ML20127N895
| ML20127N895 | |
| Person / Time | |
|---|---|
| Site: | Monticello  |
| Issue date: | 12/31/1977 |
| From: | NORTHERN STATES POWER CO. |
| To: | |
| Shared Package | |
| ML20127N882 | List: |
| References | |
| NUDOCS 9212010446 | |
| Download: ML20127N895 (16) | |
Text
..
e l
)
l l
MotiTICELLO NUCLEAR GDIERATING PIANT Docket No. 50-263 License No. DPR-22 l
August 1977 DESIGN REPORT AND SA'fETY EVALUATION FOR REPIACD1DIT OF SPDiT TUEL POOL STORAGE RACKS Incorporating:
Supplement 1 December, 1977 9212010446 771208 PDR ADOCK C5000263 P
PDR m
s-
i i
t 1
i 3.2 Fuel Storage System Construction j
The fuel storage module is a f abricated stainless steel structure composed of fuel storage tubes, made by forming an outer tube and 3
an inner tube of 304 stainless steel with an inner core of Boral*
l into a single fabricated tube. The outer and inner tubes are welded together af ter being sized to the required dimensional tolerances by a patented process. The completed storage tubes j
are fastened together to form a 13x13 storage module. Each 13x13 module is approximately 7 feet square and 14 feet high and provides j
storage space for 159 BWR fuel assemblies.
i I
Cylindrical columns, 8-inches long are welded to the underside of the 1
module base assembly. The columns transfer the module forces to the-
)
fuel storage pool slab and raise the module above the floor of the i
fuel pool sufficiently to permit natural circulation of cooling water to flow to the modules.
t 4.0 SAFETY EVALUATION i
4.1 Criticality Analysis 3
4.1.1 The principal Analytical Model i
The criticality analysis calculations were performed with the MERIT computer program. The MERIT program is a Monte Carlo l
program which solves the neutron transport' equation as an eigen-value or a fixed source problem including the neutron shielding problem. This program is especially written for the analysis of 4
fuel lattices in thermal nuclear reactors. Geometries with up to three space dimensions and neutron energies 'between 0 and 10 MeV can be handled. The MERIT program uses cross sections processed l
from the ENDF/B-IV library tapet.
4.1.2 The Model for Verification s
The qualification of the MERIT program rests upon extensive quali-i fication studies including Cross Section Evaluation Work Group (CSEWG) thermal reactor benchmarks (TRX-1,
-2,
-3, -4) the B&W UO2 and Pu0, criticals, Jersey Central experiments, CSEWG f ast reactor benchmarks (GODIVA, JEZEBEL), the KRITZ experiments, and in addition, comparison with alternate calculational methods.
Boron was used as solute in the moderator in the B&W UO, and'B&W 2 criticals, and as a solid control curtain in the J8rsey Central Pu0 exper iments. The MERIT qualification program has established a bias 2
of.005 +.002 (10) ak with respect to the above critical experiments.
Therefore, MERIT underpredicts k,gg by 0.5 percent ak.
1 i
i
- product of Brooks & Perkins. Inc. Consisting of a layer of B C-Al matrix 4
i bonded between two layers of aluminum.
Supplement 1 7
December, 1977
~--
u I
s 1
4.3.1.9 The base plate of each storage module is raised above the floor of the pool suf ficiently to permit natural circulation of cooling water flow to the modules. Analysis has confirmed that frictional forces between module support and the floor and the low seismic i
overturninE moment of the racks make them stable under all conditions
)
of storage.
5 4.3.2 Analytical Methods l
Appropriate modeling of the fuel storage module was developed for j
each structural component and mass values assigned over the height
]
in eleven mass nodes. The modules were combined into an idealized i
8-module array and the pool wall was included to determine hydro-dynamic mass effects. The modules were analyzed as a cantilever besm attached to a rigid base, using DYSEA, a GE-developed verster
]
(qualified level 2) of SAP-IV modified to derive loads in a water filled rectangular pool. These loads were derived for the horizontal and vertical accelerations specified in the General Electric BWR Systems 4
Department seismic criteria document and were compared to the allowable d
stresses in the reference documents. The analysis indicates that the d
derived loads do not overstress the modulet thus, it can be concluded I
that the modules are not overstressed for the Monticello application l
since the Monticello accelerations at the fuel pool elevation are l
0.2g (SSE) and the analysis was done for 3g (SSE).
I Monticello accelerations are those in Ref. 4.3.1.4, Bechtel, Inc.
for the Monticello building fuel pool elevations 'and those shown herein 1
in Figure 4.3-2 for the pool floor elevation.
s Generic accelerations are those in GE Document 384HA137. Rev. '1, i
"BWR-6 Seismic Design Specification".
1 j
Table 4.3-1 compares the forces for the GE specified seismic accelerations at the fixed-base module frequency of 12.17 hz with those for Monticello, l
i 4.3.3 Discussions of Results
- 4.3.3.1 The natural frequencies of the 13x13 module were calculated by accounting for the stif fness of the modules and support columns and the hydrodynamic ef f ect of the surrounding fluid. These natural frequencies were found to be greater than 8.0 hz in.the horizontal direction and 35.0 hz in the vertical direction. Frequencies of 5.5 he j
and 17.0 hz were used to obtain the spectral accelerations used in the force analysis which produce conservative values relative to the I
^ higher natural frequencies.
4.3.3.2 Maximum displacement-at the top of the modules for-the X direction 1
or the Y direction (the modules are symmetrical) is 0.07 inch.
l t
d Supplement 1 December, 1977-i 22 i
4
,-...,.,.m-
~,,,,,.....~.,-w,,._,-rd,,~._.v._,-,,.
,,,_,.mmm..m.,m,c..-
m.u-
,,,_,.~.. ~
,,,,n,-,,
.m,,
,-.,~,.
4.3.3.2 (Cont'd)
Nominal spacing between modules is 2-in. so no interaction between modules as a consequence of SSE is considered.
4.3.3.3 The only applied loads to the module are the seismic loads. These were calculated to be at the top of the fuel support members (bottom of tube). The loads in the X, Y and Z direction occur simultaneously, Since the OBE loado are 4901 of SSE loads and the OBE stress allowables (with the exception of the buckling allowables) are 50% of SSE allowables, OBE is limiting.
4.3.3.4 Thermal stresses were calculated and found to be insignificant.
4.3.3.5 The fuel pool floor loading was re-analyzed by the plant architect-l engineer and found to be acceptabic per 4.3.1.4.
1 l
4.3.3.6 The eleven-node module with fixed base was modified and analysis
)
was performed of the module plus its support structure. Since the response of the module and column support system-is primarily rigid body motion, adequate representation of the system can be made by a 2-node lumped-l miss model. The lumped mass at the top was chosen to preserve the base shear force of the first node and the height' of the model was i
selected to preserve.he over-turning moment at the base for both first node and rigid body motion.*The stiffness was selected to preserve the fundamental f requency of the module and the support columns.
i b
The critical locations for maximum compression and shear stresses are at the base of the module in the areas near the support columns.
Tigure 4.3.1 shows the path of the shear forces from the fuel elements to the support columns. Based upon this structural behavior,, shear i
stresses in the fittings, fuel support plates and bottom tube elements were developed and compared to the allowable stresses (Table 4.3-3).
None of the allowables are exceeded for either OBE or SSE conditions.
l The mechanism for transferring shear forces to the pool slab is through friction resistance provided by the normal force due to the submerged veight of the module through its support columns resting on the pool floor liner. A minimum value of 0.31-for the coefficient of sliding friction for stainless steel to stainless steel was assumed in the analysis.
This value has been verified by recent tests of stainless steel materials.
(Ref: Babinow4ct. Ernest, " Friction Coef ficient Value For a High Density Tuel Storage System," report to General Electric Co., Nuclear Energy Programs Operations, 20 October, 1977). A value of 0.31 is suf ficient to ensure that sliding does not occur for earthquake motions correspondin,g to the OBE and SSE and provides a factor of safety for sliding and over-turning greater than 1.5 and 1.1 f vr the OBE and SSE, respectively. An additional non-linear analysis for sliding was performed to determine relative displacements if the coefficient of friction were less than 0,31.
These results are listed in lable 4.3-4.
l The seismic horizontal floor time history and response spectra are those developed in 4.3.1.4 and are shown in Figure 4.3-2.
l 23 j
Supplement 1 December, 1977 l
l 1
e Table 4.3 2 (Deleted)
TABLE 4.3-3 Comparison of Maximum Shear (psi) calculated To Allowable OBE SSE Calculated Allowable Calculated Allowable.
Cambit.ed FittinS -
Normal Stress 1,960 16,500 2,160 33,000 Shear Stress 1.125 11,000 1,480 22,000 Tube Local Shear at Bottom 496 11',000 650 22,000 Support Plate Weld Shear 100 11,000 130 22,000 TABLE 4.3-4 Sliding Analysis Displacements Coefficient of Friction Maximum Non-linear Sliding Displacements (in.)
0.10 0.49 0.15 0.23-0.20-0.13 0.25 0
0.30 0
Supplement 1' December, 1977 25 l
1
t
)
'P N
I
\\l I
\\l
- idea WALL l
N f* del ELEHapf x! !
Nl l
/w!l#
x f
N
[
\\
s Hon [wG
\\
NO cAs1 LT:
rdeL PL 845 50ProgT
@\\
eAce AsseHect LoNG c ASTIrvG-
@j%coLUM9
\\
FIGURE 4cs.1 PATH OF - SEISMIC HORIZONTAL FORCES IN MODULE Supplement 1 December, 1977 25a
1 i
3 4..a 6 a.
8.88 >
l [i..,s.,I g{
h 1Whd.bn1,.9c.........
i 1:
h
.. c. m,p!. g'.qpi, l
1
)
l
... u.
a i
i
8,-
.t i
i.
i.
1 v ar a seces:s 2
Horizantal Acceleration Time History l
4 5
i i
i
" T 4 Percent DampinD
.... l, i
g
,p Target Design Spectrum i
..p L l a
i
'"r Spectrum from horizontal Acceleration E..,, [
r Time History 5
..n L
e.
i ea s.n
- i. p it.
88 as
.w m.m si.
ea Fattutatt.sagatt Floor Response Spectra _
i FIGURE-4.3-2 SEISMIC HORIZONTAL FLOOR TIME HISTORY j'
AND RESPONSE SPECTRA COMPARISON Supplement 1-4 25b.
December, 1977
l
(
P 1
i
)
i i
1 i
e Response to 10/25/77 NRC Request for Additional Infomation i
on Monticello Replacement of Spent Fuel Pool Storage Racks 1
i e
i t
i i
k I
i Q.1 Provide the number of grams of uranium-235 per axial centimeter of fuel assembly that was used in your criticality calculations. We intend to incorporate this information as a Technical Specification limit on fuel assemblies that are to be placed in these high density storage racks.
A.1 The criticality calculations did not assume a limit on grams of uranium-235. The axial distribution of U-235 in a fuel assembly is not a fully meaningful specification for Monticello fuel. This fuel is manufactured with variable rod enrictenents and burnable poison, gadolinia, integral to the fuel. The reactivity of such fuel is a function of both enrichment (or axial U-235 distribution) and gado11nia content. For this reason it is deemed prudent to use a k-infinite limit for purposes of nuclear criticality safety in storage of the fuel.
If a Technical Specification limit is desired for the High Density Fuel Storage System, we would recommend k-infinite limit (k,, <1.35 based on a BWR lattice pitch at 20 C.).
Q.2 In Section 4.1.2 of your submittal, on the verification of the criti-cality calculations, you state there were solid boron control curtains in the Jersey Central experiments. How does the areal density of boron ten atoms between the fuel assemblics and the thicknesses of the water channels next to the boron plates in the Jersey Central experiments compare with those in the proposed storage racks? Please provide any experimental confirmation that you may have for the calculated neutron multiplication factors in a BWR fuel assembly lattice with a 6.5 inch pitch, with approximately 0.25 inch water gagbetween the fuel assembly and the boral plate, and with about 1.6 x 10 boron ten atoms per square centimeter between the fuel assemblies.
A.2 The areal densgy of boron-ten in the Jersey Cgtral control curtains is 0.00597 grams B/cm2 compared to 0.013 grams B/cm for the High Density Fuel Storage System. The nominal thickness of the water channel between the surface of the hrsey Central control curtain and the outside of the fuel channel is 0.378 cm. compared to a 1.15 cm. water gap from the inside of the HDFSS storage cell to the outside of the fuel channel (when present).
There has been no direct experimental confirmation of this system.
Such experiments for a system as substantially suberitical as this one are not necessary.
. _ ~ -
?
4 4
2 i
4 Q.3 In regard to Figure 4.1-1 of your submittal, what is the origin 4
of the water gaps shown in the four corners?
1 i
A.3 The origin of the water gaps in the simplified cell is the combination of the nominal gap between storage tubes and the maximum gap between perpendicular boral plates. For conservatism and simplicity, the stainless steel in the corners was neglected and replaced with moderator.
Q.4 Provide the nominal and minimum thicknesses of the Type 304 stain-less steel in the inner and outer storage tubes.
A.4 The 304 stainless steel tube walls have dimensions as listed in the table below.
304 SS Tube Walls (Ref ASTM A-240, A-480)
Dimensions in Inches Wall Nom Tolerance Minimum Outer
.090 1'.008
.082 Inner
.0355
+.004
.031
(.t Provide the nominal and minimum dimensions of the inner storage 4
tubes.
A.5 The minimum dimension of a fuel storage space is 6.05 inches square, projected for the full length of the storage space. The nominal fuel storage space dimension is 6.250 inches square for a tube storage location and 6.261 inches square for a non-tube location.
C.(
Provide the nominal density of boron carbide in the boral and the nominal thickness of the unciad boral sheets.
A. t-The Boral product used in the High Density Fuel Storage System consists of a 0.056 inch layer of B4C-Al matrix sandwiched between two g.010 inch aluminum sheets. The density of boron-10 is 0.013 grams /cm minimum, co r-responding to a boron carbide density of approximately 0.1 gm/cm2, i
07 Provide the change in k. for this high density storage lattice with j
a small change in uranium-235 enrichment.
A.7 A variation in enrichment has no significance in BWR fuel design as stated in reply to Question 1.
In addition to average bundle enrichment.
the k-infinite of a BWR bundle is dependent on geometry, enrichment distribution, gadolinia distribution, etc.
-4
. _.., ~,...
-,,... -... _ _. - _,... -. _ _,... ~.. -, _, -. _
3 Q.8 Provide the ange in k.
for this high density storage lattice with a small chang. in the areal density of the boron ten stoms between fuel assemblies.
A.B The sensitivity to boron concentration was determined from previous analyses of a sg/cmilagrackdesign. The change in cell k,,
from.010 to.015 grams B was approximately 2% Ak.
Q.9 Provide the L for a filled lattice of these storage tubes held in a close packed condition, i.e., the minimum pussible pitch.
A.9 The k.
of the HDTSS with fuel stored in the minimum possible pitch has not been determined. It has been shown, however, that the k, of the storage cells decreases with both decreasing moderator density and increasing pitch. Therefore, the nominal spacing of 6.563 inches is the pitch that gives the maximum k-infinite.
The dimensions of the materials constituting the model cell were maintained constant for the cases evaluated. Therefore, if a lesser pitch is considered, the result of moving the fuel closer together is to exclude moderator, or effectively decrease the moderator density. Table 6 of Table 4.1-1 of the Design Report shows that the result of decreasing moderator density is a decrease of reactivity (( ).
Thus, the k, of storage tubes with minimum possible pitch will be less than that of the nominal 6.563 in. pitch case.
Q.10 Provide the amount of the in:reased pitch in case 4 of Table 4.1-1.
A.10 The pitch in Case 4 of Table 4.1-1 is 6.832".
Q.11 In Section 4.4 of your submittal, you stated that a dimensional in-spection of the neutron absorber plate locations vill be performed at the pool site. Describe in detail how you propose to use this and other tests to show that there vill be a suf ficient number of boral plates in the racks with a suf ficient amount of boron ten isotope in the plates to maintain the k,gg 10.95.
- 'l To verify thattherevillbeasufficientnumberofbgalplatesin the fuel storage modules with a st.f ficient amount of B isotope in the plates to maintain the k. 3 95, the following program has been developed.
0 The boron carbide used in the Boral sheets is certified as to its B10 iso.
topic content.
Scmp'.es of each Boral sheet are chemically analyzed to determine the boron content. These data are statistically evaluated such that the samples are representatig of the entire area of the Beral plate.
It is verified that the minimum B content, at a 95% confidence level, meets or exceeds specification requirengts. Analyses are performed to establish the correlation between the B content and the thickness of the Boral sample. The Boral sheets are dimensionally inspected and the thickness data are statistically analyzed to verify the sheet meets the minimum thickness requirement over its entire area at a 95% confidence level. These thickness data are comparg with the correlation data to provide additional assurance that the B content meets or exceeds specification requirements. The Boral is inserted into a tube assembly
e
-4 only af ter it has been verified that each of the above inspections and evaluations has been successfully performed. The Boral plates are placed between the inner and outer walls of the tube assembly, The assembled storage tube is hydro-f ormed, " locking" the Boral plates in place. The inner and outer walls are welded together at each end of the tube encapsulating the Boral. It is then not possible to remove the Boral without destroying the tube. The thickness of the storage tube vall is measured after tube assembly. These data are statistically analyzed and the entire tube assembly population, known te contain Boral, is uniquely identified by the average thickness (mean) and the standard deviation.
Presence of the neutron absorber material in the fabricated fuel storage module is verified by visual examination and dimensional inspection.
l The thickness of the Boral plate is different than commercially available i
alumunim or SSt sheets. Materials of standard thickness used in place of Boral vould be detected by the significant dif f erence in wall thickness censurements for tube vr.11s which contain Boral as opposed to tube walls which contain non-Boral materials.
The thickness of the walls of the module assembly will be measured at the reactor site before installation. These data vill be statistically analyzed such that the individual fuel storage module is uniquely identified by the mean and the standard deviation. The two sets of wall thickness data vill be statistically compared to determine, at a 95%
confidence level, if there is a significant difference between the individual tube wall thickness data and the module assembly data. If a significant dif f erence does not exist it indicates the module was cade f rom tubes known to contain neutron absorbing material and the moduit will be accepted as containing the required amount of Boral plates.
Q.12 Describe the procedures that will be used to remove the present racks and install the new ones.
Specifically, discuss how you will preclude the possibility of dropping or tipping a rack onto the spent fuel in the pool.
' 12 The spent fuel storage modification at Monticello will be carried out in tra or more phasec. The firct phase entails placement of new spent fuel racks (hereafter called modules) I through 4 in locations shown on the attached Figure 1.
Placement of modules 5 through 13 vill occur as a later phase or phases as modules become available from the manufacturer.
Presently the spent fuel pool at Monticello contains 37 standard General Electric spent fuel racks (hereafter referred to as racks) with a capa-city of 20 spent fuel assemblies each. A total of 616 spent fuel assemblies are stored in 31 of these racks. Also the pool contains two racks for defective fuel canisters or control rod blades, each rack capable of holding up to ten canisters or control blades, as needed.
One rack is designated as a control blade rack the other as a defective-fuel rack, although fuel has never been damaged to the extent that placing it in a scaled canister has been necessary. The existing pool layout is shown on Figure 2, attached.
y l
During pool preparations and module installatica the following general 1
guidelines will be adhered to:
3 A.
All racks will be emptied of their contents prior to unbolting and moving.
B.
Racks and modules will be kept low in the pool during horizontal movement. They will not be moved over spent fuel.
C.
Vertical movanent of racks and modules in and out of the pool will be done in designated areas. All racks removed from the pool in preparation for the first phase will be lifted from the southeast corner of the pool. Modules 1 through 4 will be lowered into the southeast corner of the pool. All other racks removed from the pool will be lifted fram the northwest corner of the pool.
Mbdules 4
5 through 13 will be lowered into the northwest corner of the pool.
D.
Racks will be washed down to remove contamination as they are lif ted out of the pool water.
E.
Racks or modules which are out of the fuel storage pool shall not be moved within 12 feet of the fuel pool wall except when in transit in or out of the pool in the manner specified in iten C above.
F.
New modules will be handled with the modified reactor building crane main hoist with a special lifting fixture, both of which are designed with a high safety factor.
The specific procedure currently being considered for pool preparations and module installation, which meets the above criteria, is as follows:
1.
Move rack 24 (See Figure 2) to the south end of the pool.
2.
Remove the defective-fuel rack from the pool.
3.
Move rack 24 into the location vacated by the defective-fuct rack.
4.
} ove rack 31 into the location vacated by rack 24, 5.
Remove the work table from the southeast corner of the pool.
6.
Remove racks 29, 30, 32, 33, 34 and 35. After removal from the water, move racks directly to the south away fror the pool.
7.
Install modules 2, 3, 4 and 1 respectively, approaching the fuel pool fram the south and lowering the modules into the southeast corner of the pool.
S.
Move fuel from racks 1. through 25 and 28 into the new storage modules.
(Approximately 737 assemblies will be stored in modules 1 through 4 and racks 36, 37, 26 and 27 after the Fall 1978 refueling outage.)
er r
P 4
l l l
9.
Remove racks 1 through 25 and rack 28 via the northwest corner J
of the pool. Af ter removal from the water, move racks directly j
to the west atray from the pool.
10.
Install module 7
- 11. Move fuel from racks 26 and 27 to southeast locations of module 7.
12.
Remove racks 26 and 27.
j 13.
Install modules 6, 5, 8, 9, 10, 13, 12 and 11, respectively, approaching the fuel pool from the west and lowering the modules -
into the northwest corner of the pool.
1 1
l 1
4 4
!'h 1
a i
'I i
e 4
3 W
i
~.. _. _,
,__.a_.__..,
2 G
I _)
I.
I
.d.!
6,.
. lt I o
e.
,n t.
p...
is ll 12.
l3
-~
)
')
L/
_ _.= _._.
i l
8 9
-~
i i
__}
l 6 i
i i
l!
l]
~ ", '
f,1 i.,
.I
,V
, m, t
i >_
Lt l
,c.
i l
i l
l 2
3 4
i i I
_M 6MIPPIN4 (A6K I
N'
- ' 7
?T02 A4 'i Al?E A I i
4 f
i (3--
3 (ONTRol IR/-sDE l?At K l
L
- 4
.w 1
~
. c. 3...... c.
3.
Nj-
-- T r"- r
,j t I'
...r
25 I
2 3
4 F
6 7
C' ~ ~
26 1
==
L t
)
I 8
9 10 11 12 13 14 15' c1 M
b\\
g7 A
G.. o
.=
=
u a
)il
- ~ '
j lb 17 16 19 20 21 21 23 2.4 til N
0 g'~
[
e i
(NI
}, )
til
" *i. h -
h.
.i til
- ' ' ~ ~
'il n
ro
.,s.
.i'.1}
3 7
g L2.
~
i
./ N I
./
b' t
-. =
a h*,
4 I
I N
j 32 33 3W 3F a
4 5
- M bl '
3o 37 sniei>>w4 <44x 8
6T02A4E AGE A s
i.
tt
.m_.,..__>
RAcv
-.