ML19347B071
| ML19347B071 | |
| Person / Time | |
|---|---|
| Site: | 07000371 |
| Issue date: | 09/17/1980 |
| From: | Stevenson R NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS) |
| To: | Kirk W UNITED NUCLEAR CORP., UNITED NUCLEAR CORP. (SUBS. OF UNC, INC.) |
| References | |
| NUDOCS 8010010477 | |
| Download: ML19347B071 (6) | |
Text
-- _
$'f)
I UNITED STATES
^(
NUCLEAR REGULATORY COMMISSION o
- ,2 WASHINGTON, D. C. 20555
%,...../
FCUP:RLS70-371 SEP 171980 United Nuclear Corporation Naval Products Division ATTN: Mr. William F. Kirk, Manager Nuclear and Industrial Safety 67 Sandy Desert Road Uncasville, Connecticut 06382 Gentlemen:
We have reviewed the classified application dated August 8,1980, (NIS:
80-8-9), containing your responses to the questions on your application of September 20, 1979, sent you with our letter of April 15, 1980. The responses to some of the questions, primarily those concerning units of non-spherical geometry and tiered arrays, are not satisfactory for reasons given in the enclosure. We will continue our review of your application following receipt of the additional information.
Sincerely, h-Robert L. Stevenson Uranium Process Licensing Section Uranium Fuel Licensing Branch Division of Fuel Cycle and Material Safety
Enclosure:
Comments and Questions on Application Dated August r
- 980, Docket 70-371 3
8010010 T7
Comments and Questions on Application Dated August 8, 1980, Docket 70-371 1.
To justify the application of criteria derived for spherical units to units of other geometries, you referred to Figure 3 of Document Y-CDC-13, Figure 2, in an article in Vol.12 of Nuclear Technology, and Figure 3 of Document ORNL-CDC-4.
You also provided the results of a pair of KEN 0 calculations; one calculation for spheres and one for cubes, presumably in single-tier arrays under plant reflector conditions.
(The reflector was not described.) We offer the following coments and questions to explain why we do not believe that applica-bility of the criteria to non-scherical geometries has been demonstrated and why additional justification is needed:
a.
The derivation of your spacing criteria was based on spherical units and the confirmatory KEN 0 calculations were made for single-tier, spherical units, except for the single calculation mentioned above and summarized on page 3.9-5A, Part II.
b.
Figure 3 of Y-CDC-13 concerns cubic. arrays of cylinders of limited height-to-diameter range.
l c.
Figure 2 of reference 7 (article by T. Gutman in Vol. 12 of Nuclear Technology) concerns unreflected arrays of units embracing a narrow range of unit geometries.
1 d.
As noted, Table 3 of Document ORNL-CDC-4 (there is no Figure 3),
indicates that cubic units in unreflected cubic arrays have higher multiplication factors than spherical or cylindrical units of comparable unit k-effective when spaced to produce the same average fissile density.
How does a single pair of calculations demonstrate the general conclusion that this effect of unit geometry is negligible over the entire range of geometries and fuel compositions covered by your application?
2.
" Vertical stacking" (i.e., tiering) of units is justified by reference to Table 6 of Document Y-CDC-13 and with the provision that units be spaced 12 inches edge-to-edge, a.
Again, we note that essentially all of the KEN 0 check calculations were made for arrays of single-tier, spherical units, b.
We_ note that the derived surface density limits may have involved the use of array calculations for metal units with a minimum of four units per tier (nz >4), but it is not obvious what tiering does to the margins of subcriticality inferred from the check calculations referred to in paragraph a, foregoing.
Thus, in a draft of a forthcoming paper by Thomas (to appear as Document ORNL/NUREG/CSD TM-15), there is a table showing an effect of tiering (of water-reflected infinite planar arrays) on the order of 5% on k-effective.
Clearly, it is non-conservative to apply
y 2
spacing criteria to tiered units that are based on single-tier arrays. A copy of the relevant text and table from Thomas' study are attached.
c.
Justification should be provided for the stated minimum 12-inch vertical spacing for tiered units.
Consideration should be given to the fact that the referenced studies (e.g., Table 6 of Y-CDC-13) by Thomas generally used cuboidal cells so that vertical spacing was a variable.
3.
Please explain the meaning of the new footnote on page 10A-4, Part II.
't 4
OfGf$
a Table 9.
Monte Carlo Calculated k-eff for Infinite Planar Arrays with n = 1, 2, and 4.
Unit Cell Mass Dimension Calculated n
kgU d (cm) k-eff f.g 4
1.56 15.794 0 999 0.006 i
1 7 968-0.959 0.006 4
3 23 23.528 0.996 0.006 1
11.764 0 952 0.004 4
5.12 30.728 0 992 0.006 2
21.728 0.963 0.005 1
15.364 0 944 0.005 4
13.47 55.746 1.001 0.006 2
39.416 0.983 0.004-1 27.872 0.959 0.005 4
29 53 108.466 1.004' O.006 2
76.698 0 987 0.005 1
54.234 0 977 0.004 4
>ca, s ha enu Au; O
e e
G O
4 m
4
L)TG 44 I
Infinitt Planar Arrays
(...
Uater reflected critical arrays of U(93.2) metal spheres in infin-ite planar geometry are characterized by Eqs. 5 and 6 withe = 4.58 x 10*
t cm'*(
1.8 x 10-4) and.the unreflected critical mass of 52.1 kgU.
For cells in the z-direction of the array and t.n unlimited number of n, n z and a cells, Eq. 5 can be written as y
6-(m) =
.(21) and is a constant for ns it 4.
We will remark on this n.,, restriction lat]er.
The expression for Tim) is the usual definition of surface den-sity, the mass in the array projected onto one of the array surfaces.
If we st'ppose an m and d satisfying Eq. 21, it is clear that there n
{
is a finite N unit array by Eqs.11 and 12 which will also result in criticality for the mass and spacing.
This is 41 valent to asking what values of n /n and R will result'in a e eqtaal i that 'for the inf Laite z
t planar array, et = 4.58 x 10-Y
-1 em This 1,s expressed by Eq. 12 as Dz 4
c2 0.458 E (SR 0.672_1) " {
- 1.762 = 0.2 Q = p$g( 0 Given a value of the array shape factor R, the ratio n /n is defined.
z T.i Table B illustrates values of N resulting from given values of R when n 2
-4.
O m
5 S
f0.,J,
47 C
standard deviation for a k-eff calculation, typically 1 0.005, then
$9- - 0. 01 a
and we can conservatively write Ib = 1.010N within the ability to define criticality.
This result suggests that it would be acceptable to neglect solid angle contributions to an array i
equal to about 1 percent of the O for the array.
l This analyses demonstrates the applicability of Eq.15 to infinite planar arrays.
Thecoefficientpbecomes
+
i s = F(N) A c (22) ng 2
and for water reflected metal spheres is
("
S = 0.11 nz The values of F(N) may be those of Table 8 and these are shown in Fig.
FM pi 11 for nz = 4.
Values for nt= 1 aie also given.
The, values would be scaller if n p n for a given N.
x y
Let us return to the n restriction to the application of Eq. 21.
A nu=ber of calculations were performd to evaluate water reflected infinite planar arrays specified by Eq. 21 with n = 1 and 2.
These are y,
su=mari::ed in Table 9 and show that the results produce k-eff's less q
than unity if the ru triction is ignored.
Ec;uation 21 is thus conserva-tite in appli::ation to spherical units when n is less than fcur but may result in suboriticality.'cr other unit shapes, not s
R*]O
- g~ T } Q D
' sa a lu s R 1..N L
_