ORNL-TM-2282.txt

 

_____

 

PRELIMINARY STUDY OF A MOLTEN-SALT BURST REACTOR

6AjK; RIDGE 'NATIONAL LABORATORY

operated by
UNION CARBIDE CORPORATION
| for the

u.s. ATOMIC ENERGY COMMlSSION

UNION
CARBIDE

 

 

ORNL TM- 2282

 copyno.- F

DATE - July 10, 1968

  

, ,'A M Perry
H. F. Bauman |
~ W. B. McDonald
J. T. Mihalczo

Diétribfifioh X
‘ Dr. Vlctor Rcuevskl ’ ISPRA

]
2. - M. J. Skinner
3.-5. DTIE '

e “0"05 Thxs document contoms mformahon of .a preliminary nature

and was prepared pnmanly for internal use at the Oak Ridge National

Zchorufory It is subject to revision or correcflon and therefore does
not represent a fmal report . :

¥l
VoL

' CEWIUTON OF THS DOCUMENS IS UNIMITED

 
 

 

 

—— LEGAL NOTICE-

[ Thir report was prepared as an account of Government sponsered work. Kefther the United
| States, nor the Commission, nor any pergon acting on behalf of the Commission:

i 4. Makes zuy wartanty or representation, expressed or implied, with respect to the accu- “an
| racy, completeness. or usefilness of the information contained ir this report, or that the use . S
i of any information, apparatus, methad, or process disclesed in thig report may not iniringe
| privately owned righis; or g ;

B. Assames sny liabilitles with respsct to the uge of, or for damages resuiting froin the . R
use of any information, apparstus, method, or process digclosesd in this report.

As used in the above, *“person acting on behalf of the Commisglon® includes any sm-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
such smployes or contractor of the Commigsion, or employeo of such conirzctor prepares,
disseminates, or provides access to, any information pursuant to his empivyment or contract 3
witk the Commiesion, or his employment with such contracter.

 

 

PRELIMINARY STUDY OF A MOLTEN-SALT BURST REACTOR

A. M. Perry W. B. McDonald
E. F. Bauman J. T. Mihalczo

The guestion has been posed whether one could pulse a molten-salt
reactor so as to achieve an integrated flux of 10® neutrons/cm® per
burst in a test cavity at the center of the reactor, and, if so, what
energy yield would be required and what would be the shape of the pulse?
The nominal objectives appear to be:

1) 1-2 X 10*® neutrons/cm® per burst, without a tight specification
on the neutron spectrum,

2) a burst width in the neighborhood of 10 msec or less.

Since it appears that the second objective should be easy to satisfy,
~we have looked at the possibility of achieving the Tirst without much

regard for factors which might influence the burst shape, and have sub-

sequeptly estimated the burst width for one of several reactor con-

figurations one might contemplate using.

 

1. Selection of Salt

~ The ealculations were based on use of the salt LiF (7% mole %)
UF, (27 mole %), and for each reactor considered, the critical enrich-
-menfi,of'the uranium was determined. Separated Li7 would of course be
used. |

Relevant properties of the salt are:

Melting point, °C 450
Specific heat of the melt, cal/gm °C 0.217

| Density of the melt, gm/cm’ 5.26 — §,3 X 10°% T(°C)

2. Allowable Energy Input

 

We postulate that the temperature of the salt could be allowed to

 

rise 1000°C, i.e., nominally from 500°C to 1500°C. The properties

given above yield the following values at 1000°C (average termperature):

et R

e TR e} A
‘u:’-?Efif‘nfi.tkrfw L
‘Density’ 4,33 gm/cn”®
Specific heat 4 wsec/em® °C
Thus, the integrated fiux that can be obtained in the test cavity

corresponds to a maximum energy density of h-kwsec/cm3 per burst.
3, Geometry

‘For the purpeses of this preliminary exploration, we postulate

spherical geometry, with concentric regions and dimensions as follows:

Test cavity {void) 15 enm radius

Ni shell C .1 em thick
Fueled salt 15 to 60 cm thick
Ni shell * 5 ecm thick
Graphite shell - 25 em thick

An addi%ional case was exanined with the thickness of the outer Ni con-
teiner increased to 15 cm, and the-graphite'shell omitted; this was

done for a 60-cm-thick salt region.

4. Results

The integrated fluxes obtainable in the test cavity are shown for
each case in Table 1, along with geometrical specifications and other
~ derived results. ©Some of these results are . also plotted in Fig. 1.
. -The neutron spectrum for case #8, with $2-cm-OD core (3 ft) is
shown in Fig. 2, as the integral above energy E, as & function of I,

i.e.,

[ at fm(i)(E',t)dE' ,
E . .

where the time integration is taken over the duration of the burst. For

s

S comparison, the fission spectrum is also given, normalized to 1 X 1016
neutrons. o . B

In case #3, which was the same as case #1, except that the S.cm Ni
and 25-cm graphite shells were replaced by a single 15-cm Ni shell, both
the peak-to-average power density ratio andrfhe'integrated flux in the

test cavity were the same as in case #1.
i

Table 1. Burst Characteristics vs Core Size

 

Integrated Flux in

 

 

 

Fuel® , Test Cavity
Case Cutside Fuel Critical ax Burst
Nurber =~ Diameter  Volume Enrichment 235U Mass = Total >100 kev  yield
( cm) (liters) (g (kg) avg (neutrons/ecm?® X 107*6) - (Mw sec)
h 152 1821 17.5 854 1.545 z,L 0.97 L7k
1 122 g3L 19.7 93 1.423 3,1 0.95 2625
6 112 718 21.1 398 1.378 3,0 0.94 2084
7 102 538 - P22 320 1,331 2.8 0.92 1615
8 92 . 301 2L .2 253 1.282 2.7 0.91 121k
9 &o 272 27.3 200 1.231 2.5 0.88 883
10 72 178 72,1 153 1.178 2,2 0.84. 605
11 62 108 414 120 1.125 1.7 0.77 3832

 

aFor all cases shown, cavity outside diemeter = 30 cm, inner shell thickness = 1 cm, outer
shell thickness = 5 cm, grephite shell thickness = 25 cm.
 

 

ORJL D

.y

W

~
Ve

£5-1200

 

 

hearely

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
.
S

.+ _.ORKL D45, 65-12008

3 4 %S 67891 2 3 4 567891 - 2 3 4 5 6789% .2 3 4 3567891 2 3 4 567891

~
~
»

W AU O BO.

          

       

 

~ .
Q.

K”"

W b RNNEGED-

~N
R

“a,
D
o)

W & B ANOD

s
8

An estimate of the burst width was made for case #2 (not listed
in Table 1) which was the same as case #1 except that the graphite shell
thiqkness was 10 cm. The Rossi ¢ at delayed critical, Oy s Was calcu-
lated, by use of the DTF transport code, and found to be 9 X 10° sec~t,

The burst width was estimated approximately from the expression®
O = 3.5/& ’

where =O£‘i (p—L) and p is the reactivity in dollars} Thus,; for p =
2 dollars, the burst width at half maximum is estimated to be 0.4 msec.

The reactors smaller than case #2 would have narrower bursts.

5. Discussion

v

Several interesting features are apparent from these results.
a) The reactors are all fast, i.e., from 25 to 40% of the total

~ integrated flux 1s above 100 kev, and essentially all of the flux is

abdve 1 kev.

b) - The flux of neutrons above 100 kev is qfiite insensitive to
reactor size, when the normelization is for & given maximum energy
density (e.g,, L kwsec/cm?). This is so because the peak-to-average
power density ratio rises with increasing core aize (see Table 1). The

total flux falls off fairly sharply with decreasing core size below

- .perhaps 30 in. diam.

c) A single nickel container shell, perhaps 6 in. thick, appears
.

to be a satisfactory reflector, yielding quite flat power distributions,

Nat least for the case studled. It is possible that some gains could be

realized by optimizing the container-reflector regions of the assembly.

d) Burst widths. While the burst widths have not been calculated

with great care, the results obtalned for one of the largest'cores

studied appear to confirm the expectation that the pulses of the desired

magnitude will be less than l msec in width,

 

1T F. Wimett et al., M"Godiva II —-An Unmoderated Pulse Irradiation

. Reactor," Nucl. Sci. Eng., 8(6): 691-708 (1%0).
R Very little attention has so far been given to engineering aspects
of & practical burst reactor. The guenching mechanism, depending on
expansion of the liquid fuel, can be significantly affected by details
of the geometrical arrangement, such &s the shape of the core (e.gz.,
spherical or cylindrical), and the location of the free liquid surface.
The range of temperatures that can be permitted may be extended
-somewhat by pressurizing the cover gas, and in any event the core vessel
will have to be capable of withstanding substantial mechanical shocks.
A thick vegsel will therefore probably be required.
Details of the control mechanisms, and in particular of the devices
for introducing reactivity very rapidly, will require considerable
thought. 2 L e

6. Summary

A first lock at the possibility of using a molten-salt reactor to
U produce intense, sharp bursts of neutrons indicates that fluxes of
. 2—-3 % 10%® peutrons/cm® per burst can be achieved in s central test .
- ' cavity. This can be accomplished with a core perhaps 30 in. in outer
| diameter, having a volume of about 200 liters, and with a burst yield
of about TO0C Mw sec. The neutrons are essentially all above 1 kev, and
a third of the flux is above 100 kev. The estimated burst width is less

than 1 msec.
10

Aggendix

| One of the above,reacfiers, operating in burst mode with a maximum
energy density of L Mw sec/liter, mey be compared with a similar reactor
‘operating at constant power with a maximum power density of 4 Mw/liter
and with & coolant,temperature rise, AT, equal to 1000°C times the
- residence time of the. fuel in the core, in seconds. ihe implication is
- of course that a reactor like case #10, for example, at a power level of
' 600 Mw, would produce a totel (fast) flux in the test cavity of 2 x 1016
- neutrons/cm sec. : fi
| The power density of L MM/liter is probably attainable., One is
led to speculate that such a reactor, with cooling adequate for 600-Mw
operation, could be operated in pulsed mode with & pulse rep;tition rate
of 1/sec. Since the pulses could then not be initiated from a very low
neutron level; it is doubtful that pulses as short as those cited above
could be achieved. We have not yet estimated what pulse shape might be
'obtalned at such a high repetition rate, but there appear to be some
interesting possibllities here worth further 1nvestigat10n. _' '

.
T

wyet |

o