05.08.2013 Views

ORNL-1816 - the Molten Salt Energy Technologies Web Site

ORNL-1816 - the Molten Salt Energy Technologies Web Site

ORNL-1816 - the Molten Salt Energy Technologies Web Site

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

,<br />

ANP QUARTERLY PROGRESS REPORT<br />

The addition of <strong>the</strong> swirl pump made it neces-<br />

sary, however, to replace <strong>the</strong> slinger ring, previ-<br />

ously employed, with a slinger impeller in order<br />

to buck any leakage from <strong>the</strong> swirl pump. Although<br />

<strong>the</strong> complete assembly has not yet been tested in<br />

an inverted position, <strong>the</strong> slinger impeller seems to<br />

be very effective. A similar arrangement is now<br />

being tried as a means of controlling <strong>the</strong> leakage<br />

from <strong>the</strong> main pump discharge through <strong>the</strong> clearance<br />

between <strong>the</strong> impeller and <strong>the</strong> casing past <strong>the</strong><br />

centrifuge. More conventional sealing arrange-<br />

ments were found to be ineffective because of <strong>the</strong><br />

abnormally large radial clearances (0.060 in.)<br />

specified in <strong>the</strong> design objectives. Initial tests<br />

currently in progress with <strong>the</strong> centrifuge seal<br />

impeller have been quite promising. Although no<br />

detailed studies of <strong>the</strong> off-gas system and fuel<br />

pill addition system have been made, <strong>the</strong> changes<br />

-removal system will make it neces-<br />

off-gas system connection and <strong>the</strong><br />

fuel pill addition be made through <strong>the</strong> swirl<br />

chamber ra<strong>the</strong>r than through <strong>the</strong> pump housing as<br />

i I I ustrated previous I y.<br />

CONTROL ROD DESIGN CONSIDERATIONS<br />

The design of <strong>the</strong> ART has been predicated on<br />

<strong>the</strong> belief that no fast-moving control rods will be<br />

required because of <strong>the</strong> inherent stability of <strong>the</strong><br />

reactor that will arise from <strong>the</strong> strong negative<br />

temperature coefficient of <strong>the</strong> circulating fuel.<br />

It is fur<strong>the</strong>r expected that only about 1% Ak/k<br />

will be required to compensate for xenon poisoning,<br />

since it is anticipated that with <strong>the</strong> bypass ex-<br />

pansion tank and gas-scrubbing system planned,<br />

<strong>the</strong> xenon will be removed almost as rapidly as it<br />

is formed. It is expected that compensation for<br />

burnup and <strong>the</strong> accumulation of fission-product<br />

poisons can be effected by adding pills of solid<br />

fuel having a high percentage of U235. Thus <strong>the</strong><br />

principal function of <strong>the</strong> control rod will be to<br />

control <strong>the</strong> mean temperature level of <strong>the</strong> reactor<br />

fuel. Since <strong>the</strong> temperature coefficient will<br />

probably be approximately 5 x Ak/k per OC<br />

and since a temperature level variation of <strong>the</strong><br />

order of 15OoC will probably be required, provision<br />

of a Ak/k of 1% in <strong>the</strong> control rod for temperature<br />

level control should be more than adequate. In<br />

reviewing <strong>the</strong> various requirements for control, it<br />

thus follows that a control rod having a total keff<br />

of 2% should prove to be adequate. Thisreactivity<br />

24<br />

’lbid., Fig. 3.4, p 40.<br />

can be readily incorporated in a single slow-moving<br />

control rod. It would seem entirely adequate to<br />

move this control rod at a rate of 1% Ak/k per<br />

minute, a slow and conservative value.<br />

The size and stroke of <strong>the</strong> rod are important<br />

considerations. Two items of experimental data<br />

are available to indicate <strong>the</strong> size of <strong>the</strong> rod re-<br />

quired, In <strong>the</strong> course of <strong>the</strong> second reflector-<br />

moderated reactor critical experiment (that making<br />

use of powdered fuel placed inside 1 k-in.-square<br />

aluminum tubes), it was found that with an 18-in.-<br />

dia core and a 9-in.-dia island a control rod<br />

effectiveness of about 0.6% Ak/k was obtained<br />

from a 3/,6-in.-OD stainless steel tube filled with<br />

powdered boron to a density of about 50% and<br />

inserted in one-half of <strong>the</strong> critical experiment<br />

assembly. A second indication of <strong>the</strong> effective-<br />

ness of control rod material is available in that a<br />

piece of gadolinium oxide approximately <strong>the</strong><br />

diameter and thickness of a 25-cent piece was in-<br />

serted at <strong>the</strong> center of <strong>the</strong> island in <strong>the</strong> three-<br />

region octahedron critical assembly (without core<br />

shells). The value of this gadolinium oxide wafer<br />

was found to be approximately 0.2% Ak/k. Some<br />

additional data are also available from multigroup<br />

calculations. All <strong>the</strong>se data seem to indicate that<br />

a rod approximately to ’/2 in. in diameter having<br />

a stroke of approximately 20 in., about 7% in. of<br />

which would be below <strong>the</strong> equator, should prove<br />

adequate to give a Ak/k of 2%.<br />

The heat generation to be expected in <strong>the</strong> con-<br />

trol rod will establish <strong>the</strong> amount of cooling that<br />

will be required and, to a large degree, much of<br />

<strong>the</strong> detail design of <strong>the</strong> rod. The heat generation,<br />

in turn, is dependent, in part, on <strong>the</strong> number of<br />

neutrons absorbed in <strong>the</strong> rod at full power. Multi-<br />

group calculations indicate that 1.3% of <strong>the</strong><br />

neutrons produced will be absorbed in a control<br />

rod having a Ak/k of 1%, if it is assumed that <strong>the</strong><br />

average effectiveness of <strong>the</strong> control rod is half<br />

that of <strong>the</strong> portion at <strong>the</strong> center of <strong>the</strong> island.<br />

Thus a control rod having an effectiveness of 2%<br />

Ak/k would, when in <strong>the</strong> “full in” position, absorb<br />

2.6% of <strong>the</strong> neutrons. If a value of 3 Mev per<br />

neutron absorbed is assumed, <strong>the</strong> energy associated<br />

with neutron absorption in <strong>the</strong> control rod would<br />

amount to0.039% of <strong>the</strong> total power generated in <strong>the</strong><br />

reactor. This would total 24 kw for a 60-Mw power<br />

level. The control rod cooling requirement would<br />

be a minimum if gadolinium ra<strong>the</strong>r than boron were<br />

used as <strong>the</strong> absorbing material in <strong>the</strong> rod. This<br />

. .<br />

,<br />

J

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!