ML20039B220
| ML20039B220 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 12/16/1981 |
| From: | Withers B PORTLAND GENERAL ELECTRIC CO. |
| To: | Schwencer A Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20039B221 | List: |
| References | |
| TAC-03606, TAC-3606, NUDOCS 8112220433 | |
| Download: ML20039B220 (3) | |
Text
_
i s
s Portland Cewal ElectricCompany
- !s
~
's uwEma S.
sano vaes vce neser Z
DEC 211981t> c Q cs=n=nuw e F3 December 16, I',8 "9j# /D M I "'(\\(?
Trojan Nuclear P ant Docket 50-344 License NPF-1 Director of Nuclear Reactor Regulation ATTN:
Mr. A. Schwencer, Chief Operating Reactor Branch 1 Division of Operating Reactors U. S. Nuclear Regulatory Commission Washington, DC 20555
Dear Mr. Schwencer:
Transmitted herewith are 40 copies of Revision 3 to PGE-1013, Spent Fuel Pool Storage Rack Design Report. Please have the revision inserted into your copies of this document according to the instructions enclosed.
PGE-1013 was originally transmitted to you on January 6, 1977, as an attachment to License Change Request 19.
Revision 1 and Revision 2 were transmitted to you on April 6,1977 and August 29, 1977, respec-tively. Revision 3 is editorial, incorporating into the text responses to requests for additional information concerning the design as well as the amplifying information presented during the proceedings that resulted in the approval of Amendment 34.
Sincerely, g
Bart D. Withers Vice President Nuclear Enclosure c:
Mr. Lynn Frank, Director Oregon Department of Energy -
(w/ encl)
$ck2gggfoj$
P 121 SW kmon Street Poaard, Oregon 97204
The following information and checklist are furnished as a guide for the insertion of new sheets for Revision 3 into PCE-1013, Spent Fuel Pool Storage Rack Design Report for the Trojan Nuclear Plant. This material is denoted by use of the revision number and date in the lower outside corner of the page. This instruction sheet should be inserted immedi-ately following the front cover. New sheets should be inserted as listed below:
Discard Old Sheet Insert New Sheet Title Page Title Page i
i 11 11 iv iv vi vi Section 1.0 1-1 1-1 1-2 Section 2.0 2-1 2-1 thru 2-3 Section 3.0 3-1 3-1 thru thru 3-23 3-29 i
Table 3-6 Table 3-6 Table 3-7 Table 3-7 i
Section 4.0 4-1
'4-1 thru thru 4-9 4-11 8112220436 81121'6 PDR ADOCK 05000344 i
P PDR Revision 3 l
(December 1981)
Discard Old Sheet Insert New Sheet Section 5.0 5-17 5-17 thru 5-19 Table 5-14 Table 5-15 Section 7.0 7-1 7-1 7-2 7-2 Appendix A A-1 A-1 Table A-1 (Sheet 1) thru thru A-17 Table A-6 (Sheet 2)
Apps. dix B B-1 B-1 thru thru B-4 B-7 Table B-1 Table B-2 Figure B >
Figure B-10 l
i l
Appendix C l
l C-1 C-1 l
thru C-3 l
l l
1 O
Revision 3 (December 1981)
PCE-1013 Revision 3
. December 1981 SPENT FUEL POOL STORAGE RACK DESIGN REPORT FOR THE TROJAN *.DitEAR PLANT 6
December 1976 Portland General Electric Company 121 S. W. Salmon Street Portland, Oregon 97204 0
SPENT FUEL POOL STORAGE RACK DESIGN REPORT FOR THE TROJAN NUCLEAR PLANT CONTENTS Section Title Page 1.0 Introduction...................
1-1 2.0 Summary..................
2-1 3.0 Spent Fuel Storage.
3-1 3.1 Description and Design Bases.
3-1 3.1.1 Structural and Thermal Expansion Analyses.
m 3-3 10 3.1.2 Criticality Analysis.
3-10 3.1.2.1 Analysis Overview for Delayed Modification.
3-16 3.1.3 Seismic Analysis.
3-17 3.1.4 SFP Liner Plate and Fuel Clad Integrity.
3-19 3.2 Spent Fuel Pool Cooling and Demineralizer System.
3-21 0
3.2.1 resign Bases.
3-21 3.2.2 Design Description.
J-22 3.2.3 System Description and Operating Modes.
3-24 3.2.4 Instrumentation.
3-28 4.0 Safety Evaluation.
4-1
/^
4.1 Spent Fuel Pool and Spent Fuel Pool Rack Safety Evaluation.
4-1 1
4.2 Spent Fuel Pool Cooling and Demineralizer System Safety Evaluation.
4-7 I,S 5.0 Radiological Evaluation.
5-1 L
5.1 Source terms.
5-1 5.1.1 Activities in Spent Fuel.
5-1 5.1.2 Activities in Reactor Coolant and Refueling Water 5-2 5.1.3 Tritium.
5-6 l
5.1.4 Activities in Spent Fuel Pool Cooling and g
l Demineralizer System.
5-7 l
5.1.5 Activities in Ventilation Air.
5-8 5.1.6 Environmental Releases.
5-9 5.2 Radiation Doses.
5-12
(
5.2.1 Doses to Plant Personnel.............
5-12 5.2.1.1 Direct Radiation Dose From Spent Fuel Assemblies 5-12 5.2.1.2 Direct Radiation Dose From Activity in Water.
5-13 O
i 5.2.1.3 Dose From Airborne Isotopes.
5-13 5.2.1.4 Miscellaneous Sources of Exposure.
5-15 5.2.1.5 Plant Man-Rem Doses.
5-16 5.2.1.6 SFP Modification Exposure.............
5-17 5.2.2 Site Boundary Doses.
5-18 5.3 Disposition of Existing Racks.
5-18 6.0 Need for New Spent Fuel Storage Racks.
6-1 v
7.0 Tests and Inspection.
7-1 i
Revision 3 i
(December 1981) i
t SPENT FUEL POOL STORAGE RACK i
DESIGN REPORT FOR THE TROJAN NUCLEAR PLANT CONTENTS Section Title Page 8.0 References.
8-1 j
Appendix A Responses to Requests for Additional Information.
A-1 Appendix B Summary of Models.................
B-1 h
Appendix C Supplemental Seismic Analysis of 7 x 8 Rack Module.
C-1 gg i
i
}
i i
O 11 Revision 3 (December 1981)
,,., - -.... ~,. - - -.. -. _. - - - - - - - _ -.. _. -. -. -. - -... _,
b N
SPENT FUEL POOL STORAGE RACK DESIGN REPORT FOR THE TROJAN NUCLEAR PLANT TABLES Table Title I
5-8a Expected Change in Release Rates of Radioactive Gaseous g
Effluents from Spent Fuel Area (Ci/yr) v l
5-9 Dose Rates at Pool Surface and Edges From Isotopes Contained in Refueling Water During 8th Refueling (ares /hr) 5-10 Inplant Dose Rates From Airborne Isotopes During 8th Refueling (arem/hr) g v
5-11 Refueling Manpower Requirements (3-1/3-Core Storage) 5-11a Refueling Manpower Requirements (1-1/3-Core Storage) 5-12 Maximum Site Boundary Doses From Refueling and Fuel Storage Operations (area /hr) 3 5-13 2xpected Change in Site Boundary Doses From Spent Fuel Area g
(ares /hr) v 4
5-14 Factor for Inplant Radiation Exposure Estimates for Spent Fuel Storage Rack Work 5-15 Radiation Exposure Accumulated During SFP Modification (mrem) 8 B-1 Ratio of Allowable Stress to Fuel Assembly Component Maximum Stresses for the Uniform Loading Case at 400 Lb B-2 Ratio of Allowable Stress to Fuel Assembly Component Maximum Stresses for Triangular Loading Case at 850 and 1090 Lb i
C-1 Combined Stresses for 7 x 8 Module - SRSS Method (First 8 Modes) i lO 1
iv Revision 3 (December 1981)
I.-...
- _, -,. ~,..
SPENT FUEL POOL STORAGE RACK DESIGN REPORT FOR THE TROJAN NUCLEAR PLANT FIGURES Figure Title Ib B-1 General and Typical Module B-2 6 x 8 Module B-3 6 x 7 Module B-4 7 x 7 Module O
B-5 7 x 8 Module v
B-6 Computer Model, 2 Module Floor Truss B-7 Computer Model, 1 Module Floor Truss B-8 Rack Module Model Detail B-9 Fuel Assembly Deflected Shape 400 Lb Reaction Load w
B-10 Fuel Assembly Deflected Shapes for Triangular Load Distribution l
l r
i 1
vi Revision 3 (December 1981)
. _... -.. _. _. _. - _ _, - _ -,. ~..
[~ ]/
1.0 INTRODUCTION
x_,
This report is submitted in support of Portland General Electric Company's (PGE) Amendment 34 of the Trojan Nuclear Plant Operating h
License, NPF-1, to more fully utilize the storage capacity of the spent fuel storage facility.
The modifications to the spent fuel storage facility allowed increased flexibility in scheduling offsite shipment of spent fuel. The modifi-cations were accomplished without significant changes to the facility by replacing existing spent fuel racks with racks of a high-density
-w design.
Section 2.0 is a summary. Section 3.0 is the description, design bases and supporting analyses for the modification. Section 4.0 and Section 5.0 are respectively the safety and radiological evaluations.
Section 6.0 explains the conditions that necessitated the modification.
(']
Section 7.0 includes the tests and inspections involved in the fabri-cation and installation of the racks. Section 8.0 lists references.
3 Appendix A includes a summary of NRC requests for information and
-u references PGE documentation of compliance on each of the issues.
Appendix B is a summary of the models used to analyze the racks.
Appendix C is supplemental detailed information demonstrating compli-
]
ance with Regulatory Guide 1.92 and NRC Standard Review Plan, Section 3.8.4.
Revision 3, being af ter the fact, incorporates into this text PGE commitments made during the hearings that eventually resulted in the approval of Anendment 34.
Throughout this text, " existing" denotes b
status prior to the expansion modification, whereas "new" implies po st-modifica tion.
The design of the spent fuel racks was performed by the prime contractor j'S Programmed and Remote Systems Corporation (par), St. Paul, Minnesota.
\\
)
v The static and seismic structural analyses were performed by 1-1 Revision 3 (December 1981)
1 i
i i
i t
- 1..
Environmental Services Irc. (ESI), Mir.'eapolis, !!innesota. The criti-cality and heat transfer analyses were performed by Nuclear Associates J
international Corporation (NAI), Rockville,liaryland. Bechtel Power i
4.
Corporation,' San Francisco, California, verified the compatibility of the Spent Fuel Pool (SFP) Cooling and Demineralizer System with.the new SFP design temperature limit. The radiological. evaluation was performed by PCE.
1 4
t t
1 1
i i
i I
i i
i i
i
\\
i i.
i i
4 t
t l
1-2 Revision 3 (December 1981) w e rev
,,,-----...,,,w e e,mv s e n.,-
w, ~-, m,,
.m,_
,,m
.,e,--
n-,,.,,,,,,
~.,.e,,n,
.,,,.v.,,mw.,,.,,.
...,.,,,,,,w
i 2.0
SUMMARY
,b The modification provides safe storage for up to 651 spent fuel assem-blies in the SFP and included replacing the existing spent fuel racks with those supplied by par. The modifications did not alter the structure of the SFP or the supporting cooling systems.
The new spent fuel rack design consists of arrays of stainless steel to form cavities having a nominal center-to-center spacing of 13.3 in. and a wall thickness of 3/16 in.
The preparatory work took place in two phases.
Phase 1 was scheduled to be completed near the end of October 1977.
In detail, Phase 1 included the following work:
- 1) To facilitate rack removal and reinstallation, existing spent fuel rack leveling shims which were tack-welded originally to the SFP liner plate were ground loose and tack-welded to the existing spent fuel racks. This action permitted the existing racks to be easily removed and
-~s reinstalled.
- 2) To facilitate this work required removal of the existing spent fuel racks a few at a time. Sufficient rack capacity was being retained in the SFP during this phase to permit discharge of a full core.
- 3) The installation of the duct heater in the SFP ventila-tion exhaust system described in Section 3.2.2 and the reorientation of the SFP Cooling and Purification System diffuser header to improve circulation of the cooling water in the SFP.
Phase two, scheduled to commence in November 1977 involved the follow-ing work:
- 1) A sufficient number of the existing spent fuel storage (s\\
(,)
racks were removed from the SFP to provide adequate room 2-1 Revision 3 (December 1981) i
m for work described below. As many of the existing spent fuel storage racks as possible were retained in the SFP.
Sufficient existing spent fuel storage racks could be replaced before the Trojan Technical Specifi-cations permit fuel to be moved into SFP (100 hr).
- 2) A sleeve was field-welded around each embedment stud.
Since this sleeve did not extend above the SFP liner plate, it did not interfere with the reinstallation of the existing spent fuel storage racks. The welds were inspected.
- 3) The alignment sleeves and embedment modules locating frames were located over the embedment studs. The embedment module locating frames were welded to the align-neat sleeves. Note at any point these components could be removed leaving the embedment stud ready to receive the existing spent fuel storage racks.
J c
v
- 4) The module support cups were welded to the embedment module locating frames after they were located using a template and the new rack modules. The welds and tolerances were inspected.
- 5) The new rack modules were positioned in the SFP.
- 6) The new rack modules were leveled, and the top module-to-module was fit and welded.
- 7) The new rack modules and embedment module locating frames were removed and the existing spent fuel storage racks reinstalled.
The design description is comprised of structural, criticality, th'ermal
-e and seismic analyses that conform to applicable NRC regulations, NRC guides, and industry standards. The design is supported by a safety 2-2 Revision 3 (December 1931)
j
(
and radiological evaluation and an assessment of the need and schedule for installation of new spent fuel storage racks at Trojan.
C3 4
v
.The 1974 edition of ASME Section III and all applicable addenda up to the date of the purchase order between PGE and par (January 29, 1976),
were used for the design.
i-The design ensures that an effective multiplication factor (keff) Of 1ess than 0.95 will be maintained; that. adequate cooling during postu-lated normal and special operating conditions will be provided; and l
that the structure will withstand safe shutdown earthquake (SSE) g loadings. The modifications for the Trojan Nuclear Plant provide safe storage for up to 651 fuel assemblies and are compatible with the plant design as provided in the Trojan Final Safety Analysis Report (FSAR) and Operating License (NPF-1).
I i
i 1
i
)
t U
2-3 Revision 3 (December 1981)
3.0 SPENT FUEL STORAGE The design basis for the spent fuel storage facility has been changed f rom the storage of four regions to the storage of 10 regions described lb in this report. The operating temperature design limit of the SFP water is increased from 125'F to the design limit described in this report of 140*F.
lb 3.1 DESCRIPTION AND DESIGN BASES The SFP is a reinforced concrete structure with seam-welded stainless stcel plate liners enclosing a pool volume of approximately 51,900 cu f t.
The reinforced structure of the SFP is designed in accordance with codes descrfbed in FSAR Section 3.8.1.
The SFP structure and spent fuel racks are designed in accordance with Seismic Category I require-ments. Gaseous radioactivity in the area of the SFP is designed to be maintained below the 10 CFR 20 limits.
f~h V) t The SFP provides a safe and reliable means of storage and facilitates handling of irradiated fuel. No equipment or materials other than g
spent fuel components and associated tools will be stored in the SFP.
The safety function of the SFP is to retain water and passively support spent fuel assemblies in a position amenable to natural circulation cooling. The SFP is designed to accommodate 10 regions (plus one spare cavity) of spent fuel in a suberitical array such that a keff,$0.95 is j
j maintained.
Accommodation in the SFP for 10 regions or 651 spent fuel assemblies allows the concurrent storage of one full core (193 asser blies), the spent fuel assemblies from seven normal refuelings (65 assemblies each),
and one spare cavity. The spent fuel assemblies are stored in cavities in parallel rows and have a center-to-center distance of 13.3 in. in both horizontal directions. Burnable poison rods removed from the reactor are stored in spent fuel assemblies.
O k-)
3-1 Revision 3 (December 1981)
. =- -
.. - - -.. -.-.= _ -
t The general arrangement of the storage space is illustrated in
}
Figure 3-1.
The embedment frame arrangement is shown in Figures 3-2 g
l through 3-21; details of Figures 3-1 and 3-2 are shown in Figure 3-3.
A typical module arrangement is shown in Figure 3-4.
Figure 3-3a was submitted in anticipation of underwater installation. The decision to use the details in Figure 3-3a was made in September 1977, when the
^
licensing and refueling schedules indicated that the rack installation might have to be performed underwater.
i The water level in the SFP is maintained to a normal value of 91 ft 8 in.
This provides at least 23 ft of water above the top of a spent fuel assembly in the storage racks and at least 10 f t above the actual fuel rods during fuel trau.sfer operations. This water barrier serves as a radiation shield, enabling the gamma dose rate at the pool surface to be maintained at or below 2.5 ar/hr.
A drainage system is below the liner plates of the SFP, cask loading pit Ib l
and fuel transfer canal. Mancal diaphragm valves-are left opened to Ib m
monitor drainage into a manifold connected to the Dirty Radioactive Waste 10, l
System. The drainage system will be routinely inspected for leakage-g I
during shift tours.
Adjacent-to the SFP are two zaaller pools - the fuel transfer canal and the cask loading pit. The fuel transfer canal is connected by the fuel transfer tube to the Containment refueling cavity. A leaktight door is provided between the SFP and the fuel transfer canal. The cask loading pit is connected to the SFP by another leaktight door on the opposite side of the SFP from the fuel transfer canal.
The gaseous radioactivity from the atmosphere above the SFP is control-led by a ventilation system. The SFP Vent Monitoring System (PRM-3) continuously monitors noble gases in the vents from the SFP areas and alaras if the gaseous airborne radioactivity level reaches a preset b
limit. This system is described in Trojan Nuclear Plant FSAR Section 11.4.2.2.2.
3-2 Revision 3 (December 1981)
(N After passing through HEPA and charcoal filters, the exhaust from the SFP is combined with the Fuel and Auxiliary Building Exbstsat System.
This combined flow is continuously monitored by the Auxiliary Building
{}
Vent Exhaust Monitoring System (PRM-2) for noble gases, iodine and par-ticulates. This system is described in Trojan FSAR Section 11.4.2.2.2.
The SFP Area Radiation Monitoring System is provided for personnel pro-tection and general surveillance of the SFP area. Equipment in the control room provides continuous monitoring, recording and alarms.
Audible and visual indicators are provided locally. Potentially radio-active crud at the bottom of the SFP can be removed by an underwater portable vacuum cleaner.
72 The importance of the SFP as a potential sabotage target has been recog-nized by PCE in developing the Trojan security plan. PCE has identified the SFP as a vital area. To elaborate on security measures refer to PGE's most recent security plan.
p.
3.1.1 STRUCTURAL AND THERMAL EXPANSION ANALYSES The cavities in the SFP are on 13.3-in. nominal center-to-center spacir.g and are welded into the following modules, arranged as illustrated in Figure 3-1.
l l
Module Module Total Size Quantity Cavities (Ibs)
(lbs) 6x8 7
336 20,400 142,800 6x7 5
210 17,850 69,250
{3 7x7 1
49 20,825 20,825 7x8
_ l, 56 23,800 23,800 14 651 82,875 276,675 I
The new racks occupy the same exterior space envelope in the plan view as the existing racks.
O
\\~ /
3-3 Revision 3 (December 1981)
The racks and their interfacing structures to the existing SFP floor V
embedments are constructed almost entirely of Type 304 stainless steel.
The module threaded feet and top tie bolts are ARMCO Type 17-4PH stain-b less steel heat-treated to condition H-1100. These are prec!.pitation-hardened stainless steel because of the gallin8 and higher strength requirements. Material properties for ARMCO Type 17-4PH were taken from the ARMCO Steel Corporation, Advanced Materials Division Publication, and have been verified to be the same as established in the ASME Code Section II; ie, the ARMCO Type 17-4PH minimum yield values used are the same as ASME SA 564 for Type 630 steel minimum yield values. With b
regard to the NRC comments concerning stress corrosion characteristics of Type 17-4PH for heat treatments below H-1100 series, the vendor's proposed use of heat treatment H-900 on the top tie-bolts was not per-mitted by PGE, and all Type 17-4PH material remained at H-1100.
This ic compatible with with the pool liner plate and embedments of Type 304 7
stainless steel.
O The analysis of structural loads imposed by dynamic, static, seismic and thermal forces is m'ade in accordance with NRC Standard Review Plan, g
Section 3.8.4, and is discussed in Section 3.1.3 of this report. The combined stress ratios are less than the allowable interaction factor, therefore the corresponding safety margin is at least the difference b
between the allowable and the yield stress.
Allowable stress limits for combined loading conditions are in accor-dance with Section III, Appendix I, and Appendix XVII of the ASME Boiler & Pressure Vessel Code. Allowable stress limits for linear com-ponent support welds are defined in Iable hT-3292.1-1 of Section III, Subsection hT of the Code. All materials in these calculations are based on Type 304 stainless steel manufactured to ASTM specifications
]
of A-240 (sheets and plates), A-276 (angles and flats) and A-312 (pipe).
- 1) Minimum Yield Strength "F"
[see Table 1-2.2, " Yield Strength of Austenitic Steel", Appendix I, ASME P & V
(~T Code,Section III, and see ASTM-A240 specification):
Q 3-4 Revision 3 (December 1981)
-.~
7.s F - 30,000 psi at 0.2 percent permanent set at 100*F.
(v) y F = 25,000 psi at 0.2 percent permanent set at 200*F.
y
- 2) Minimum Ultimate Strength "F ~8 u
F = 75,000 psi at 40 percent minimum elon3ation.
u O
v
- 3) Modulus of elasticity "E" (Type 304 stainless steel)
(see Table 1-6.0, " Module of Elasticity of Material",
Appendix I, ASME P & V Code, Section IIIj:
E = 28.3 (10 ) psi at 70*F.
6 6
E = 27.7 (10 ) psi at 200*F.
The individual cavities of a module are welded as an open-ended box
[
section, 8.96 in. sq by 14 f t 2-1/2 in. long with a 3/16-in. wall thick-Each cavity has a welded cruciform (see Figure 3-2j) in the bot-ness.
tom end to support the fuel bundles. The cavity length is such that the top of the fuel assembly plus a rod cluster control is 3-3/8 in. below the cavity funnel top.
The cavities are first welded into columns of six or seven (called 1 x 6 or 1 x 7 arrays) as shown in Figure 3-4.
The top funnel structure, the lower horizontal channels and the "X" bracing The provide t'c.a structural integrity for interconnecting the cavities.
arrays are then welded into seven or eight rows to make up the modules.
The side channels and "X" bracing are the interconnecting members; The adjacent funnels are welded to form shear ties between the arrays.
end cavities of the end row array have welded supports in the bottom of the cavities which form the four-corner supports for the module. Each support has a threaded pad which is used to level the module by means of a long-handled tool reaching down the centerline of the corner cavi-ties. The corner supports raise the racks 11-7/8 in. above the SFP for water floor, thereby providing an unimpeded passage of this height
)
circulation under the racks.
v 3-5 Revision 3 (December 1981)
g3 (a)
The modules are supported in the SFP by the existing anchor bolt embed-ments and by tie plates between the tops of the modules. The loads, which are mainly due to seismic forces, are transmitted from the sup-ports of the modules to the embedment anchor bolts by stainless steel locating frames as shown in Figures 3-2 through 3-21.
The frames are designed to evenly distribute the loads from the supports to the anchor bolts and climinate, insofar as practical, increased shear due to moment couples on the bolt groups which would be inherent without the frames.
The stresses in the locating f rames are shown in Table 3-4.
The modules are designed to be installed in the dry pool. A sleeve is field-welded around each stud to increase its shear capacity to accommodate the cal-culated loads; the existing embedments are adequate for the calculated loads. The locating frames are placed over the studs and alignment sleeves field-welded to the frame plates after precisely aligning the sleeves relative to the studs. The cups for the module supports are field-welded to the frame as shown in Figure 3-3.
Special lifting and welding fixtures are provided to assure safe handling and precise align-ment.
The ties between the tops of the modules shown in Figure 3-1 (v)
(detail B in Figure 3-3) stabilize the modules against seismic forces and maintain the top spacing for criticality considerations.
The differences in thermal expansion between the pool floor and the modules are accommodated in the interface (stud to plate sleeve) between the locating frames and the embedments so that there is no significant shear on the embedment for the case of thermal expansion over the 75'F bh to 212 *F operating temperature range. The total clearances provide for the dif ferential thermal growth between the racks and the floor. There-
[3 fore, no significant induced horizonts? thermal loads are on either the rack or the embedment studs. Rack temperature gradients horizontally due to a full rack next to an empty rack are not significant. Also, as described above, any resultin ; thermal growth in this direction is fh unconfined and no stresses result.
The largest vertical temperature gradient through the water inside a n
(
)
cavity tube, as calculated f rom the thermal analysis, is 23*F.
If it is
~
assumed that the cavity tube has this same temperature distribution 3-6 Revision 3 (December 1981)
(ignoring external water surrounding the tube which reduces this gra-dient), and if one tube is heated to a mean temperature produced by the vertical temperature gradient and all other surrounding tubes are not (confining the heated tube), the resulting thermal axial stress in this O
tube is approxiuately 3100 psi, which gives an axial interaction ratio of 0.19.
When this maximum thermal stress interaction ratio is com-bined with those due to dead, live, seismic, and fuel bundle impact, the t'otal interaction ratio on the most highly stressed cavity tube is 0.88, which is less than 1,30 and therefore acceptable.
The fuel pool 1/4-in. thick stainless steel liner was initially designed and functions as a leaktight barrier, and is not a fuel pool structural element. As described, the liner is capable of resisting the maximum accident temperature conditions of 212*F without compromising leak tight-Although no code load combinations with respective stress limits ness.
apply directly in this case, the fuel pool liner has been reevaluated for the governing load combination of dead, live (hydraulic), seismic g
(SSE) and new thermal loads (140*F). Under this load combination, mem-brane stresses remain below the code value for structural elements of 1.6 S per Standard Review Plan Section 3.8.4.
Similarly, combined stresses for the reinforced concrete spent fuel pool walls remain within allowable limits for load combinations which include the increased nominal design temperature of 140*F.
The cups which horizontally restrain the bottom module support have a 0.015-in. radial clearance. Each sleeve has an outside and inside diam-eter respectively of 1.75 in. and 1.062 in.
Each stud has an outside diameter of 1.0 in.
The locating frames which hold the cups have a
^
0.062-in. ;adial clearance with the' sleeved embedment studs. These 3
clearances provide thermal expansion space for the modus frames.
Alternate details for underwater installation, using the same clearances for shear transfer from the module cur.= to the existing embeds, are shown in Figure 3-3a.
3-7 Revision 3 (December 1981)
l
\\
The tops of adjacent modules are clamped together with shear bolts which have c. 0.008-in. radial clearance. The shear bolts are 3/4-in. bolts.
Under thermal expansion from 70*F to 212*F, the top of the end module moves further than the bottom because the bottom is restrained by the g
embedment; this results in a 0.012-in. deflectic,n of the top of the module. This deflection is approximately 1/10 of the deflection exper enced under seismic loading and does not significantly increase modulo loading.
Finite element models used to determine module t.eriber stress for static dead loading and seismic response conditions are described in Section 3.1.3.
Model assumptions are that the floor trusses sre ideal planar frames composed of flexural beam - column elements, that module.
~
feet reactions are applied at the cup location and that the embi4ments m
are pinned to the floor. The model was analyzed using the SAGS com-puter code. Member stresses and embedment bolt stresses were compared L
against allowable limits as set forth in Section 3.1.3 of this report.
3
,a The rack configuration and tolerances are designed to the requirements
. imposed by the seismic, criticality, thermal, hydraulic and interfacing considerations. For example, a bar is provided in the design on the out,.
,l side modules to preclude a drapped assembly from locating closer than i
6 in. to the nearest stored fuel assembly. This dropped assembly condi-tion is discussed in Section 4.0 of this report.
The design also accommodates an impact resulting from a 2000-lb rigid object in the size and shape of a fuel assembly dropped from the maximum crane hook height above the rack funnel (10-in. drop). A prototype module was constructed consisting of four cavities connected to form a i
module which is representative of a cluster of cavities of the actual module design. An impact plate 8 in. sq was anchored to the bottom of a test weight to simulate the bottom fitting of a fuel bundle. The test weight was suspended over the prototype module and a guiding structure was constructed around the module so that the impact plate would strike
[J
\\
the module at the funnel. intersection.of the four cavities. Three load cells were mounted on the module base and connected to an oscillograph 3-8 Revision 3 (December 1981)
" ~
m i,,
r -
- (
^$
W
~
-j
-to record the force-time response during impact. Vertical drop tests were then conducted (in dry conditions) to determine the effects of the 29,000-in.-lb impact energy on the prototype module, and to provide l
data' for further analyses. Observations and measurements after full-sc)1e test.ing showed only very localized deformation at the funnel top (on-thefderof1/16in.),whichwouldnotaffect fuel bundle inser-ition or teinoval. The recorded impact time history was a triangular x
N-
~
'inpulse lasting 0.006 sec. The equivalent static load was calculated, f
s, E-applied - t'o the center of the largest full module in the computer model,
. and the'resulting member stresses were calce: lated. For this condition
~
all structural members except the impact interface were below 90 percent 7
of yield.
=,
\\
Under the condition where th'e. dropped fuel bundle does not strike the top of the rack 'and tr'avels unimpeded down through the cavity, the fuel
~
bundle will strike the%ttom fuel support cruciform. The cruciform is designed to shear-out under this impact energy such that the fuel bundle
/
[
will come"to rest on the fue Nal liner plate af ter shearing out the y
m.
cruciform. Resulting stressis on the remainder of the rack module mem-
')
! 1 bers will be less than yield.'
- Ihe effect of a fuel bundic dropping vertically and then pivoting and hitting the rack was examined. The total net impact energy for this con-dition (approxima'tely 13,300 in.-lb) is less than the energy of the strafght Mrtical drop and resulting member stresses are less than yield.
The hypothetical dropped fuel bundle accidents described above have been prototype-tested and analyzed to demonstrate design adequacy under worst-case conditions. The criterion that the fuel supporting structure remains elastic under accident impact conditions, such that the center-to-center spacing of the stored fuel is maintained, is therefore satisfied.
In the unlikely event of an actual fuel rack impact incident, all cf fected structural elements would, of course, be immediately investigated and corrective action taken, as necessary, to ensure structural integrity.
. O and safe fuel assembly storage conditions.
i V
3-9 Revision 3 (December 1981)
The effects of upward loading due to the maximum force that the crane would exert on a fuel assembly stuck inside a can was considered and the stresses due to this condition are well below yield stresses, and
^
are within code-allowable limits. The racks are designed for stresses due to loads in the downward direction which are greatly in excess of any load that could be applied by the crane in the upward direction.
3.1.2 CRITICALITY ANALYSIS The criticality analysis of the spent fuel storage rack design was per-formed by means of a series of diffusion theory calculations utilizing the CHEETAH-P/PDQ-7 model. CHEETAH-P is a modified version of the original LEOPARD code and uses a modified ENDF/B-II cross-section library. NAI has extensively tested the CHEETAH-P/PDQ-7 model through benchmarking calculations of measured criticals as well as tl,.> ugh core physics calculations of operating power reactors. Based on NAI bench-i marking experience, the maximum model uncertainty is 0.008 Ak.
This b
represents the largest difference between analytical results and mea-b sured criticals in actual reactor cores. Table 3-1 lists some of the excellent agreements achieved for PWR beginning-of-life critical configurations.
To ensure that the criticality analysis followed a very conservative j
approach and conformed to the general guidelines for safety analysis, h
the calculations were performed with the following assumptions:
- 1) Enrichment:
3.5 wt% U-235
- 2) Fuel:
fresh and nondepleted
- 3) Burnable poison rods:
replaced by water holes i
- 4) Control rods:
replaced by water holes L
i 3-10 Revision 3 1
(December 1981)
j
- 5) Minor structural members: replaced by water J
- 6) Fresh pool water
- 7) Hard thermal spectrum in the water gap
- 8) SFP water temperature of 68'F
- 9) Macroscopic thermal absorption cross sections (calculated):
a) Water at 68'F - 0.2636 cm" b) Stainless steel at 68'F - 0.0222 cm".
The physical parameters assumed for the reference case (or base case) analysis are listed in Table 3-2.
The basic cavity dimension is 13.3 sq in.
The rack cavity, which is made of Type 304 stainless steel of 3/16-in. thickness with a design tolerance of +0.05 in. and -0.00 in.,
,v has an outer dimension of 9.335 sq in.
Since the fuel assembly itself is 8.426 sq in., there exists a free space of 0.267 in. on each of the four sides between the fuel assembly face and the inside wall of the cavity. The reference cavity geometry is shown in Figure 3-5; Figure 3-6 gives a detailed sketch showing the dimensions inside a fuel assembly.
A two-dimensional X-Y four group PDQ-7 model, including an axial buckling to account for the axial leakage, was used for the criticality analysis.
A zero-current boundary condition was employed on all four outer bound-l aries of a storage cavity, as shown in Figure 3-5, to produce an infinite array effect. The reference case calculation result is keft = 0.9136 by the PDQ-7'model. In Figure 3-7 the reference case result is compared with the results of an independent calculation using the multigroup, multidimensional Monte Carlo neutron-transport KENO-II computer code with the 16 group Knight-modified llansen-Roach cross-section library.
The comparison confirms the validity of the diffusion theory calculation.
The sensitivity of the PDQ-7 model results to the basic cavity spacing V
3-11 Revision 3 (December 1981)
,.,(')
was evaluated by performing calculations with the spacing varied from 12.5 in to 14.0 in.
Figure 3-7 shows the calculated behavior of k gg
'~
e as a function of the cavity center-to-center distance.
Because of the free space existing between a properly centered fuel assembly and the cavity wall, it is possible for an assembly to be loaded off-center in a cavity. Two extreme cases of off-center loading configurations were evaluated to determine k gg.
The first case was e
a 16-assembly cluster with assemblies loaded in their cavities off-center and preferentially ieaning toward the center of the cluster as shown in Figure 3-8.
The zero-current boundary condition applied to the cluster outer boundaries produces an effect of an infinite ceray of these 16-assembly clusters in both directions of the X-Y plane.
For this off-center loading of the assemblies, all other dimensions in the configuration were maintained as in the reference case.
The second configuration, as illustrated in Figure 3-9, is four assem-
[h blies loaded off-center with all assemblies leaning toward the center i
/
of the cluster.
The zero-current boundary condition also produces an infinite array of four-assembly clusters in the X-Y plane. In addition to the effect of clustering, this second configuration includes the worst condition design geometrical and mechanical tolerances. The center-to-center spacing was reduced by 0.06 in, from 13.30 in. to 13.24 in.
The Type 304 stainless steel canisters with the reference wall thickness were also enlarged by 0.06 in., making each side of the square 9.395 in. instead of 9.335 sq in, as before. This reduces the intercanister spacing by 0.120 in. from 3.965 in. to 3.85 in.
1 l
The total effect of the clustering configuration and of the geometrical and mechanical tolerances is shown in Table 3-3.
The values of k gg e
for the 16-assembly and the four-assembly reduced-tolerance off-center cluster configurations are calculated respectively to be 0.9209 and 0.9241.
From a comparison of the results from the four-assembly off-center cluster calculation (k gg = 0.9241) and the reference capa e
(k gg = 0.9136), the effect due to possible off-center loading is
[~')
e
\\
O.01 Ak.
From an an intercavity spacing-sensitivity calculation to 3-12 Revision 3 (December 1981) i l..
account for any possible bowing and tilting of the racks, the reactivity effect is 0.002 ak. Thus, the total possible reactivity effect due to dimensional and positional tolerances is 0.012 ak.
Additional measures were taken to assure that a dropped fuel assembly in any possible configuration will remain 6 in. away from the rack sides.
Figure 3-4 shows a deflector on the side of the rack module approxi-mately 13 ft 0 in above the SFP floor. A second identical deflector was included in the design at approximately 1 ft 0 in, above the floor.
This second deflector will keep the base of an assembly at least 6 in.
from the bottom of the rack sides.
In addition, a criticality analysis was performed showing that even if the assembly was assumed to be tipped sideways and somehow pushed in contact with the rack module between the deflectors, the keff would be less than 0.95.
This calculation was performed using the Monte Carlo neutron-transport KENO-IV computer code with the 16 group Knight-modified Hansen-Roach cross-section library.
The spent fuel racks were assumed to be in the actual array of 21 by 31 assemblies surrounded on all sides by water and approximately 5 ft of concrete. A reduced pitch of 13.24 in, and assembly clustering was assumed. The result of this analysis was:
keff = 0.9180 1 0.0088 (95 percent confidence interval)
This analysis, including a 0.002 ak for temperature ef fects, yields a maximum k gg of 0.929.
e Subsequent analyses show that the keff should be 0.949 i 0.0094 (95 percent confidence interval). The following additional assumptions were included in these analyses:
- 1) The steel in the assembly grids was included in the calculation model.
O g/
3-13 Revision 3 (December 1981)
- 2) The AMPX library and the discrete fuel pin representation s
which had been used la other licensing applications (eg, Prairie Island) were used in place of the Hansen-Roach library.
The ef fect of temperature on reactivity has been found to be less than 0.2 percent ak. The effect cf boiling was evaluated by simulating the presence of equal void content inside as well as between cavities. The analysis results, which are plotted in Figure 3-10, show a continuous i
decrease in reactivity as the amount of voids increase equally inside and outside the cavity. This void-reactivity behavior will hold true as long as the void content inside the cavity is equal to or greater than the void content between or outside the cavities. As calculated in the thermal / hydraulic analysis, this is always the case for this rack design.
Calculations using two types of neutron spectra were -compared to assure f-'
conservative PDQ-7 results from the base set of cross-section values
.k generated. The conservative PDQ-7 results were used throughout the criticality analysis.
The results of the criticality analysis of the proposed spent fuel storage racks for Trojan are summarized below:
k gg, reference case 0.914 e
Dimensional and posi-tional tolerance, ok 0.012 Temperature effect, ak 0.002 Model uncertainty, Ak 0.008 Total 0.936 A calculated margin of 0.014 ok under the design limit k gg 0.95 is-e maintained even when all the possible positive reactivity effects listed above are added to the reference case. The SFP normally con-tains 2000 ppm of borated water; this concentration is estimated to f )h decrease the total keff from the criticality analysis (keff = 0.936) w 3-14 Revision 3 (December 1981)
p by 20 percent. Therefore, the normal condition is estimated to result V
in a keft of 0.75 with borated water in the SFP.
The large difference between the borated and unborated water conditions gives a in keff significant margin of safety below both the design limit (k gg = 0.95) e and that calculated from the criticality analysis.
The sensitivity of k gg to stainless steel thickness, fuel enrichment e
and burnup was considered in the design. Values for these parameters are as follows:
- 1) Stainless steel thickness of 0.1875 in. for the Type 304 stainless steel cavity wall. Under the reference condi-tions as stated in this report, the reactivity. worth of Type 304 stainless steel in the neighborhood of the nom-inal 0.1875-in.-thick container wall is approximately 0.03 percent ak/ mil (thickness) of stainless steel.
p
- 2) Maximum projected enrichment of 3.5 wt% U-235.
No
(.
enrichment sensitivity study was specifically.per -
formed for the Trojan SFP racks, however, based on information from other calculations, it is estimated that under the Trojan SFP conditions, approximately b-0.6 percent Ak in reactivity can be associated with an increase or decrease of 0.1 wt% enrichment around the base enrichment of 3.5 wt% U-235.
- 3) Fresh fuel as a conservatism. The spent fuel actually expected to be loaded in the spent fuel racks will be approximately 30 percent lower in reactivity than the ak/k assumed in the analysis because of lower initial enrichment, depletion of U-235 and buildup of fission products. The reactivity effect due to buildup of plutonium in the fuel is more than compensated for by depletion of U-235 and buildup of fission products.
The rod bowing that has been observco in spent fuel from
(/
PWRs has been so slight that it would not significantly 3-15 Revision 3 (December 1981)
affect the reactivity of the fuel. Similarly, any (gj damage that might be expected to occur to the fuel g
during normal operation would also not significantly increase the reactivity of the fuel.
The assumption of fresh (unborated) water in the pool ensures that there need be no limit on the minimum boron concentration in the SFP in the 3
Trojan Technical Specifications.
3.1.2.1 Analysis Overview for Delayed Modification If the reracking was delayed until af ter refueling, it was not expected to be started until several months af ter the refueling outage. However, calculations showed that if the rerack was not started until at least 2 months after reactor shutdown for refueling, the offsite doses from the postulated accident will be less than 10 CFR 100 guidelines. The 2-month decay time was calculated using the assumptions of Regulatory Guide 1.25 as described in the Trojan FSAR Section 15.5.9.2.1 except for the following:
- 1) It was assumed that all the rods in all 65 assemblies were damaged.
b
- 2) A radial peaking factor was not included because of the assumption that all the assemblies were damaged.
- 3) The spent fuel was assumed to have decayed for 30 days.
Using the above dose calculational parameters of Trojan FSAR Section 15.5 and the above assumptions, the calculated 0- to 2-hr site boundary doses we re :
Beta and Gamma Whole Body Dose - 2.5 rem (10 CFR 100 guidelines - 25 rem).
(q/
Thyroid Dose - 5.6 rem (10 CFR 100 guidelines - 300 rem).
3-16 Revision 3 (December 1981)
1 Even if no credit is assumed for the SFP Ventilation Exhaust System d
charcoal filter, the offsite doses are well below 10 CFR 100 guidelines.
A criticality anslysis using the LEOPARD code and the following assump-tions was performed to demonstrate that a minimum boron concentration in the SFP water of 2000 ppm would ensure a keff of less than 0.95:
1) 2.1 wt% initial enrichment of U-235.
The exposure of all assemblies was set at 8000 mwd /Mtu, which corresponds to the minimum nodal exposure expected at the end of Cycle 1.
- 2) The assemblies were assumed to be stacked in an infinite array at a minimum pitch (assembly width). -
S
- 3) No control rods or burnable poison assemblies were assumed to be present.
y/
- 4) The water temperature was assumed to be 120*F.
- 5) A boron concentration of 2000 ppe. was assumed.
Using these assumptions, a k rg of 0.87 was calculated.
e Therefore, should reracking be delayed until after refueling, PGE proposed that the minimum boron concentration be established during the reracking evolution at 2000 ppm and that the reracking not begin until at least 2 months after the reactor shutdown for refueling.
3.1.3 SEISMIC ANALYSIS Seismic analysis and design is in accordance with the criteria speci-m i
w l
fied in NRC Standard Review Plan, Section 3.8.4.
Horizontal and-vertical seismic response spectra are described in Trojan FSAR Section 3.7.1.1.
Specifically, for structural' design of the racks, the following absolute 3
sum load combinations and factored allowable stresses were used:
i 3-17 Revision 3 (December 1981)
i U
D + L + E + I.<. S O,
D + L + E' + I + Ta i 1.5S-D + L + FD 5 1.5S where j
D=
stresses resulting f rom dead load of rack structure L=
stresses resulting from live load of fuel assemblies E=
stresses resulting from the OBE hh 4
E' = stresses resulting from the SSE Ta=
stresses resulting from thermal loads at accident condition-l 1
I=
stresses resulting from fuel bundle " rattling" impact in cavity FD = stresses due to equivalent static loads resulting from fuel drop of the module S = allowable normal operating stress limits as delinested in ASME Section III, Appendix XVII.
Each of the four configurations of storage rack modules were modeled and
,) _
analyzed for static and seismic loads as described above using the SAP IV( } finite element computer program.
It was assumed for the mathematical model that the fuel assemblies add mass but no stiffness to the general. structure. The surrounding external and entrapped water.
was conservatively calculated to be equal to approximately 30 percent -
of the combined dry mass of the fuel and module.
3-18 Revision 3 (December 1981)
Seismic modal and spatial responses were combined in accordance with x
/
)
i,,,/
NRC Regulatory Guide 1.92, Revision 1.
The first eight modal frequencies were determined, which included the first and second modes in each of the three spatial coordinate directions. The analysis and results are further described in Appendices B and C.
3.1.4 SFP LINER PLATE AND FUEL CLAD INTEGRITY The SFP is a reinforced concrete structure with a seam-welded stainless steel liner plate. The liner plate is 0.25-in.-thick Type 304 stainless steel. The linar plate specification requires all welds to be smooth and allows grinding to produce smooth and/or clean finished weld surfaces.
Also,Section VIII of the ASME Code permits grinding of these welds.
Therefore, the grinding flush of either plug or seam welds is consistent with the vendor drawings, the liner plate specifications, and the ASME Code.
Written instructions do not impose special means to control grinding 7-~
(
)
heat input to affected zones because PGE's metallurgical engineer's evaluation concluded that the degree of microstructural sensitization of the base metal where weld grinding is performed is not increased by heat generated during the grinding operation. In comparison with initial welding, heat input by grinding is not considered to be significant.
Similarly, residual stresses that could be induced due to grinding, within the bounds of normal good workmanship, are not considered to be of sufficient significance to warrant special concern.
As on all Q-listed work, supervised experienced workmen were utilized to accomplish the work, and quality control was performed by qualified inspectors. The welds were specified to be ground flush. Visual and liquid penetrant inspections were performed af ter the grinding, and all areas of the liner plate where grinding was performed were inspected by the liquid penetrant method. Any unacceptable conditions (overgrinding, gouge, etc) was rejected and repairs performed and requalified before acceptance by Quality Control inspectors.
i
)
s 3-19 Revision 3 (December 1981)
~
4 i-1 A drainage system is provided - beneath the liner plate to collect leakage if it should occur. The drainage system is designed so that the loca-tion and magnitude of a leak can be identified and repaired.
i Small amounts of leakage can be identified visually by checking for evi-dence of water at the drainage system collection manifolds.- Routine
-sw shift and tour inspections of the drainage system assures early detection
- of leaks.
4 The SFP water chemistry will be sampled and maintained in accordance with Tables 3-6 and 3-7.
The SFP water chemistry limits and sampling j
frequency are. specified by Westinghouse and are incorporated into the plant operating procedures. Therefore, from a materials viewpoint, the.
environment to which the SFP liner or spent fuel is exposed is relatively benign and not conducive to corrosion. For example, Berry ( ). gives corrosion rates of approximately 0.02 mils /yr in 500*F water for zircaloy and 0.17 mils /yr in 600*F water with 1600 ppm H3B04 for Type 304 stainless steel. Berry further states.that zircaloy is not affected tnr
(
H B04 in the range of 1500 ppm.
Thus, given a fuel' clad thickness of i
3 22.5 mils, a liner plate thickness of -250 mils and the corrosion rates
{}
above, over.1100 and 1400 yr, respectively, would be required to pene-trate the fuel clad and liner plate.
The temperature increase of 15'F (125'F to 140*F) as a result of the rack modification is not of long duration (Section 4.2) and is well within the normal operating temperature limits for stainless steel and zircaloy.
Based on the above, it is concluded that neither the SFP liner nor spent fuel clad integrity should be affceted significantly by the rack modification.
I w/
3-20 Revision 3 (December-1981)
3.2 SPENT FUEL POOL COOLING AND DEMINERALIZER SYSTEM p
V 3.2.1 DESIGN BASES The SFP Cooling and Demineralizer System is designed to remove the decay heat from spent fuel in the SFP and to continuously purify the system b
water inventory. The SFP Cooling and Demineralizer System is designed to perform the functions described below.
- 1) Maintain the SFP borated water below 140*F by removing the decay heat from seven regions of 33,000-mwd / tonne-U stored af ter each of seven annual refueling operations.
The heat load under these circumstances will not exceed 1.89 x 107 Btu /hr.
- 2) Maintain the SFP borated water below 140*F by using the Residual Heat demoval (RHR) System in the event that seven regions with a burnup of 33,000-mwd / tonne-U are
[m)
(j stored for 1, 2, 3, 4, 5, 6 and 7 yr, respectively, and a complete core is unloaded 150 hr after shutdown. The heat load under these circumstances will not exceed 4.12 x 107 Btu /hr.
- 3) Maintain the clarity and purity of the borated water and l
control potential fisson product releases and other con-taminants. Even after appropriate scaling of present FSAR values, the resulting pool activity does not lead to apprecirbly larger releases than the previously estimated liquid or gaseous releases from the Plant. The SFP puri-fication subsystem can be used during the time an RHR train is being used for SFP cooling. A core flow path also exists using the normal letdown to the Chemical and Volume Control System (CVCS) demineralizers to provide additional j
purification.
l (O) l l
l 3-21 Revision 3 f
(December 1981)
- 4) Maintain the clarity and purity of borated water in the
,- s
/
\\
(,)
refueling water storage tank.
- 5) Supply makeup (normal and emergency) water to the SFP.
The SFP, refueling canal, the emergency makeup supply line, and the interconnecting lines to the RHR System are designed to Seismic Cate-gory I requirements. The SFP Cooling and Demineralizer System has been designed to Seismic Category II requirements.
3.2.2 DESIGN DESCRIPTION No equipment modifications are required fcr the SFP Cooling and Deminer-3212er System; however, the SFP Cooling and Demf aeralizer System opera-tion under normal and special conditions is changed. During normal operation, with one to seven regions stored in the SFP, both SFP cooling pumps will be operated initially to maintain the SFP water temperature below 140*F.
The RHR System and SFP Cooling and Demineralizer System
(
)
components are not adverse.. affected by increasing the SFP temperature design limit from 125*F to 140*F.
The refueling cavity and Reactor Coolant System (RCS) water temperatures are limited to 140*F during refueling operations in accordance with the Trojan Technical Specifica-tions. The proposed changes will make the SFP tempercture design limit compatible with that of the refueling cavity and RCS.
I When normal cooling of a full pool is considered, the individual assem-blies within the pool are cooled by natural circulation. For worst-case design conditions, the cooling water enters the pool at 112*F and is bb mixed with 140*F water at the sides and below the racks. This provides an inlet temperature to the individual channels of 130*F.
For this con-dition, analyses of the SFP Cooling and Demineralizer System show that the average SFP water temperature will be maintained below 140*F for the design bases listed in Section 3.2.1.
For a maximum power fuel assembly, conservatively assumed to be unloaded l
at the Technical Specification limit of 100 hr following reactor shutdown,
./
3-22 Revision 3 (December 1981)
i the maximum clad temperature in the fuel assembly and the assembly h
coolant outlet temperature are calculated using the IIYDROPOOL computer l}.
code to be respectively 183*F and 154*F.
HYDROPOOL is an NAl-derived computer code from Control Data Corporation - computer code IIYDROP, CDC-84004500 - This analysis assumed that 100 percent of the decay heat was generated within the fuel assembly. Under these conservative condi-tions, no voiding was calculated to occur in a cavity containing a maximum poser fuel assembly.
Natural circulation of pool water between the fuel racks removes the 4
decay heat that is deposited between the racks. Slots at the top of the racks permit natural circulation between the racks. The decay heat generated between the racks was calculated to be 6.5 percent of the total decay heat;-3.5 percent of the total was calculated to be generated in the stainless steel channel. To determine the natural circulation con-ditions between the racks, a total of.10 percent of the decay heat generated was assumed to be deposited in the space between the racks.
Based on this conservative assumption, the maximum coolant temperature between the racks was calculated to be 146*F for the regions adjacent to maximum power assemblies. Under these conditions no voiding will occur
- in this spau.
b The vapor produced frem the SFP could cause condensaticn in the SFP h
exhaust system that could adversely affect the charcoal filtration. To maintain the efficiency of the charcoal filter, a four-stage 90-kW heater
'is located in the duct ahead of the charcoal filter. The relative humidity of the air entering the charcoal filter is continuously moni-g tored. The heater operates as necessary to maintcin the relative humid-icy below 70 percent. The design conditions for this heater are as l
follows:
- 1) Temperature of air entering SFP ventilation exhaust -
145*F.
1
- 2) Relative humidity of air entering SFP ventilation Q
.(_)
exhaust - 100 percent.
i 3-23 Revisior. 3 (December 1981)
l-s
-l
- 3) Temperature of air entering charcoal filter - 155'F.
&O i
~
- 4) Relative humidity of air entering charcoal filter -
<70 percent.
j.
3.2.3 SYSTEM DESCRIPTION AND OPERATING MODES 4
The SFP Cooling and Demineralizer System consists of.the following components:
- 1) Two half-capacity cooling pumps (P-207A and B).
i j
4
- 2) Two half-capacity heat exchangers (E-205A and B).
- 3) One purification pump (P-208).
i
- 4) ~0ne purification filter (F-201).
1 b) i
(,
.5)
One demineralizer (T-224).
- 6) One demineralizer af ter filter (F-211).
i
- 7) One skimmer pump (P-209).
l-l
- 8) One skimmer filter (F-203).
- 9) Valves and piping.
- 10) Instrumentation.
l I
FSAR Tables 9.1-1 through 9.1-9 give physical data for the heat exchangers, f
purification pump and skimmer pump, purification filter, demineralizer af ter-filter and skimmer filter, the demineralizer, and the SFP. FSAR Figure 9.1-4 shows the piping and equipment for the SFP Cooling and
- y v
l Demineralizer System.
i O 3-24 Revision 3 (December 1981)
'The SFP Cooling and Demineralizer System is a closed-loop system consist-h ing of three subsystems - cooling, purification, and skimmer.
3 The cooling subsystem utilizes two half-capacity cooling pumps and two half-capacity heat exchangers. The cooling pumps draw suction from the SFP and discharge back to the SFP through the heat exchangers.
The suction piping from the SFP to the cooling pumps is connected to the RRR System through a normally closed 10-in.-diameter line. One 8-in.-
diameter return line from the RHR System ties into the 10-in.-diameter header returning the discharge from the heat exchangers to the SFP.
The purification subsystem utilizes the purification pump to divert 250 gpa of the total flow through the purification filter and/or the demineralizer. Both the filter and the demineralizer have sufficient capacity to recirculate the entire SFP volume (390,000 gal) approximately once daily. One 4-in.-diameter line from the refueling water storage tank (RWST) is conected to the suction sides of both the purification
()
and cooling pumps. This feature also enables the purification system to D
recirculate and purify the RWST water.
The SFP purification subsystem will be operated continuously during refueling operations and intermittently thereafter as required to main-tain SFP water clarity and purity. When it is not being used to purify SFP, the SFP purification subsystem can be used to purify RWST.
The operation of the SFP cooling pumps is indicated locally. Plant operating procedures require that these pumps be inspected once a shift for proper operation. This surveillance includes inspection for abnormalities such as excessive vibration, and improper suction ano discharge pressures. Plant operating procedures also require that both SFP cooling pumps and externally-actuated valves be tested quarterly to D
verify adequate system performance. This testing includes running the SFP cooling pumps on normal recirculation flow, exercising the valves with the SFP ourification pump idle, and recording system performance
[ j\\
t characteristics such as pump suction and discharge pressure.
v 3-25 Revision 3 (December 1981)
Modificatiori of the SFP racks increases only the storage capacity of the SFP and not the refueling frequency or the amount of spent fuel moved during the refueling. Thus, the ancunt of corrosion products which are introduced into the SFP water primarily during refueling operations will remain at about the same level regardless of the SFP storage capacity.
Increasing the SFP capacity may, however, increase the amount of fission products introduced into the SFP water. However, the evaluation in Section 5.1.4 shows that the purification system is capable of accommodating the potential increase in fission products even if a leaking fuel assembly is transferred to the SFP.
Therefore, the SFP purification system is adequate for the modified rack design.
E e need for filter cartridge and demineralizer resin replacement is
{}
based on the differential pressure across the filter cartridge and the decontamination factor of the demineralizer resin. Replacement of the filter cartridge and demineralizer resin will result in approxi-3 and 50 ft3 of solid waste, respectively.
It is anticipated mately 2 ft that filter cartridge and demineralizer resin replacement fr3Juency will
)
not increase significantly above the expected annual replacement rate as a result of the rack modification.
The purification subsystem can be used with the RHR System to maintain SFP water purity. The purification subsystem, including the demineral-izer resin, can accommodate a maximum water temperature of 140*F without degraded performance.
The skimner subsystem utilizes the skimmer pump to draw suction from the surface of the SFP, fuel transfer canal, and refueling cavity. The flow is then discharged through the filter back to the SFP.
Borated makeup wa ter to the SFP is provided by the holdup tank recirculation pump.
Demineralized makeup water is provided by the demineralized water trans-fer pumps.
Emergency makeup water supply is provided by the service water pumps.
Overflows from the SFP and the fuel transfer canal are directed to the s
/
s
(
clean waste receiver tank. Drain and overflow connections have also been provided for the refueling cavity.
3-26 Revision 3 (December 1981)
i Provisions have been made to pump a portion of the SFP flow to the CVCS
,_s
!,.,I holdup tanks to maintain the required boric acid concentration.
Ventilation is provided by a once-through ventilation system described in FSAR Section 9.4.2.
The discharge is monitored for radiation by the Process and Effluent Radiation Monitoring System, discussed in FSAR Section 11.4.
During normal operation,1/3 of the reactor core fuel elements are stored in the SFP.
Initially, both cooling pumps are operated to main-tain the SFP temperature at or below 140*F.
When the decay heat emitted by the spent fuel decreases, one cooling pump is stopped, and the remain-ing pump is sufficient to provide cooling water. Af ter one cooling pump is stopped, one of the two heat exchangers may be taken out of service for maintenance purposes.
A portion of the SFP flow is pumped by the purification pump through the
(
~
purification filter and/or the demineralizer for removal of impurities,
(,)
fission products, and other contaminants present in the SFP.
The cooled and partially purified water is discharged into the SFP through a diffusion header located near the bottom. Boric acid concen-tration in the SFP is maintained at minimum 2000 ppm boron. Provisions have been made for supplying borated water from the CVCS holdup tanks and demineralized water from the demineralized water storage tank, in con-junction with pumping the required volume of water to one CVCS holdup tank.
I The skimmer pump operates intermittently to prevent dust and debris from accumulating on the surface by pumping a small amount of SFP volume from near the surface through the filter and returning it to the SFP.
Provi-sions are made to filter and purify the content of the refueling water storage tank during normal operations by isolating the purification pump frcm the main cooling loop and opening two automatic valves in the line connecting the suction of the purification pump and the refueling water
(
)
storage tank.
3-27 Revision 3 (December 1981)
During cold shutdown and refueling operation, the refueling cavity, refueling canal, fuel transfer tube, and the SFP are kept full of borated water from the refueling water storage tank utilizing the residual heat removal pumps.
The cooling subsystem lineup procedure during the refueling operation is similar to that during normal operation. The refueling cavity water can be cooled by circulating it through the SFP Cooling and Demineralizer System heat exchangers. The refueling cavity water can also be filtered and purified by circulating it through the purification filter and demineralizer.
Air circulation velocity in the SFP area is kept sufficiently low so as not to generate ripples on the surface that would inhibit clear vision from the Fuel Building crane and transfer platform operator-into the SFP.
Af ter completion of the refueling operation, water from the refueling cavity and refueling canal is pumped into the refueling water storage tank. Special operating conditions are described in Section 4.2 of this
/N report.
(
)
.x /
3.2.4 INSTRUMENTATION A level switch is provided in the SFP to transmit high and low water level annunciation signals to the control room. This continuous monitoring will ensure rapid detection and appropriate corrective action for significant leakage. SFP water temperature-sensing element, transmitter, indicator, and control room high water temperature annunciator have also been provided.
Thermowells are pravided in the piping upstream and downstream of the heat exchangers to permit measurement of temperatures when required. Differen-tial pressure indicating switches are provided across the filters to moni-tor their effectiveness. Pressure indicators are provided upstream and downstream of both cooling pumps for monitoring satisfactory pump operation.
3-28 Revision 3 (December 1981)
. =. _
m i
Automatic air-operated Seismic Category I control valves (located on the
] sd suction line from the'RWST) are designed to fail in a clocad position and also to close on receipt of a safety injection signal, thereby isolating the Seismic Category II SFP Cooling and Demineralizer System from the
' Seismic Category I RWST.
{
The SFP cooling pump characteristics and RWST isolation valves are tested ij' every 90 days. The position of the Containment isolation valves is g;
l checked monthly either by visual verification or by checking the control room locked valve list.
l 1
i Nau 3-29 Revision 3 (December 1981)
.I
TABLE 3-6 SFP CHEMISTRY SPECIFICATIONS Analysis Value Remarks Cross Camma NA Determine purification requirements, evaluate leakage from spent fuel; analytical l
sensitivity 5 x 10-7 pCi/al.
Gross Beta NA Monitor activity level.
Tritium NA Evaluate in-Plant buildup of tritium; analytical sensi-tivity 10-5 pCi/al.
pH 4.0 - 4.7 Determined by concentration of boric acid present.
>2000 ppm B Prevent positive moderator coefficient.
Conductivity 1-40 paho/cm Consistent with pH.
t Chloride f0.15 ppa Prevent chloride stress corrosion.
u Fluoride f0.15 ppm Prevent fluoride stress corrosion.
<1.0 ppm Prevent deposition.
Magnesium fl.0 ppm Prevent deposition.
Suspended Solids fl.0 ppm Reduce deposition.
Sodium 31.0 ppm Control soluble impurities.
DEMINERALIZER INFLUENT:
Cross Gamma NA Evaluate performance of demineralizer. Analytical sensitivity 10-5 pCi/ml.
DEMINERALIZER EFFLUENT:
Gross Gamma NA Evaluate performance of demineralizer. Analytical sensitivitf 10-5 pCi/ml.
>25 Determine resin replacement time.
O Revision 3 (December 1931)
-.~.
.-... - - -. - - =.
)
F
?
TABLE 3-7 i
SFP SAMPLING SCHEDULE 1:
l Analysis Frequency Gross Gamma Monthly Gross Beta Quarterly 1
Tritium Monthly i
pH Weekly 7
I Boron Weekly
- i Conductivity Weekly Chloride Weekly l
Fluoride Monthly 2
~
AR[b]
l Calcium AR[b]
Suspended Solids
~ Monthly Sodium
.keekly i
{
DEMINERALIZER INFLUENT:
i Gross Gamma Monthly DEMINERALIZER EFFLUENT:
+
I Gross Gamma Monthly
Monthly
[a] In Mode 6, th maximum time between r
samples is 72 hr.
[b] As requested; these analyses would normally be requested as an explor-atory measure following other out-i of-specification conditions, eg, high sodium or high-suspended
- solids, 4
i LO i
Revision 3 (December 1981)-
1 1
,w,-~~--,,dm.w.n,n-,,0.-,,-m,,.,ga,~,n..,
,_,-nn_ln,-,,,.,4,,,mmn-,,,._..;,,nden.,,,--,,.,,,,
,,,,.,nn...,,r
..., e
4.0 SAFETY EVALUATION The safe storage of irradiated fuel depends on maintaining the integ-rity of the fuel cladding as the primary barrier against the release of radioactive materials. The protection of the fuel cladding is, therefore, a basic design requirement for the SFP and associated systems at Trojan.
The design considerations of the new spent fuel storage racks are almost the same as for the existing storage racks. The installation of new spent fuel racks was compatible with the previous storage facility, which was designed in accordance with General Design Criteria 1, 2, 3, 4, 5, 61, 62, and 63, as discussed in Trojan FSAR Section 3.1.
The applicable NRC regulations, NRC guides, and industry standards which are met are 10 CFR 20; Regulatory Guides 1.13, 1.29, g
1.92, and 3.41; Standard Review Plan, Sections 3.84, 9.1.2 and 9.1.3; ANSI N45.2 and N18.2; and ASME Boile'r and Pressure Vessel Code,Section III, Subsection NF.
Safety analyses pertaining to possible dropped spent fuel assemblies and failures in cooling and cleanup capabilities are presented in this section, which will replace the I
safety evaluation of the spent fuel storage facility des'cribed in Trojan FSAR Sections 9.1.2.3 and 9.1.3.3.
A radiological evaluation for refueling and normal SFP operation is presented in Section 5.0 of this report.
4.1 SPENT FUEL POOL AND SPENT FUEL POOL RACK SAFETY EVALUATION h
The design of the new spent fuel storage racks is such that it is impossible to insert the spent fuel assemblies in other than prescribed locations. Suf ficient center-to-center distance is maintained between adjacent spent fuel assemblies to ensure a kef g,<_0.95 even if unbora-ted water and fresh nondepleted fuel, enriched in U-235 to 3.5 wt%, are j
in the SFP (see Section 3.1.2 of this report).
~
The possibility of accidental criticality was investigated for.two cases of dropped fuel assemblies, one in which an assembly is dropped
\\d 4-1 Revision 3 (December 1981)~
l l
l
on top of the spent fuel racks, the other along the periphery of the h
rack array. Because the depth of the water between the top of assem-blies already inside the cavities and a hypothetical dropped assembly laying on top of the rack is approximately 9 in., no adverse reactivity effect is calculated to occur from dropping a fuel assembly during fuel handling on top of a fully loaded storage rack.
It is also possible to drop a spent fuel assembly parallel to an assembly in the storage array because of the 2.5-f t unobstructed space between the periphery of the storage array and the side walls of the SFP (see Figure 4-1).
A conservative analysis was performed for an assembly, assumed to be dropped during handling, which lodges parallel to an off-center assembly in an outer cavity. The analysis assumed reflective boundary conditions in three directions out to the three zero-current boundaries. A fourth boundary, the zero-flux boundary, was used as noted in Figure 4-1.
Including the dimensional and posi-tional tolerance effects, temperature ef fect, and model uncertainty, as discussed in Section 3.1, the analysis shows that a 6-in. minimum separation between the dropped assembly and the nearest outer storage array face is acceptable to maintain the overall suberitical keff below 0.95.
The summarized results follow.
Drop case with dimensional and positional tolerance, k yf 0.920 e
Temperature effect, Ak 0.002 liodel uncertainty, Ak O.008 Total 0.930 The structural design specifies deflector plates on the outside modules to preclude a dropped assembly from locating closer than 6 in. to the nearest stored fuel assembly, as shown in Figure 3-4 and described in Section 3.1.2.
Revised versions of cask drop accident were submitted in Trojan FSAR Amendments 1 and 5 to resolve NRC questions. The design of the spent O
-\\
f v
4-2 Revision 3 (December 1981) 1 J
fuel handling equipment is summarized in Section 9.7 of NRC Safety n
[
j Evaluation Report for Trojan Nuclear Plant dated October 1974; it
%d concludes the design was acceptable.
^
The layout of the fuel handling area is such that the spent fuel casks will never be required to traverse the SFP when spent fuel assemblies are in the SFP.
Physical obstructions identified as wheelstops will be in place and evidence that they are in place determined prior to any movement of the SFP building crane. Wheelstops prevent the crane from d
physically moving over the SFP and provide compliance with Technical Specifications regarding movement of heavy loads over the SFP.
- Thus, it is considered incredible that a spent fuel cask would ever fall into the SFP.
Nonetheless, all the fuel rods in one assembly are assumed to be damaged in a nonmechanistic manner in the fuel handling accident analyzed in Section 15.5.9 of the Trojan FSAR. The proposed modifica-g tion does not affect the assumptions and calculated consequences of the accident. This conclusion is based on the fact that it would take four independent concurrent failures to allow a spent fuel cask to fall into the SFP. These four failures are:
v
- 1) The mechanical stops must fail.
- 2) A mechanical device on the crane must fail, eg, the crane hook.
- 3) The safety slings must fail.
- 4) Administrative procedure control must be violated.
l The heaviest load that can be carried by the Fuel Building bridge crane is a loaded spent fuel cask for rail shipment; it weighs <195,000 lb.
If this cask was lifted to the maximum height and dropped, the impact would not damage the SFP.
This is accomplished by providing in the b
design of the spent fuel storage facility a-separate cask loading pit with a suitable gating arrangement to facilitate underwater transfer of
[]
fuel from the storage area to the cask, as shown in Trojan FSAR V
4-3 Revision 3 (December 1981)
Figure 9.1-1.
The dimensions of the cask loading pit preclude the tipping of the spent fuel cask while it is in the pit.
In the unlikely event of a sideways or end-first cask drop during tM lif t from the loading pit, the loading pit floor could fail, but the SFP structure would not be affected. If the loading pit floor failed, then only the water in the loading pit would be released since the refueling gate g
would be closed during all periods of cask movement. This volume cf water would not result in the flooding of any safety-related equipment.
A diagram showing the path for spent fuel cask movement is shown in Figure 9.1-10 of the Trojan FSAR.
Initially, the spent fuel cask is placed in the wash pit for cleaning. Nex t, the cask is moved to the
^
SFP cask loading pit where spent fuel can be transferred into the cask.
w Finally, the loaded cask is returned to the wash pit for decontamina-tion prior to exiting the building. Cask handling throughout this evolution will be performed by the Fuel Building bridge crane.
If a spent fuel cask were dropped onto the concrete floor above a CVCS O
holdup tank, the tank could be damaged. However, the content of the
(>
tank would be safely contained in the watertight concrete enclosure around the tank. For this event, the safe shutdown capability of the plant would not be affected. Any airborne contamination from a damaged g.
tank would be removed by the Fuel Building Ventilation Exhaust System.
7 If a spent fuel cask were dropped to the ground above the buried diesel fuel oil and service water supply lines, no breaks would occur in these lines because of the depth they are buried in the concrete.
The movement of loads over or around the SFP is limited in accordance with Technical Specification 3.9.7.
PGE has Plant procedures in the form of' Administrntive Orders regarding the Fuel Building bridge crane b
and Fuel Handling Procedures for the use of the SFP bridge crane, which wi11 limit loads and lif t heights over the SFP. There should normally be no movement of heavy loads in the vicinity of the SFP other b
than the spent fuel casks. Heavy equips.ent that is to be moved through mu O
the Containment equipment hatch, which is accessed through the Fuel U
4-4 Revision 3 (December 1981) et
---_._________N
Building on the same elevation as the SFP, will be transported along a set of rails that run perpendicular w the hatch centerline. The
' g nearest rail is 11 ft from the SFP at its closest point. Thus, heavy loads moved in and out of the Containment should come no closer than within 11 ft of the SFP.
In addition, since the load would weigh less than that'of a spent fuel cask, accidental dropping of this heavy load 3
would not affect the SFP integrity.
Nonetheless, administrative actions consisting of briefing and training riggers and secondarily involving design aspects such as safety factors used ir atruts, cables, hooks and rigging gear were f aplemented. In
'y aedition, it is expected that such work will be supervised at luast_by Plant staff personnel, although there may be some outside craft people involved in the work.
To provide further assurance that the spent fuel rack modification does not increase the consequences of an object striking the spent fuel,-
g freshly discharged spent fuel (that discharged less-tha 1 approximately
]
1 yr) will be stored no closer together than in every other storage position in the new racks, except for the following short periods of time when the storage distence may be reduced:
- 1). Initial reracking of the SFP.
Followicg raracking, the spent fuel will be located in at least every other storage cell if such work is completed within approxi-mately 1 yr after the reactor was shutdown for refueling.
C
- 2) Removal of a rack module to facilitate repair of a liner plate leak may require storage of the freshly discharged spent fuel assemblies closer than in every other storage location.
Following module reinstallation af tar repairs, the spent fuel will be located in at least e ery other storade cell, if such work is completed within approxi-mately-1 yr following reactor shutdown for refueling.
t
'a 4-5 Revision 3 (December 1981)
f
- 3) Discharge of a full core into the SFP. !!ost freshly discharged assemblies will be stored in every other-storage cell. Seventeen spent fuel assemblies will be stored in the diagonal storage locations between the other freshly discharged spent fuel. assemblies at a center-to-center
.2 9tance of 16.6 in.
l This storage configuration will result in the freshly discharged fuel I
being stored on a nominal center-to-center spacing of 26.6 in. compared to the previous nominal center-to-center spacing of 21 in.-
Spent fuel older than approximately 1 yr is stored in the positions between those occupied by the freshly discharged spent fuel. Any damage to the older g1 spent fuel will have negligible impact on the offsite radiation dose consequences since the isotope of concern, 1-131, will have decayed at least 45 half-lives. Thus, the consequences of any object strdlng the SFP was no greater with the modi?ted racks than with the existing racks.
i Q
The process of changing the SFP racks did not endanger stared fuel.
b Ihe sequence that was developed in changeout of the spent fuel racks I
j did not require carrying any acw or existing spent fuel rack over actual spent fuel storage positions. The plan was developed to
- minimize risk in handling heavy loads over spent fuel and required first that the fuel be located in essentially one corner of the SFP in l
the e,:isting Westinghouse racks. When sufficient existing racks were removed for placement of new t acks with suf ficient storage capacity, the fuel was moved from the existing racks to the new storage racks.
That sequence of removal was essentially the same as for the initial i
Westinghouse racks.
In other words, we continued to remove existing racks without carrying them over the actual spent fuel locations.
The layout and desiga of the Fuel Building bridge crane, including the use of interlocks, meet the requirements of Regulatory Guide 1.13, Spent Fuel Storage Facility Design Basis, as stated in Trojan FSAR b
Section 9.1.4.
The later more elaborate single-failure design features which were identified in Regulatory Guide 1.104, Overhead Crane Handling
-A)
(%.J 4-6 Reviston 3 i
(December 1981) s-+n er.
r-,,.
.--.,,---,we.,,,--r m e m m,
n,.
n
,.ng,
,,-~-e,
,+-.n-e,-,,wm-,,m-s~,,,,.,,v.e-,-re-emwwgmr+-m,vms,
f Systems for Nuclear Plants, are not applicable to the Trojan Nuclear
[~'}
Plant as noted in the distribution of Segulatory Guide 1.104 and its s
value Impact Statement (NRC letter f ram Guy A. Arlotto to Distribution g) dated February 12, 1976).
4.2 SPENT FUEL B'OL COOLING AND DEMINERALIZER SYSTEM SAFETY EVALUATION The SFP Cooling and Demineralizer System is not required for the safe shutdown of the plant.
If in the event of a loss-of-coolant accident, preferred power is available, the SFP cooling pumps would continue to operate if they were operating before the accident.
If only standby power is available, the pu93s would be shed f rom the power supply q
buses. Although the Seismic Category 11 portion of the system may not l.
be operable following a design basis seismic event, a Seismic Category I makeup water supply from the Service Water System.(SWS) is available to maintain the desired water level in the SFP as shown in Trojan FSAR
-sw Figure 9.1-4.
The only requirement ta assure adequate cooling for the spent fuel is to maintain the water lovel in the SFP so that the spent hel elements are not exposed.
a To prevent any possibility of loss of water f rom the SFP due to failure of inlet piping, a siphon breaker la (hc form of one 1/2-in.-diameter hole in the inlet pipe has been provided. The outlet piping'is connec-ted above the minimum permissible SFP water level to render impossible
,)
4 a loss of SFP wacer by failure of outlet piping below this level.
l Also, since the SFP cooling pumps are inspected once a shif t, the longest time interval between a pump failure and its detection is 16 hr.
This maximum time interval will occur only if the first inspection is-l performed at the beginning of the shift and the second inspection is l
performed at the end of the shJft.
If the SFP water-temperature
'}
limit is exceeded as a result of the failure of SFP cooling pump, this failure vill be detected sooner as a result of the investigation 1
(
performed following receipt of the alarm. The setpoint for the SFP aater-temperature alarm is 135*F.
l V'
4-7 Revision 3 (December 1981) l
a The four conditions that are discussed below were analyzed to deter-O mine the limitations of the spent fuel storage facility to cool the spent fuel stored in the SFP.
Condition 1 - For this condition, a calculated maximum decay heat load of 4.12 x 107 Btu /hr is used. Is perform this calculation, seven h
spent fuel regions are assumed in the SFP, each with an average burnup of 33,000 mwd / tonne-U.
A comp
- e core is also assumed to be unloaded with an average burnup of 22,000 mwd / tonne-U.
The regions are assumed b
to have been unloaded annually for 7 yr; the full core is assumed to have been placed in the SFP 150 hr after reactor shutdown. The decay heat factors used are consistent with the decay heat factors taken f rom g
~
NRC Standard Review Plan, Section 9.2.5 (revised November 24, 1975).
If a full reactor core is unloaded into the SFP at a time when seven regions are already stored in the SFP, then one RHR train aligned with the SFP would keep the temperature below 140*F.
The SFP temperature would be above 125'F for only 11 days af'ter reactor shutdown. The RRR train and SFP cooling subsystem cannot be used aimul.aneously to cool the SFP because the pipe diameter for the inlet pipe restricts the flow'
\\
rate to the SFP.
However, the SFP purification subsystem can be used with the RHR to continue to maintain SFP water purity.
3 i
Condition 2 - For the condition where the SFP is loaded with seven regions with a deca, heat load of 18.9 x 106 Btu /hr, a thermal analysis was performed for various cooling configurations. The seven j
regions are assured to each have an average burnup of 33,000 Mid/
tonne-U; six of thm. regions would be stored for 1, 2, 3, 4, 5 and 6 yr.
If the seventh regiva is discharged 125 hr af ter tne reactor is shut down, then both SFP Lw'.ing and Cleanup System trains aligned with the SFP would keep the temperature below 140*F and under 125*F af ter 20 3
days.
If one of the two SFP cooling pumps fails, the remaining pump and two heat exchangers would maintain tb9 SFP wat :r temperature below 145*F.
j However, under this sir. cation, the refueling cavity and the E3 l
temperature could approach the Trojan Technical Specification litait. of 140*F.
One RHR train aligned with the SFP would keep the temperature 3
O 4-8 Revision 3 (December 1981) i
below 140*F.
The SFP purification subsystem can be used with the RHR
'N train to continue to maintain SFP water purity.
%/
Condition 3 - For the condition where all forced flow to the SFP is a
lost, and SFP boiling and fuel temperature analysis of the maximum power fuel assembly was perforned.
It is assumed for this analysis that the water temperature begins at 140*F, that the maximisa power fuel assembly decay heat load is 3.03 x 105 Btu /hr, and tht' the water level is maintained at the siphon-break elevation of 83 f t 11 in.
The evaporation rate f roci a boiling SFP is approximately 90 gpm and the SUS makeup capability is greater than 200 gpm. The HYDBPOOL computer program was used to. analyze the heat transfer from the maximum power
-w fuel asser.bly. HYDFPOOL is an NAI-derived computer code from Control Data Corporatiun :omputer code llYDROB, CDC-84004400. The program evaluates the physical properties of water in the pressure range from 15 to 50 psia and includes the effect of static pressure changes on the saturation properties of water along the fuel channel.
P. valuating the Dittus-Boeltes correlation, using team properties, the HYDbPOOL program determines a lower and thus a more conservative heat transfer coefficient than would nora: ally be expected for a boiling condition.
The steam and liquid velocities are assumed equal for the void fraction predicted by the HYDBPOOL program, This assumption tends to.overpre-dict the void fraction for liquid velocities in the range of 1 to 3 ft/sec.
The results of the HYDBPOOL analysis of the maximum power fuel assembly 3
yield a void f raction of 0.23 and a steam quality of 0.034 percent in the top 1 f t of the maximum power fuel assembly where net boiling occurs. The coolant temperature is calculated to increase -to boiling in approximately 3.5 he assuming no heat loss from the SFP water. The calculated maximum fuel clad temperature is 249'F, and the calculated 7
maximum centerline fuel temperature is 254*F.
The integrity of the spen t fuel is assured under these conditi mc.
An examination of the thermodyna~b state in the space between the spent b
fuel racks was also perf ormed, aesuming that 10 percent of the gamma v
4-9 Revision 3 (December 1981)
energy must be removed by natural circulation between the racks. A
/'~h small amount of-boiling resulted at the top of the racks, but not at elevations opposite the fuel assemblies. A void fraction of 0.40 and an exit steam quality of 0.075 percent were calculated to occur in the hf space adjacent to but above the elevation of the maximum power fuel assembly.
Condition 4 - If all forced flow to the SFP is Ir,t af ter the spent fuel assemblies have decayed for some time interval, then the time taken for the SFP to reach a boiling condition is extended beyond t he 3.5 hr calculated for Condition 2.
Table 4-1 summarizes the maximum heat loads for two design cases:
4 (t) return of the plant to power operation fetlowing a refueling when the SFP has been filled with spent fuel, and (2) following a shutdown when the full core has been discharged to the SFF. lleat loads for both the existing and modified rack designs are provided.
r'~'
Table 4-1 shows that the additional SFP cooling system heat load because of the rack modification is approximately 14 and 6 percent,
~_s respectively, for Cases 1 and 2.
The increase in heat load is small because the decay heat rate from spent inel decreases rapidly. - For
,sw example, the decay heat rate from spent fuel decreases by approximately a factor of 10 from 150 hr after shutdown when the spent fuel is discharged to the SFP and 1 yr af ter shutdown when the next refueling is assumed to take place.
i Table 4-1 also shows that the rack modification will not significantly affect the Component Cooling Water (CCW) System operation. The CCW heat load increases approximately 2.3 and 3.6 percent for Cases 1 and 2, respectively. The maximum heat loads are well below the CCW design heat load for one CCW heat exchanger. Two CCW heat exchangers are provided.
)
(_/
4-10 Revision 3 (December 1981) l
{
Ihe plant heat dishrge rate to the river increases approximately 1.3 i
i and 2.9 percent, respectively, for Cases 1 and 2.
In both cases the
+
i heat discharge rate to the river is below. the NPDES limit.
i a
M i
v Therefore, it was concluded that the existing cooling systems were-i adeguate to accommodate the small increased heat load which resulted from the rack modification.
b i
i 1
}
i i
I r
f 6
4 i
i i
(-
s I
i i
i l
i I
i 6
?
f 1
4-11 Revision 3 (December 1981) i e
-,-,w-m ---ww,,,,
m n-m...
.._,,s-
~~~~-...-----~~~-~a=
5.2.1.6 SFP Modification Exposure
\\J Exposure estimates were made for three separate cases. The factors used in developing these radiation estimates are shown in Table 5-14.
Each case, with its asscciated man-rem estimate, is li.rted below:
1)
Installation of new racks prior to 0.4 man-res refueling.
- 2) Final installation of the new spent fuel 17.7 man-rem storage racks is delayed until af ter refueling, but all the preparatory work described in our letter to the Atomic Safety and Licensing Board of Septem-ber 20, 1977 is completed.
- 3) Final installation of the new spent fuel 50.3 man-res g
storage racks is delayed until af ter g
refueling, but Phase 2 of the preparatory work described in our letter to the Atomic Safety and Licensing Board of September 20, 1977 is not completed.
The estimates displayed a severe radiation exposure penalty associated with the installation of the spent fuel storage racks af ter refueling.
The new spent fuel storage racks were designed for installation in a l
dry spent fuel pool. The amount of time required to install the new l
racks af ter refueling was considerably greater than for other plants' racks which were designed to be installed in a water-filled spent fuel pool.
lu actuality, doses were substantially lower than the 17.7 man-rem projected during licensing hearings. As recorded in Table 5-15, the total whole Sdy exposur.e was 1.4 man-rem.
5-17 Revision 3 (December 1981)
The lower doses occurred because the SFP water had a lower radio-
{V'}
_ activity level than was anticipated and used for the origiani estimate.
c v
This lower radioactivity level was due to the determined effort to purify the SFP water prior to and during the rack replacement.
5.2.2 SITE, BOUNDARY DOSES Radiation doses have been evaluated at the north site boundary (662 m),
the offsite location having the highest annual average atmospheric dispersion factor. The submersion dose from noble gas releases and the inhalation dose from halogen, particulate and. tritium releases are evaluated with the same models as described in Section 5.2.1.3.
Continuous occupancy at the site boundary is assumed, and a breathing rate of 20,000 t/ day is applied. Atmospheric dispersion and deposition models presented in Regulatory Guide 1.111 together with onsite meteo-
-5 3
rological measurements yield a X/Q equal to 4.7 x 10 sec/m and p
plume depletion factor equal to 0.87.
V Site boundary doses corresponding to estimated atmospheric release rates in Tnbles 5-8 and 5-8a are given in Tables 5-12 and 5-13, respectively.
5.3 DISPOSITION CF EXISTING RACKS Although no spent fuel had been stored in the SFP, the SFP was used to temporarily store liquid radwaste in July 1976 prior to its processing b
in the radwaste system. As a result, low-level radioactivity was present on the lower 80 percent of the existing spent fuel racks. The 2
radioactivity levels ranged from an average 70,000 dpm/100 cm to a maximum of 180,000 dpm/100 cm. The predominatri nuclides were Co-58 and Co-60.
General radiation levels in the SFP were less than 0.1 mren/hr.
%s 5-18 Revision 3 (Decembar 1081)
=
N i
i i
A high pressure water spray was used to partially remove the cadiosc-tivity from the racks. However, it was expected that 80 percent of the existing spent fuel racks would have to be disposed of as low-level solid waste.
j S1 The packaged volume of the racks is not known precisely as this evalua-
]
tion has not been completed. liowever, based on experience at other plants, it is anticipated that the final packaged waste volume should 1
not exceed 1500 ft3 i
J l
- l i
J l
i I
I I
\\~
5-19 Revision 3 (December 1981) i
i i
TABLE S-le 6
i h
FACTOR FOR INPLANT RADIATION EXPOSURE ESTIMATES 4
d FOR SPENT FVEL STORAGE FtCK WORK Dose Rates in b'ork Area t'
General area dose rate in dry spent fuel pool (seasured dose rate).
0 1 area /hr l
1 General area dose rate around dry spent fuel pool (seasured dose rate).
Negligible General tres dose rate in spent fuel pool after refueling. Calculated based on 10.6 area /hr 0.12 parient failed fuel and corrosion product input f rom spent fuel movement and
]
rack replacement. Measured dose rates at other plants:
~
l Fort Calhoun 10-15 area /hr Prairie Isis 20 arca/hr General area dose rate around spent fuel pool af ter refueling. Nessured dete fates 1 ares /hr 3
at other plants:
}
Fort Calhoun 2-3 area /hr
[
Prairie Island 2 5 area /hr Contact dose rates on existing racks prior to refueling (measured).
0.2-0.5 area /hr Contact dose rates on exisitng racks af ter refueling. Measured dose rates at another 10 ares /hr 8
plants l
Point Beach 10-70 ares /hr af ter washdown Man-hours in' Radiation Area [a]
Installation of new racks [b] prior to refuellag.
26% man-hr O
Installation of new racks after refueling. but all prepa stor is completed CI. y work described 5288 man-hr in our letter to ASLB of September 20, 1977 Installation of new racks af ter refueling, but Phase 2 of preparatory work described 8504 man-hr in our letter to ASLB of September 20, 1977 is not completed.
Disposal of existing racks.
325 man-hr
[a] Includes only work involving radiation exposure and does not include tasks such as unloading new racks after shipment, etc.
[b] Installation o+ the new spent fuel storage racks involves the following steps:
i (1, e roove ext sting racks.
(2) Fit and weld embedment stud sleeves (56 pieces).
j i
(3) Inspect embedment stud sleeve weiss.
a (4) locate alignment sleeves and embedaent module locating f rases over embedeent studs. Weld embedsent f
module locating f rames to alignment sleeves (11 embedment module locating f rames and M alignment sleeves).
(5) Locate and weld support cups to embedment module locating frases using template and new rack modules (56 support cups).
l (6) Inspect fitup of specified tolerances and welds.
(7) Locate new rack modules into SFP.
j (8) Level rack modules and fit and weld the top module-to-module connections (40 connections).
l (9) Perform final alignment checks and inspections of new racks to confirm correct installation.
[c] Underwater work in spent fuel pool is escinated to require approximately three times the man-hr of comparable work in a dry spent fuel pool.
Revision 3 V
(December 1981)
TABLE 5-15 RADIATION EXPOSURE ACCUMULATED DURINC SFP MODIFICATION (area)
Work Description Exposure SFP rack, clamp and bolt removal 124 Lining as.
uel rack boxes 36 Diver setup and testing 17 Moving spent fuel and poison rods 18 General carpenter and labor support 261 SFP' rack removal 323 m
M New rack installation (excluding divers) 450 Drag testing new racks 2
Repair of spent fuel handling tool 2
j Scaffold placement for crane stop movement 33 Removal of lights from pool 0
i Divers in SFP (TLD data) 152 1418 Note: With the exception of the exposure listed for Divers, all 1
data was obtained by pocket ion chamber which will be conservative for these small doses.
2 i
l i
O Revision 3 (December 1981) i 4
.l i,_.._____,.,,___,,.,_,,,_._,___ _,, _ _ _ _ _., _ _ _ _ _ _. _
7.0 TCSTS AND INSPECTION
,r )
The spent fuel storage racks were designed, f abricated and installed
~~
as safety-related Seismic Category I equipment. As such, the quality assurance program for the spent fuel storage racks is governed by 10 CFR 50, Appendix B.
To satisfy this requirement, PGE and the 7) contractor, Programmed and Remote Systems, Inc. (par), adopted and implemented quality assurance (QA) programs that conform to ANSI N45.2-1971. All work was subject to inspection and approval by PGE's inspector for conformance to specifications, drawings and quality control requirements.
It was not considered necessary to impose requirements beyond ANSI N45.2.
Revision 1 of par QA Program was approved in March 1977, and the deletion of the ANSI referenced standards was incorporated into Revision 2 of PGE-1013. The foregoing is consistent with Section 3.2.3 of ANSI N45.2.13-1974, which does not require imposing QA requirements
{}
beyond ANSI N45.2.
The deletion of the referenced ANSI standards did f' N not cuange any of the material requirements. Material test reports t
}
\\_/
provide conformance with applicable ASTM specifications. ASME Section II, in turn, endorses ASTM standards.
The PGE contract with par to desige and f abricate the spent fuel racks was signed in January 1976, and the first independent inspection by Bechtel Procurement Inspection, as an agent of PGE, was made in April 1976. Also, the par QA program was reviewed and approved in April 1976.
FGE verffled the implementation of the par QA program in two separate audits, which did not detect any significant findings.
O w
During the manufacture of the spent fuel racks, Bechtel Procurement Inspection performed the tasks associated with " Customer Witness and Hold Points" as well as normal supplier inspection activities.
The PGE QA Department also audited Bechtel Procurement Inspection activities relating to their responsibilities as an agent of PGE.
h
%j 7-1 Revision 3 (December 1981)
par is an experienced fabricator, and the contractual and regulatory
(}#
requirements imposed contained suf ficient controls to assure that the final product met the design / contract requirements.
Af ter assembly of the spent fuel storage racks, tests were conducted by par with a representative fuel assembly (dummy) at least 0.27 in, wider than an actual fuel assembly. The dummy assembly was placed in each cavity to ensure that no binding occurs. Following field installation and before using the racks, tests were also conducted by PGE with a representative fuel assembly which more closely duplicates an actual fuel assembly than the dummy used by par. The insertion and withdrawal drag forces were measured with a load cell and were required to be less than 50 lb during both the PGE and par tests. Each guide assembly 7) centerline was aligned so as not to deviate by more than 0.125 in. from its true vertical per the design specification requirements.
Equipment in the SFP Cooling snd Demineralizer System is accessible for inspection under normal operating conditions. Maintenance of the SFP l
/~'s water below 140*F is an indication of satisfactory operation. The SFP
)
cooling, purification and skimmer pumps were tested in accordance with s-the standards of the liydraulic Institute and found to satisfy the respective performance requirements. The SFP cooling pumps were tested i
one at a time for their ability to operate locally by manually operat-ing the respective push-button switches.
Pump characteristics, heat i
exchanger capacity, filter and demineralizer performance, and Contain-ment isolation functions are monitored periodically during operation.
e (n}
%/
7~2 Revision 3 (December 1981)
a 1
i i
APPENDIX A RESPONSES TO REQUESTS FOR ADDITIONAL INFORMATION Appendix A provides a summary of the requests for additional informa-tion that preceeded the NRC approval for Amendment 34 to Facility f
Operating License for the Trojan Nuclear Plant.
mw v
This Appendix encompasses the supplements submitted by PGE in support of the initial license change application dated January 6, '1977.
4 Listed with each summary is the applicable PGE docuraent that demon-1 j
strates compliance with ear.h request.
I If further clarification is needed, PCE letters to the NRC, as listed in Amendment 34, may he referred to.
i I
i.
I i
i i
1 i
V j
i a
1 A-1 Revision 3 1
(December 1981) l 4_,.-._--._.,_..,__,____.-______._
Summary of NRC Request PGE Documentati on of J.
eN for Additional Information Compliance (PCE-1013)
J In the footnote on Page 3-1, indicate if Rev.1, Section 3.1 Figure 9.1-1 of the FSAR will be replaced by Figures 3-2, 3-3 and 3-4 of your report.
Provide detailed sketches of the locating Rev. 1, Figure 3 3-2a frames and the cruciforms at the bottom of thru 3-2j,
the rack cavities. State the weights of the Section 3.1 rack modules in the table on Page 3-3.
- Also, provide the dimensions of the stud sleeves on the embedment assemblies and the bolts in the 5
tie plates at the top of the modules illus-i trated in Figure 3-3.
1 Provide sketches of the mathematical models Rev.1, Appendix B of the fuel pool, the fuel storage rack, and
-sw the fuel assembly system which were utilized
[J
\\
in the analysis. Illustrate on the sketches R.
the mechanism of shear and load transfer to the fuel pool floor slab. Discuss how the effects of sloshing water, the surrounding i
external and entrapped water, and the ex-1 pansion allowances in the connections were considered. Also, provide the resulting 1
significant nodal frequencies of the fuel racks in air and water and the corresponding mode shapes and participation factors. State i
the damping values assumed for the fuel racks.
s Indicate if the combination of the spatial Rev.1, Appendix C responses and the modal responses for the fuel storage rack seismic systen is in accordance I
with kegulatory Guide 1.92.
C)
\\_ -
A-2 Revision 3 (Dscember 1981) q
Summary of 'NRC Request PGE Documentation of for Additional Information Compliance-(PCE-1013) d Your reference to the FSAR and the ASME Code Rev. 1, Saction 3.1.3 Section III, Subsections NA and hT, regarding the loads, load combinations and acceptance
- criteria utilized in the design of the racks is not sufficient for an adequate review of the design. Therefore, provide a summary for these items.
Compare the most severe temperature distribu-Because of the dif-tion considered for the structural design of ferences in design the fuel pool structure for both the original documentation, it rack design and the new rack design. Also, was not meaningful provide the most severe temperature differ-to make the required Q
ential through the pool utilized for the new comparison.
rack design.
Provide a summary of the highest stresses, the Rev. 1, Section 4
~
corresponding safety margins, the locations in 3.1.1, Appendin B the rack structure where these occur, and the maximum displacements at the top of the racks i
for the loading conditions considered in the.
analysis of the rack structure.
Provide the details of your analysis consider-Rev.1, Appendix B ing the impact of the fuel assemblies against the rack walls. Show how you incorporated the local effect into the total effect on the rack design.
I l
i
!O A-3 Revision.3
-(December 1981)
s t
Summary of !!RC Request PGE Documentation of for Additional Information Compliance (PGE-1013) a Provioe a detailed summary of the stress Rev. 1, Tables 3-4 margins in the anchor bolt embedment assem-and 3-5.
blies, tie plate components, locating frames and the fuel. pool floor for the loading Because of the dif-imposed by the new rack assemblies for the ferences in design critical load combinations. Compare numeri-documentation, it cally these results to those for the previous was not meaningful rack structure.
to make the required comparison.
On Page 3-3, it ' has been mentioned that the Rev. 1, Section 3.1.1 fuel racks are designed to withstand the effect of a dropped fuel assembly. State the assumptions regarding the masses, the kinetic energy of the dropped assembly and the ductil-icy factor of the target in absorbing the energy of impact, indicating any possible effect from buckling. Provide the results of your analysis.
t On Page 4-3, it is stated that there is no Rev. 1, Section 4.1 1
threat to the integrity of the pool if a l
s pent fuel cask is accidentally dropped into l
the pool from the maximum height. Stute the assumptions regarding the kinetic energy of the dropped cask and the ductility factor of the target in absorbing the energy of impact and provide the results of your analysis.
C]
k A-4 Revision 3 i
(December 1981) i
Susanary of NRC Request.
PCE Documentation of for Additional Information Compliance (PCE-1013)
. ~
. ( )\\
On Page 3-3, it is stated that "the racks and Rev. 1, Section 3.1.1.
their interfacing structures are constructed almost entirely of Type 304 stainless steel".
Provide a list of all materials utilized for the design of the fuel rack structures, their corresponding functions, and their applicable specifications.
State clearly the codes which are utilized for Rev. 1,.iections the design, fabrication, and installation of 3.1.1 and 3.1.3; the rack structure.
Specify the codes from Section 7.0.
which the maximum stress limits (at appro-priate temperatures) for the materials and welds were obtained.
1 The utilization of a 25 percent increase in Rev. 1, Section 3.13, g
the normal operating stress limits of ASME Appendices B and C.
Section III, Appendix XVII as the acceptance criterion in the OBE load combinations on Page 3-12 of your report is unacceptable.
The normal operating stress limits of ASME Section III, Appendix XVII, without any increase, should be utilized. In addition, a comparison of Tables B-6 and C-1 indicates that the combination of the three earthquake components by the SRSS method generally yields higher stress resultants within a rack module than if the resultant earthquake stress is taken as the greater of the absolute sum of the vertical component with each of the individual horizontal components. Therefore, the earthquake components should be combined utilizing the SRSS of the three individual
' O components.
kJ A-5 Revision 3 (December 1981) f
Summary of URC Request PGE Documentation of 7x for Additional Information Compliance (PGE-1013) k
)
x_-
The load combination results in Tables 3-4, Rev. 2, Table C-1 B-3 through B-6, and C-1 indicate that the I
ASME Section III, Appendix VII interaction Equation 21 was utilized for the calculation of all results. The code states that Equa-tion 21 shall be used in lieu of Equation 19 only when the applied axial stress is less than or equal to 15 percent of the allowable axial stress. Therefore, utilize the appro-priate interaction equations in these tables.
In addition, provide a sumnary of the resul-tant shear stresses within the rack modules.
C The assumed values of "K" (effective length Rev. 2 Appendix C factors) were not indicated in Figure B-1 and Table 3-4 of your report.
In a telephone con-s
/
)
versation with representatives of the NRC,
_ l par, and PCE on May 24, 1977, it was stated that K=.5 was utilized throughout the analysis.
Justify your utilization of this value, rather than the value of.65 as recommended in the Commentary Section of the seventh edition of the AISC Code. In addition, since the built-up tunnel sections at the top of the rack modules are welded continuously to *.he tops of the canisters and adjacent inverted funnels, thereby providing for tul ? lateral support, clarify whether gross or local buckling was considered in the calculation of allowable axial stresses.
o I
\\
(._/
A-6 Revision 3 (December 1981)
Susanary of NRC Request.
PCE Documentation of n -
for Additional Information Compliance (PCE-1013) i t
Your utilization of 3 percent additional Rev. 3 Appendix B; damping.due to rack submergence is unaccepta--
References 27 and 28.
ble. Based upon an examination of the current literature on the effects of submergence on damping, indications are that there is a neg-ligible increase in damping due to submergence for objects the size of the rack modules in an environment similar to that which exists in a fuel pool. Therefore, the damping values should be taken as 2 percent and 4 percent for the OBE and SSE' conditions, respectively.
Justify the use of a maximum modal frequency Rev. 2, Appendix B.
cf 15.70 Hz as indicated in Table B-1 of your report. The maximum modal frequency con-h sidered should be 33 Hz unless a lower cut-off frequency can be quantitatively substantiated.
The allowable stresses presented in Rev. 2, Appendix B.
Tables 3-4, B-1 through B-6, and C-1 indicate that a yield stress of 30 KSI.was utilized, the yield stress for stainless steel at 100*F.
The yield stress at the appropriate tempera-ture should be utilized, f
Consider the possible loading arising from a Rev. 2, Section 3.1.1.
temperature g,radient through a rack module, (eg, consider the case of an empty set of canisters adjacent to a full set of cani-sters). Also, dia:uss the effects of the increased fuel pool temperature on the fuel pool walls and liner.
ws A-7 Revision 3 (December 1981) 1 L
Summary of NRC Request PCE Documentation of for Additional Information Compliance (PCE-1013)
On Page 3-7 of your report,.the discussion of Rev. 2 and 3, the impact of a fuel assembly is not complete.
Section 3.1.1 The local and gross effects on the rack modules must be discussed and quantified for the following three possible cases of a fuel-assembly drop:
a) A straight drop on the top of a rack module.
b) An incline 3 drop on the top of a rack mod ule.
c) A straight drop through a can with the g
~
fuel assembly impacting the cruciform at the bottom of a can.
v Include the kinetic energy and the drop 4
height considered for each of the three cases.
Also, discuss the effects of several impacts of dropped fuel assemblies affecting the struc-tural integrity of a rack module.
Impacts on l
the built-up funnels of a module, which are relied upon as structural members, will af fect their load carrying capacity. Assurance that the effects of the incident on the structural integrity of the rack will be assessed immediataly following the incident.would be suf ficient to ensure the safe storage of the i
fuel assemblies.
l I
l
,p A-8 Revision 3 l
(December 1981)
-~
~
-~ _-
4 Summary of NRC: Request PGE Documentation of for Additional Information Ceapliance (PCE-1013)
In' addition, consider-the effects of the
-loading which will result from a fuel assembly sticking inside a can (this loading is defined-j in' ANSI N210-1976)., The upward loading should be the maximum force the crane is allowed to
)-
exert on a. fuel assembly.
In Appendix B,-your reference to ANSI Commit-Rev. 2, Appendix B.
tee 20.2 draft report entitled " Design Basis for Protection Against Pipe Whip", dated June 1973, for the calculation of the effects I
of fuel impact on the rack structure during a I
seismic event is not suf ficient. Provide the
_ u j
basis of the method and the details of your calculations. Also, discuss the effects of the impact locally on the can and. a the fuel
/ h assemblies themselves for both OBE and SSE.
(/
The additional bolt loads out of the plane, Rev. 2, Appendix B.
illustrated in Figure B-8, arising from-the i
vibration of the rack modules while containing I
the worst possible unsymmetric' array of fuel assemblies, should be considered, i
l Provide the water chemistry which will be Rev. 3, Tables 3-6 maintained in the spent fuel pool.
Include and 3-7.
the. boron concentration, pH, and the chloride, the fluoride and any heavy metal concentrations.
i-l
.I, /"')
!- V A-9 hevision 3 (December 1981)
J b
i
.m
i Sunmary of NRC Request ~
PGE Documentation of for Additional Information Compliance (PGE-101.Q You have stated the average burnup of the Rev. 2, Section 5.1.1.
- spent fuel in the pool is 33,000 mwd /MT in Sections 3.2.1 and 4.2 of your April 1977 submittal and 35,000 mwd /MT in Section 5.1.1 of your April 1977 submittal. What is.the average burnup expected for the spent fuel in the pool?
Discuss the expected change in the radio-Rev. 2, Sections logical gaseous effluents from the spent fuel 5.1.5 and,5.1.6; i
Tables 5-8a and 5-13.
pool area because of the proposed modifica-tion. Include in your discussion the impact of the increase in the maximum pool bulk water l
temperature from 125'F to 140*F because of the propoaed modification.
-l 1
C.
What is'the weight of the fuel in the core in
'Rev. 2, 4
metric tons?
Section 5.1.1; Trojan FSAR Table.4.3-1.
1 What will be done with the material to be Rev. 2, Section 5.3.
removed from the spent fuel pool- (eg, spenc l
fuel racks) because-of the proposed modifica-
- tion? If the material to be removed will be disposed of as solid radwaste, what is the' volume of the packaged waste?
i What. is the volume of solid waste generated Rev.-2, Section 3.2.3.
. by the replacement of a cartridge filter in the Spent Fuel Pool Cleanup System? For the proposed modification, what is the frequency l
of operation end the expected flow rates through the cartridge filter and demineralizer "N
)
during a year? What is the expected frequency l
J I
A-10 Revision 3 (December 1981) i
'4
Summary of NRC Request PCE Documentation of for Additional Information Compliance (PCE-1013) j,_k)
{
of replacing the cartridge filter and deminer-alire'r, and what is the basis for their I
replacement?
In Section 3.2.2 of your April 1977 submit-Rev. 2, i
tal, you stated no equipment modifications Section 3.2.3.
were required for the Spent Fuel Pool Cleanup System. Explain why the Spent Fuel Cleanup System is adequate to maintain low pool water concentrations, so that there are reasonably low exposure levelu in and around the spent-fuel pool area, during and af ter the modifica-
~
l.
tions of the pool. Provide applicable experi-(;
v ence from the other plants te support your estimates.
In Section 5.1.6 of your April 1977 submittal, Rev. 2,' Sections f
you assumed credit for decontamination factors 5.1.6, 3.2.2 and 3.1.
of 100 and 10 for HEPA filters and charcoal
-beds to estimate releases to the environment.
Is the Spent Fuel Pool Exhaust System (SFPES) used continuously to filter air discharged i
j from the speat fuel pool area? What are the design conditions of the heater for the System, and does it operate continuously? Where and how do you sample for iodine being released i
to the environment from the spent fuel pool area?
f 8
4 I
I I
/
f A-ll Revision 3 (December 1!81}
f l
l L
.s,,
..,, _, _ _ ~, -, _ _,. - -.
_.. -. - - -, - ~, - -, - -, -, - - -
k i
Summary of NRC Request PGE Documentation of
'for Additional-Information Compliance (PCE-1013)
O Table 5.11 of your April 1977 submittal Rev. 2, Section
. itemizes refueling operations with the respec-5.2.1.5.
tive man-rem dose. equivalent that will be
- received by personnel from each operation during a normal refueling. Describe the impact of the proposed modification on the man-rem exposure of the items listed (eg, fuel storage and inspection operations').
i The dose equivalent rate at the pool surface Rev. 2, Tables 5.9 l
is given in Table 5.9 as 7.8 mrem /hr. Explain and 5-11.
why this value is not in the listed exposure rates of Table 5.11 for spent fuel pool 3
operations.
Y Is the overhead handling system, including
- Trojan FSAR Section
()
rigging gear, single-failure proof?
9.1.4; NRC letter from Cuy A. Arlotto 4
dated February 12, 1976.
What heavy loads, other than casks, may be Rev. 2 and 3, moved in the vicinity of the spent fuel pool.
Section 4.1.
i
' Describe the travel path of the spent fuel Rev. 2 and 3, I
cask in the vicinity of the spent fuel pool.
Section 4.1.
l What is the gross weight of the spent fuel Rev. 2 and 3, cask to be used?
Section 4.1.
' F) -
\\~ -
A-12 Revision 3 (December 1981) 1 i
- -- - - - _ _. - - - - - _. - - - - _ - - - _ _ _ _ _ - - - _ _. - _ _ - - - _ _ _ _ _. -.. - - _. - _ _. _ _ - -. _ _, _ - - - - _ - - - - _ - _. _ _ _. _ _.. _ _. _ _ ~ _ _ - - -.. - - - - - _ -
.-.- w---
4 Summary of NRC Request PCE Documentation-of for' Additional Information C6mpliance (PGE-1013) i. J
.Giscuss any possible scenarios by which it is Rev. 2 and 3, conceivable th.at a spent fuel cask or heavy Section-4.1.
load could fall or tip into the spent fue l J
pool.
Are there any plans for preferred' spent fuel nev. 2 and 3, storage. configurations in the spent fuel pool?
Section 4.1.
Clarify the statement that dif ferences in fuel Rev. 3, Appendix B.
rack seismic res ponse using a 2 percent increase in damping due to submergence and added conservatism in the virtual water mass are essentially compensating.
C v
Clarify the effect of the increase in the fuel Rev 3, Section 3.1.1.
(
pool nominal design temperature from 125*F to 140*F in terms of applicable load combination stress limits.
It is stated that stresses in the racks due
.Rev. 3, Section 3.1.1.
to the maximum pull-out force exarted by the crane on a stuck fuel assembly are well below yield. Clarify whether or not thesa stresses j
are within the applicable Code acceptance criteria.
It is stated that combined stresses in the Rev. 3, Appendix C.
fuel racks for load combinations including i
fuel bundle impact inside cavities remain below yield. Yield stresses are not the acceptance criteria. Clarify whether or not stresses are within Code allowable values.
%.s A-13 Revision 3 (December 1981) l
[
1 i
j Summary of NRC Request PCE Documentation of r
' for Additional Information Compliance (PCE-1013)
(
x Describe the effects on the fuel bundle of impact inside-a cavity as previously requested.
3 d
j From the referenced description, it does not Rev. 3, Appendix B.
appear that torsional conditions dne to possible unsymmetric fuel storage in the fuel i
rack airay were analyzed for the tie-bolt design. Justify the tie-bolt design approach with respect to the possible torsional effects.
How will the failure of either one or both of.
Rev. 3, Sections the Spent Fuel Pool (SFP) cooling pumps be 3.2.3 and 3.2.4.
detected?
t$
4 1
I)
What will be the length of the longest time
.Rev. 3, Sections
. U intervel between a spent fuel. cooling pump 4.2 and 3.2.3.
failure and its detection?
f The brief description ~provided in the Licen-Rev. 3, Section 2.0.
l see's letter to the Licensing Board dated PGE 's Answer to l
September 20, 1977 is not sufficient to gain David B. McCoy's
[
a ;ull understanding of the SFP work that is Motien for Disclosure l
currently.In progress. Describe in detail the and cease and Desist,'
I l
work ' that has been performed and -will be per-
. dated October 5, l
formed in the SFP prior to a dec tston with 1977.
e l
respect to the proposed amendment, and. include F
'your justification and safety evaluation as to I
why these activities do not involve an unreviewed safety question or otherwise
(
require prior NRC' approval.
r i
l x.-
l-l A-14 Revision 3 (December 1981)
Summary of NRC Request PCE Documentation of L
for Ad fitions1 Information '
Compliance (PCE-1013)
Describe the effects. and forces upon < & el Rev. 3, Appendix B.
assembly at impact on the side of storage cavity resulting from a seismic event. The additional information provided in the letter of Sep* ember 27, 1977 (Goodrin to Schwencer) did not provide sufficient specific data to support the numerical conclusions.
i Clarify the comparison of calculated to Rev. 3, Appendix B.
design allowable stresses for the welds and top tie-bolts.
i The fuel assembly stress results, presented Rev. 3, Appendix B.
in terms of ratios of allowable stresses to 4
calculated stresses, were based on unir-
{}
l radiated fuel assembly component material
{
}
properties at 70*F.
Provide -information on 3
allowable fuel assembly component stresses i
for irradiated materials at the operating temperature of 140*F.
Clarify the ASME Code edition and addenda that Rev. 3, Section 2.
were used for design.
1 What is the SFP water temperature alarm' Rev. 3, Section 4.2.
setpoint?
I
.i In PGE-1013, Amendment 2, the Quality Rev. 3, Section 7.0.
1 Assurance Program referenced ANSI Standards, except N45.2, have been deleted. Did the deletion of :these standards af fect' material
~
i testing requirements?.
,!Ob A-15 Revision 3 (December 1981)
I
~-
d Summary of NRC Request PGE Documentatica of
-s(
for Additional Information Compliance (PCE-1013)
What is PGE 's intent with respect to the use Rev. 3, Section 3.1.
of the design details shown in Figure 3-3a of PGE-1013?
Refer to the PGE letter to the NRC, Rev. 3, Section 3.1.1.
November 8, 1977. Provide information on the reference used to establish allowable stresses for the ARMCO 17-4PH stainless-steel material used for the top tie-bolts and module-threaded i
feet components.
With reference to existing plug and seam welds Rev. 3, Section 3.1.4.
in the spent fuel pool floor liner plate, it is stated, "The vendor drawing for the liner plate installation specified that these welds W
i were to be ground, but did not indicate the final contour of the velds in the cases".
Please clarify the weld grinding requirements for the initial liner plate installation.
i How do the procedures used to perform the plug Rev. 3, Section 3.1.4.
and seam weld grinding provide for control of heat generat(d by the grinding process, and effects of the grinding performed?
Will PGE repair all plug and seam weld ground Rev. 3, Ser!!on 3.1.4.
areas that do not pass liquid penetrant inspection?
Are other areas (other than plug and seam Rev. 3, Section 3.1.4.
weld areas) where the fuel pool liner plate is ground also inspected by the liquid P
penetrant method?
O A-16 Revision 3 (December 1981) i
Summary of NRC Request PCE Documentation of for Additional Information Compliance (PGE-1013)
,- s
)
V Provide a criticality analysis to confirm that Rev. 3, Section 3.1.2.
a dropped fuel assembly leaning against the pool vall (with its base adjacent to the rack bottom) will result in an artay for which k gg e
is less than 0.95, or provide additional measures (eg, deflectors) to assure that a dropped fuel assembly in any possible configu-ration wi:1 remain 6 in. away from the rack sides.
As discussed in the Safety Evaluation issued Rev. 3, on November 11, 1977, minimum boron concentra-Section 3.1.2.1.
tion in the SFP water and decay times for hh spent fuel should be proposed in the event that the re-rack evolution should take place after the first refueling. These measures
[J
)
would provide assurance that k gg would e
u.
remain below 0.95 and the potential radiologi-cal doses below 10 CFR 100 guidelines in the event a seismic event should cause racks containing spent fuel to be upset. As an alternate, an analysis demonstrating that the seismic restraint capability of both old and new racks would be maice.ained is acceptable.
Provide inplant radiation exposure estimates Rev. 3, associated with the spent fuel storage rack Section 5.2.1.6.
work.
A (v)
A-17 Revision 3 (December 1981)
APPENDIX B f'
~
i
)
V SLMMARY OF MODELS Figures B-1 through B-5 show the finite element models used to determine module member stresses resulting frem static and dynamic loading con-ditione. Fuel assemblics are considered to add mass but no stif fness.
The peak SRSS resultant module feet reactions under seismic loading were used at input to the model shown in Figures B-6 and B-7 to determine max-i=um locating frame member stresses and embedment loais. The locating frane plates will easily span the existing leak detection channels and transmit maximum module reactions.
Figure B-8 shows the model used to determine top tie-bolt loads. The tie-bolt loading condition used in the design conservatively considered the maximum " diaphragm shear" in the critical direction of translation to be resisted by only those bolt assemblies oriented normal to each direction of motion, analyzed separately (thus, about 1/2 of total
{}
f-s tie-bolts available were considered eff 3ctive in each horizontal di ec-(
)
\\._/
tica). This approach was determined in the design to be conservative based on bo.h direct shear and torsional shear tie-bolt analyses of individual rack medules. A planar analysis of the model shown in Figure B-8 was done for the.iorth-south direction (X axis), which produces :he highest tie loads because of the shorter model base.
These tie loads were then applied to the detailed three-dimensional SAP IV models of the racks in both north-south and east-west directions, thereby conservatively considering the three-dimensional effects of the tie loads on each rack.
The rack modules and ties were designed accordingly to transmit these maximum loads.
l Subsequent torsional analysis of the entire fuel rack array under maximum eccentric loading conditions has shown that the naximum vec-torial combf.ned tie-bolt shears are about 1/2 the values used in bh the design approach. Thus, torsional conditions result in tie-bolt loads appreciably less than those used in the design.
/~m i
I sv B-1 Revision 3 (December 1981)
i l
The top elevation of the modules is approximately 15.5 f t above the fuel pool f 2 0or and the minimum water elevation is 24 f t above the U
/~
k 'h/
floor. The rinimum water cover is thetefore 8.5 f t and the racks are well below any t arface wave effect induced by seismic activity.
Therefore, no additional loads are generated by surface water LIcc hing.
In the dynamic calculations of the submerged module, the total " virtual" horizontal mass conservatively included all water trapped within the module and the volume of water trapped between the cavity face and the wall of the pool in the direction of motion. The added mass of the water is equal to approximately 30 percent of the dry mass of the fuel and the module. Only the submerged retural frequencies were calculated.
The first eight modes of vibration were analyzed which encompassed the first and second modes of vibration in each of the three spatial coordinate directions. For the modified 7 x 8 module (see Appendix C),
calculated modal frequencies in ascending order were:
8.0 Hz, 8.2 Hz, 12.3 Hz, 15.6 Hz, 16.0 Hz, 16.5 Hz, 17.2 Hz, and 17.3 Hz. Results of the
[N analyses show that only the first three response modes contribute
\\
'I t
s1 nificant stresses in the fuel rack components.
3 In addition to the 30 percent increase in mass, a 2 percent increase in damping was used in the analyses to account for changes in response due to the submerged conditions of the fuel rack (
The design conservatively considered the trapped water within cavities displaced t
water between cavities, and displaced water between the fuel racks aad pool wall in the direction of motion all to be effective over the full height of the racks for the virtual water mass determination (about 30 percent of the total weight of fuel and rack structure). A 2 percent increase in damping, based on the referenced experimental data, was also g) used in the design Jr. consideratier of the underwater response cundi-tions (hydroolic drag, etc, on the rr.cks).
The experimental data indi-cate that the amount of water mass that should be considered effective is a direct function of the f eel rr.ck response displacements relative to the water mass. Accordingly, since the fundemente.1 translational mode
('~N, shapes are essentially linear (supported cantilever displacements), the V
B-2 Revision 3 (December 1981)
virtual water mass used in the design is approximately twice use amount
/~^)
suggested by the~ references.
If a virtual water mass proportional to response displacement is used, and no credit is given for increa:ed dampi.m due to submergence ' response equivalent ta that in air),
calculated fuel rack member stresses would be increased between 10 and 15 percent. All materials for the SFP racks have been delivered to the fabricator, and rack fabrication is proceeding. Ma*.erial mechanical property certificates show actual material yield stresses varying from a minimum of 37,000 psi to a maximum of 51,000 psi in comparisan with
_u the nominal code value of 30,000 psi.
rherefore, design allowable stresses based on actual rack material minimum yield stresses, with temperature corrections included, can be increased at least 20 percent and no changes to member section properties are required.
The analysis of fuel rack seismic response, considering that no credit is given for increased damping due to submergence, shows that the stresses resulting in the welds and the top tie-bol's are within acceptable code limit. No credit was taken for actual material mechani-
/]
cal properties.
OBE response spectra at 4 percent dampi:*g and SSE response spectra at 6 percen; dampice were utilized. Comparison of these spectra show that in the range of fuel rack response frequencies, spectral values are nearl identical. Therefore, because of lower allowable stresses, the load combinations which include the OBE governed design.
I The seismic modal and spatial responses were combined in accordana i
with NRC Regulatory Guide 1.92, Revision 1, as described further in b
Appendix C.
Load cotabinations and allowable stress limits used in the design are in accordance with NRC Standard Review Plan, Section 3.8.4.
Material properties for Type 304 stainless steel at 140*F were used in
= 28,000 paa the analyses. Specifically, a material yield stress of Sy 6 psi, were used.
and a modulus of elasticity, s = 28 x 10 The effects of inpact, under a seismic occurrence, of fuel assemblies l
inside fuel rack cavities were analyzd by conventional energy balance l \\
/
B-3 Revision 3 (December 1981) i t
-=
metheds similar to those described in ANSI Committee 20.2 Draf t Report,
" Design Basis for Protection Against Pipe Whip" (June 1973). The equivalent static force, R, developed by a fuel assembly impacting against a fuel rack cavity, can be described by the following equation:
4
~
1
~
aa m " "tA 6,(u-1) + g 1
2p _
4 where M
the total mass of fuel and fuel rack structure per cavity =
=
t 2638 lb mass i
A = the maximum seismic acceleration of the fuel rack calculated f rom the response spectrum when no gap exists between the fuel and fuel rack cavity C
A = maximum cavity bundle gap (fuel considered to, be tight against a corner of the cavity opposite the direction of motion, ie,
,,,s
}
maximum potential energy) = 9.020 in. - 8.424 in. = 0.596 in.
a = ratio of mass (weight) of fuel, M, to total mass (weight) of fuel and cavity, M, = 1616/26$8=0.613 6 = limiting elastic horizontal displacement a: module center due - to uniform load along its height. This displacement was determined in the finite element analysis of the module by applying a static uniform load and increasing the module center horizontal displacement until the module diagonals (limiting elements) were at yield:
6 - 0.395 in.
p = allowable material ductility ratio = 10.
The equation was derived by summing:
(1) the virtual work produced by the inertia force of the fuel assembly mass acting through the maximum gap between cavity and fuel, M Aa (here the module peak response f
acceleration, A, was used considering the fuel as a rigid body); and (2) the virtual work produced by the inertia force of the combined fuel B-4 Revision 3 (December 1981) 2
and ravity structure csss acting through the limiting elastic horizontal module displacement, M A6. To preserve energy balance (conservatively neglecting energy lost during impact), this total virtual work must be equal to the area under the force versus deformaticn curve, R (6 - 6e/2),
representing the internal strain energy absorbed by the fuel rack module (6 - maximum displacement = u6-).
e Numerical substitution into the above equation gives R,= M A(1.150).
t The ef fect of a feel bundle impact on a storage rack cavity is then 1
calculated to be equal to 15 percent of the seismic response that would occur if each assembly were considered to be rigidly attached to each ctvity. This 15 percent increase was added to the peak fuel rack l
seismic response (which considered fuel assemblies :-igidly attached to all cavities) to include fuel impact effects.
It is'noted that tie t s impact force is equal to the inertia force of the fuel bundle acceler-ated to about 90 percent of the maximum loaded fuel rack acceleration.
The fuel rack module components were designed to resist these together w
(%
with otner static loads.
\\
Probable maximum local effects of fuel mssembly impact on a cavity were also evaluated. Because of the relation between the natural frequency-of the rack and the duration of the fuel assembly impact, it is statis-tically icprobable that e.uh fuel assembly would impact simultaneously and in the same direction.
By representing the probability of this l
occurrence by the SRSS metnod, the equivalent static force (representing the impact of a fuel assembly on each cavity) would have to be approxi-(
mately 60 percent of the maximum seismic response to have the 15 percent overall effect as calculated, ie, 7 __(F)2 + (0.57F)2__
1
=. 15 The localized fuel impact ef fect is equivalent to a static force of approximately 400 lb at the top of each cavity. Static force tests have been made on fuel impact portions of individual cavities for concentrated loads of up to 20,000 lb without exceeding localized yield
%)
l B-5 Revision 3 (December 1981)
I I
stresses. Therefore, local and overall fuel bundle impact effects have
[N been considered and provided for. in. the fuel rack design.
(
Pursuant to the NRC request for clarification of the effects on the fuel elements resulting from the governing seismic event (SSE), supple-mental analyses were perfctmed in addition to those previously described.
For the analyses, the mathematical model considered the fuel assembly to be initially located eccentrically inside the fuel rack cavity such that the maximum gap of 0.596 in., uniform over the assembly height, existed between the assembly and the opposite rack cavity wall. The appropriate Westinghouse eight grid fuel assembly mass and nonlinear stiffness properties were used in the model. Discrete loads, representa-tive of seismic inertial effects, were mathematically applied to each mass point and iterated in increments up to values resulting in the total impact reactions bounding the results described. Two separate analyses were performed using the WECAM finite element computer code.
(
The first analy;is was performed for two iterative grid load distribu-b b]/
tions corresponding to a total impact reaction of 400 lb:
(1) uniform rec tangular load distribution, and (2) inverted triangular load distribu-tion (maximum at top of assembly). Results in terms of ratios of allowable stresses to maximum calculated fuel assembly element stresses are presented in Table B-1.
The deflected shapes for both cases are provided in Figure B-9.
The second analysis was performed for two sets of iterative inverted triangular grid load distributions corresponding to total impact reactions of 8f0 lb and an arbitrary upper limir of 1090 lb.
For this second analysis, only the inverted triangular load distribution was used becsuse it was considered to be a better representation of the assembly discrete mass inertial force distribution corresponding to the assembly fundamental mode response. Results of this second analysis are presented in Table B-2.
The deflected shapes for both loading conditions are provided in Figure B-10.
O B-6 Revision 3 (December 1981)
i J
l The conclusions reached by the supplemental analyses performed, as e
described herein, are t' it governing fuel assembly stresses due to 1
seismic (SSE) induceo
' pact in the fuel rack are below allowable i
nu values by a factor of abo t two.
The reaction at the foel assembly v
base nozzle is air-less than that required to cause fuel asseebly sliding, and ti...,
the mathematical model used is a valid representation.
1 i
1
?
i i'
i I
h i
i t
t i
4 1
4 j
l I
r i
B-7 Revision 3 (December 1981)
, =.
-~..
.~.
~
I l
i i
i TABLE B-1 i
RATIO OF ALLOWABLE STRESS TO FUEL ASSEMBLY COMPONENT MAXIMUM STRESSES i
FOR TIIE UNIFORM LOADING CASE AT 400 LB a
1 t
Allowablell } Stress Allowablell} Stress l
Limit (Pg)[2]
Limit (Pg + P )E I B
f Loading Case Component Uniform Stress (og)
Combined Stress _(og + og) 6 Uniform loading Thimble 5.75 1.97 I
i l
Uniform loading Fuel rod 174 32.4 l
C i
Triangular loading Thimble 7.52 2.53 1
i Triangular loading Fuel rod 266 40.7 i
I 70*F.
[1] Based on unirradiated properties at 4
[2] Explanation of symbols:
1 I '9E
'g = Maximum allowable membrane stress P
Q$
PB = Maximum allowable bending stress i
EO og = Calculated membrane ~ stress n
+
h,
'og Calculated bending stress.
i r
1 4
i.
f' i-1, 1
t l'
i TABLE B-2 1
'(~)
I RATIO OF ALLOWABLE STRESS TO FUEL ASSEMBLY COttPONENT MAXIMUM STRESSES FOR TRIANGULAR LOADING CASE AT 850 AND 1090 LB
-Allowablefl} Stress Allowablell} Stress j
Linit (r3)[2]
Limit (Pg + Pg)l2}
Load (1b)
Component Uniform Stress (og)
Combined Stress (oM + CB)-
850 Thimble 4.75 2.05 i
850 Fuel rod 260 33.1 1090 Thimble 4.75 1.96 1090 Fuel rod 260 38.6 t
i l
[1] Based on unirradiated -properties at 70*F.
4
[2] Explanation of symbols:
f'~N Pg = Maximum allowable membrane atress P3 = Maximum allowable bending stress og = Calculated membrane stress og = Calculated bending stress.
J.
l l
6 l
l I'.
t.
i t
Revision 3 (December 1981) f a
- _ -. _ - _.. _.... -, _. _.., -. _..,, - _., _ _., _ _ _ -. _, _ _.. _ _ _ _.. _.. _ _..... - - -., _,... _ _,... ~. _ _, - -.. - _.
TN TN l
\\
i 3RIO I GRID I k
j N
N
\\
GRID 2 GRID 2 N
GRID 3 GRID 3 g
N N
N GRID 4 GRID 4 N
GRID 5 GRID 5 N
N GRID 6 N
GRID 6 N
N N
GRID 7 GRID 7 N
N N
N GRID 8 GRID 8 l
N BN BN O
05 1.0 O
O.5 1.0 a) Triangular Load b) Unifom Load Figure B-9 Fuel Assembly Deflected Shape Revision 3' 400 Lb Reaction Load (December 1981)
.. _,... _ _ _ _.. _., _.. - _.. _ - _. _ - _. _ _ _,,. _ -. _ _ _ _ _ _..... _ _,.. -. _ _,.. _. _ _. _ _.. ~ _. -.. _ _ _. _ _ _ -. _,. - _..
/
TN TN GRIO I N
GRID l I,
I i
N I
3 I
l N
i i
I GRID 2 -
GRID 2'-
I s
N GRID 3 GRID 3 N
N GRID 4 GRID 4 N
u N
~
GRID 5 GRID 5 i
N N
GRID 6 GRID 6 N
GRID 7 GRID 7 N
GRID 8 GRID 8 N
BN BN a) REACTION FORCE = 850 lb b) REACTION FORCE = 1090 lb Figure B-10 Fuel Assembly Deflected Shapes for Revision 3 Triangular Load Distribution (December 1981)
APPENDIX C i
\\- '
SUPPLEMENTAL SEISMIC ANALYSIS OF 7 x 8 RACK MODULE The initial analysis and design of the SFP racks were {wrformed using the criteria specified in ~frojan FSAR Sections 3.7 and 3.8 for load combinations and allowable stress limits, and for methods of combining modal and spatial seismic responses. Subsequent reanalysis was per-formed to assure compliance with load combinations aad allowable stress limits specified in NRC Stendard Review Plan, Section 3.8.4, and compliance with NRC Regulatory Guide 1.92, Revision 1, for combination of modal and spatial seismic responses. This appendix summarizes results of the reanalysis.
Rack component member properties are the same for all modules (see beam section properties, Figure B-1).
The 7 x 8 module, which is the largest and most heavily loaded, was found in the initial analysis
{}
to govern design of the most highly s:ressed module members. The e'N reanalysis was therefore performed for the 7 x 8 module and for the
(
i N s' locating frames. Model characteristics used are as described in Appendix B.
Funor structural modifications of the 7 x 8 rack were found to be necessary as a result of the reanalysis. Stif feners were added to the module legs (Figure B-1, beam section properties, Section 9), and gusset plates were added to the base 12-in. channel (Section 5).
The channel with added gusset plate is referred to as l
Section 10.
These modifications have been incorporated on all of the rack modules, ie, 6 x 7, 6 x 8, 7 x 7 and 7 x 8.
In the redesign, member slenderness ratios in accordance with the AISC Code, 7th edition, recommendations were utilized.
l l
Table C-1 summarizes the stress results for the reanalysis of the modi-fled 7 x 8 reck.
The combined stresses in the fuel rack members for the governing load C3 s-l ('~x combination, which includes fuel bundle inpact, are within code allowable L./
l C-1 Revision 3 l
(December 1981) 1 l
a limits. The combined stress interaction ratios are illustrated in
+
Table C-1.
Fuel bundle " rattling" impact stresses in the fuel rack members have been accounted for in the design as described in Appendix B.
The ef fects of this impact on the fuel bundles do not otherwise enter into the design of the fuel racks and, therefore, have not been analyzed as part of the modification design. Ilowever, comparisons show that translational mode response accelerations are nearly the same in the new racks and existing racks, and " rattling" impact effects on the i
fuel should be essentially the same in either of the rack configurations.
Generic impact analyses and test data for Westinghouse prototype 17 x 17 fuel bundles (iJentical te Trojan's fuel), per Westinghouse Topical Report WCAP-8288, show that fuel bundle component stresses and impact i
forces due to a simultar oous LOCA and seismic event (having a peak horizontal acceleration of up to 0.4 g) indicate that the fuel bundle
~<
w.
('"N design is structurally acceptable based on established allowable
\\--
design limits.
Fuel bundle impact forces determined in the analysis of the new spent fuel racks for the Trojan SSE, having a peak horizontal acceleration of 0.25 g, are much less than those used in the Westinghouse generic impact analyses.
Unirradiated material properties at 70*F were used in the analyses because they are the more conservative values; ie, they result in lower allowable stresses. Radiation affects the zirconium alloy so as to increase its yield strength. There is some attendant reduction.in ductility tr.at occurs due to irradiation; however, 'with the calculated stress -levels being on1? ibout.1/2 of the allowable values and within i
the clastic ' range, a renaction in ductility is not of concern. The B
allowable stresses used in the stress datio comparisons presented are considered to be conservative. Unirradiated properties of zir-
"'N[
conium alloy were also _ used to establish conservative allowable C-2 Revision 3-(December'1981) 4 m.
. m.-
stress limits for normal operation and/or worst-case combined seismic and blowdown fuel assenbly stress analysis as described in Trojan FOAR
^w Section 4.2.1.1.2.
i l
l l
J C-3 Revision 3 (December 1981)
,,, _. _.. _