1958. USAEC Report BAW-1046.

2. Bapcoek & Wincox Co., Ligutd Metal Fuel Reactor; Technical Feasibility
Report. USAEC Report BAW-2(Del.), June 30, 1955.

3. Bascock & Wincox Co., 1958, USAEC Report BAW-1048.

4. W. L. RoBBa et al., Fission-product Buildup in Long-burning Thermal
Reactors. Nucleonics 13(12), 30-33 (1955).

5. Bascock & Wincox Co., A Review and Evaluation of Maintenance Concepts
for Liguid Metal Fuel Reactors, USAKC Report BAW-1047, March 1958.