IR 05000260/1989031
| ML20247H530 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 07/13/1989 |
| From: | Cheng T, Terao D Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20247H523 | List: |
| References | |
| 50-260-89-31, NUDOCS 8907310051 | |
| Download: ML20247H530 (11) | |
Text
.
-- - - - -
_ -.
-
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_
__ _ _ _
_
_ _ _ _ _ _ _
_
_ _ _ _ _ _
_
-[
I
o,,.
UNITED STATES.
,- [
g
- NUCLEAR REGULATORY COMMISSION
-
tj WASHINGTON, D. C. 20555
,
$f
%.. **,,*
.
- UNITED STATES NUCLEAR REGULATORY. COMMISSION
0FFICE OF NUCLEAR REACTOR REGULATION l
TVA PROJECTS DIVISION i
Report No.:
50-260/89-31
'
Docket.No.:
50-260
'
- Licensee
. Tennessee Valley Authority 6N 38A Luokout Place
1101 Market Street
'
l --
Chattanooga, Tennessee 37402-2801 Facility Name:
Browns Ferry Nuclear Plant,. Unit 2 Inspection'At:
1. Bechtel Power Corporation (BPC)
San Francisco, California 2. General Electric Company (GE)
San Jose, California 3. Nuclear Regulatory Commissivn (NRC)-
Rockville, Maryland Inspection Conducted:
April 6 - June 28, 1989 Inspector:
- r4 [
7N8M9 u
ThomasM.Cheng,Teampeader Da te '
Consultant:
Tom Tsai
%
/[G[I
'
'O I Approved By:
OdVid Ierao, Cnief note
Engineering Branch TVA Projects Division Office of Nuclear Reactor Regulation l
8907310051 890717
PDR ADOCK 05000260 c,
G PNU
'
_ ______
___.__
__
_'
i
f
.-
.
,
'
.
Special Inspection Relating To Design Calculations for the Seismic Design Program 1.0 INTRODUCTION AND 8ACKGROUND As documented in NRC Inspection Report (IR) 50-260/88-38 dated April 19, 1989 (Reference 5.1), the staff conducted its first inspection of the Browns Ferry Nuclear Plant Unit 2 (BFN-2) Seismic Design Program. A total of 32 items were identified as summarized in Enclosure 3 of IR 50-260/88-38.
One of these 32 items (CSG-9) concerned the potential impact of the newly generated amplified response spectra (ARS),_ based on the updated building model, on the existing
,
evaluations of systems and components including the nuclear steam supply
system (NSSS). TVA committed to complete its assessment of this item before the restart of BFN-2.
During the meeting held on January 5, 1989 (Reference 5.2), TVA stated that the impact evaluation of the primary components of the NSSS (reactor pressure vessel, reactor internals and supports) will be completed using time history analysis with the El Centro 1940 earthquake time history as input.
Further-more, the evaluation would use damping ratios as described in the FSAR and the newly-developed dynamic model of the primary system. The staff found this approach acceptable.
TVA also explained that Bechtel Power Corporation (BPC)
would perform the seismic analysis, using the updated detailed primary system model provided by General Electric (GE) Company coupled with the reactor building model to calculate the seismic loads (shears, axial forces and moments). GE would then use these loads to complete its impact assessment of the NSSS components.
2.0 PURPOSE AND SCOPE During this inspection period, from April 26 thruugh June 28, 1989, the staff performed inspections related to the seismic analysis and the design interface between BPC and GE.
The purposes of the inspections were (1) to audit the analysis and calculations performed by BPC for generating the seismic loads of the primary system (including RPV internals), (2) to ensure that adequate procedures were followed to transfer the seismic loads from BPC to GE, and (3)
to audit the impact assessment results completed by GE using the new seismic loads obtained from BPC.
3.0 INSPECTION FINDINGS On April 26-28, 1989, the staff and its consultent conducted a special inspec-tion of the TVA impact evaluation of the BFN-2 primary system at the BPC office in San Francisco, California and at the GE office in San Jose, California.
The attendees of the meetings conducted in the first three days are listed in Enclosures 1, 2 and 3.
After subsequent discussions between the staff and TVA
.
_ _ _ _ _ _. _ _ _ _ _ _ -
.
.
,
,
to verify the validity of the unresolved issues and to establish the adequacy of the corrective actions for the unresolved issues identified during the inspection, an exit meeting was held on June 28, 1989 at the NRC offices in Rockville, Maryland.
The list of attendees at the exit meeting is con-tained in Enclosure 4.
As discussed in Section 2.0 above, the three major areas covered by this inspection involved (1) generation of seismic loads, (2) transfer of seismic loads, and (3) impact evaluation of the primary system. The following subsections document the inspection findings, staff's conclusions, and TVA's verbal commitments made during the exit meeting.
3.1 Generation of Seismic Loads The inspection of the seismic load generation was conducted at the BPC's office.
In calculating the seismic loads (shear forces, axial forces, and moments) for the evaluation of the primary system and structural components inside drywell, TVA and BPC used a multiple-stick lumped-mass structural model. The lumped-mass model included the reactor building structure, drywell, pedestal, biological shield wall, and a detailed reactor pressure vessel (RPV) model provided by GE.
The analysis used the damping ratios as specified in the FSAR and a time history analysis technique using the 1940 El Centro earthquake ground motion time history as documented in the FSAR. Due to the limitations (capability of handling the hydrodynamic water mass in the horizontal RPV model which gives rise to off-diagonal terms of mass matrix) of the BPC in-house computer codes (CE917, CE920 and BSAP) which were previously accepted by the NRC staff, a combination of computer codes, SASSI (fixed base version) and BSAP, were used in the analysis. As discussed during the meeting held on January 5, 1989 (Reference 5.2), the use of a multiple-stick lumped-mass model combined with the time history analysis technique using the 1940 El Centro earthquake ground motion time history to calculate seismic loads for the primary system evaluation was found acceptable to the staff.
The staff and its consultant reviewed the dynamic model developed by BPC, calculations, and analysis results (shear forces, axial forces and moments)
as well as the verification package of SASSI computer code and found that the l
calculations and the analysis results appeared reasonable.
The use of the I
SASSI computer code for generating the seismic loads of the structural elements (not including primary components) was found acceptable based on the small differences between the results obtained from SASSI analysis and those i
from the computer codes previously accepted by the staff (within 5%).
j l
In addition, the staff and its consultant also reviewed the comparison of the I
horizontal seismic loads for the RPV (not including RPV internals and attached components), biological shield wall, and concrete pedestal with the corres-ponding seismic loads documented in Figures 12.2-34 through 12.2-39 of the FSAR. As expected, the two sets of horizontal seismic loads were reasonably consistent with each other because the refinement in the RPV modeling was not I
expected to significantly change the responses of the supporting structures.
Although the analysis as originally described in the FSAR did not consider the flexibility of structural elements in the vertical direction, the staff found-2-I
_ _ _. _ _ _ _ _ _ _
_ _ _ _. _ _ _ _ _
- _ _ _ _ _
. _ _ _- ______
_
.
.
,
,
"
I i
i I
' that.the same assumption, theory, and computer code used for the horizontal I
analysis were also applied to the new vertical seismic analysis. Therefore,
.l the staff concludes that both horizontal and vertical seismic loads generated by BPC have been adequately calculated.
3.2 Transfer of Seismic Loads-After the seismic loads were calculated, BPC submitted to TVA, a seismic analysis report in which all seismic member loads (shear, axial forces, and bending moments) of the RPV, RPV internals, and components attached to the RPV were tabulated. After TVA's review and approval, this report was documented in TVA's RIM system (RIM No. B22 890425 108).
This report was subsequently transmitted to GE for its RPV (including internals and attachments) evaluation.
The staff reviewed the interface procedures and related documents used to transmit the loads and found that the seismic loads calculated by BPC were properly transmitted to GE for the primary system evaluation.
3.3 Impact Evaluation of Primary Systems The General Electric Company performed an assessment of the impact of the newly-generated seismic loads on the primary (NSSS) system.
The scope of the impact assessment. included an evaluation of 13 components:
fuel elements, stabilizer, vessel stabilizer bracket and adjacent shell, vessel support skirt and ring girder,' shroud support, top guide, core support, incore housing, control rod drive (CRD), orificed fuel support, CRD tube, CRD housing, and CRD housing support. When the primary system was evaluated by GE with the seismic member loads (shear forces, axial forces, and bending moments)
obtained from BPC, the component was considered qualified if one of the following methods was satisfied:
Method 1) The new seismic load or stress is less than the original seismic design load, stress or allowables specified in the FSAR.
Method 2) The new seismic load or stress is less than the design load o'r stress of the similar component in other licensed plants of the same vintage as Browns Ferry for which GE was the NSSS supplier.
This approach is referred as a " sister plant approach."
Method 3) The new seismic load or stress is less than the design capacity established generically by GE for the same component.
According to GE, all of the 13 components were qualified for the new seismic loads and the methods used for qualifying each of the 13 components are tabulated below:
-3-L
!
.
M
,.
.
.
.
.
!
q COMPONENT METHOD 1 METHOD 2 METHOD 3 Stabilizer.
X
. Stabilizer Bracket &
l Adjacent Shell X
RPV Support Skirt &
Ring Girder X
Vessel Shroud Support X
Top Guide X
Core Support X
Fuel X
.Incore' Housing X
'
Control Rod Drive (CRD)
X Orificed. Fuel Support X
Control Rod Guide Tube X:
CRD Housing X
CRD Housing Support X
Based on its review of GE's calculations, handouts, and design drawings, the staff concludes the following:
1) The staff was unable to review the original design basis and criteria for the Browns Ferry primary system, because Appendix J to the FSAR which documented this information had been deleted from the FSAR.
To alleviate the staff's concern, TVA committed to establish the reasons for the deletion (and when it occurred) and redocument the design basis and criteria used for the original primary system design prior to the next seismic design program inspection.
2) The staff and its consultant reviewed the three methods used by GE for the evaluation of the 13 primary components.
The staff's findings are summarized as follows:
(a) Method No. 1 is acceptable.
(b) Method No. 2 is acceptable provided:
-4-
.
_ _ _ __ _.____ _ _ _ _ _
_. _
__
- - - _ _ - - - _
_ _ - - - _ - - -. - - - _ - - _
_
.'
.
,
I
'
.
L
,
.
l, (1) The sister plant components are identical to the BFN components, i
(ii) the design basis and criteria used for the sister plant
!
components have been previously reviewed and accepted by the staff, and
,
I (iii) BFN design basis and criteria are more stringent or equivalent l
to those used for the sister plant components.
(c) Method No. 3 is acceptable provided that the GE report which documents the design capacity has been reviewed and accepted by_the staff.
3) The staff reviewed the model of the reactor vessel (including internals)
developed by GE model (TVA Calculation B22 890425 108) which was trans-ferred to BPC for the overall reactor building seismic analysis.
The staff's review found that restraints (modeled as spring constants K and chn)cretepedestal.were assumed by GE to tie the control rod drive (CR K
The effect of these restraints resulted in stabilizing the cantilevered portion of CRD housing and reducing the bending moment at the fixed end of CRD housing (the moment was reduced by a factor of ten when compared to the original design moment). The restraints (K and K )
were not shown on the original Browns Ferry design drawing (GE Dhawing
104R935). Because the model incorrectly assumed the existence of these seismic restraints on the CRD housing, the model was not a valid represen-tation of the as-built condition.
The absence of CRD housing restraints was subsequently confirmed by a TVA field walkdown.
Therefore, the use of the seismic loads generated by Bechtel Power Corporation (BPC) based on the incorrect model for qualifying the primary system was not acceptable to the staff.
As a result of this finding it is unclear to the staff when the error might have first occurred and on which RPV model the design basis and results documented in the BFN FSAR are based.
In order to clarify these uncertainties, the staff requested that TVA provide:
(1) the original seismic analysis results (reports) used to license BFN, (2) the revision date of the current FSAR and the version of RPV model documented in the current FSAR, (3) the revision date of the current Table C.0-5 of the FSAR and model used to generate the stresses tabulated
in this table (4) a copy of GE Report 22A2016, Revision 0,1 and 2, and I
(5) a copy of Appendix J originally found in earlier versions of the FSAR.
In a conference call, TVA agreed to provide the requested information for review when it is available.
In addressing this issue, TVA developed a modification consisting of a set of seismic restraints (i.e., K, and K,3 in the dynamic model) to tie the bottom flanges of the CRD housing to the reinforced concrete pedestal.
During the exit meeting, TVA presented its conceptual design of the restraints and comitted to complete the final details of the modifications and its installa-tion prior to restart of the plant.
The conceptual design of the modifications-5-
.
_.
,
h
-
'
.
l
. appears reasonable and the staff will review the final analysis and details of the modifications when available. The staff considers this inspection finding as an Unresolved Item (URI 89-31-01).
In addition, the staff will review the new amplified response spectra (ARS)
and seismic loads generated based on the K and K consistent with the new
modification of the CRD housing supports. yThe staff will also review the
'
impact of these newly generated seismic loads and ARS on the RPV internal and attached piping which were analyzed by the input based on the originally assumed K and K '
y
4.0 CONCLUSION Based on the review findings discussed in Section 3.0 above, the staff concludes that the coupled dynamic structural model including related calculations, use of computer code SASSI, and analysis results are acceptable. In addition, the staff finds that proper procedures were used to transfer the seismic loads from BPC to GE for primary system evaluation.
However, three open issues were identified as a result of the staff's inspections.
The open issues are as follows:
1) The methods used by GE for primary system evaluation are acceptable provided that additional information is needed to justify the adequacy of Method Nos. 2 and 3.
TVA agreed to provide this information when available.
2) Due to deletion of Appendix J from the current BFN FSAR, the staff was not able to review the original design basis and criteria for the primary systems. TVA agreed to redocument Appendix J and to find out the cause of deleting this appendix if possible.
)
3) When the dynamic model of the RPV, internals, and supports was developed, TVA incorrectly assumed the existence of seismic restraints at lower end of CRD housings. This inspection finding was confirmed by the original design drawing and a field walkdown. TVA is developing a modification for the installation of seismic restraints on the CRD housings and will
!
report to the staff when finalized.
This open issue is considered as an Unresolved Item (URI 89-31-01).
5.0 REFERENCES 5.1 NRC Inspection Report No. 50-260/88-38, dated April 19, 1989.
5.2 Summary of meeting between NRC and TVA held on January 5,1989 to discuss civil / seismic issues, dated March 22, 1989.
-6-
.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _
, _.
___.
,;.:
.
e
.
I
.4
.
ENCLOSURE 1
'
ATTENDANCE NRC REVIEW 0F BFN/WBN SEISMIC ANALYSES APRIL-26, 1989
,
NAME ORGANIZATION W..A Massie Bechtel/TVA 0. Gurbuz Bechtel/TVA Rick Cutsinger--
TVA/BFN D. R. Denton-TVA W.~S. Tseng Bechtel/TVA
.-
. Tom N.C. Tsai.
NRC Consultant
- Thomas M. Cheng NRC
'
,.
- _. _. -. - - _ _. _ _ - -
., - _ _ -. _ _ _ - - _. _ - -. -.... _... - - -. - _ _ _.. _..... - - _. - - -. _. -
- -. -.. - - -. -... ~.. -. _ _ _ -...
- - - _ _.. -. _ _ - - _. - - -
--
~
.
,.
.
_ _ _ _ _ - _ _ _ -
p
,-
'
v
.
- 5 i
I ENCLOSURE 2
,,
TVA/NRC/GE MEETING APRIL 27, 1989
?:
NAME ORGANIZATION ~
J. Wallach GE Nuclear-Energy:
N. Shirley.
GE-L&CS B. Koepke GE J. M. Ioakem GE Wen S.'Tseng.
Bechtel, SF, Civil / Structural Dept.
Wayne A.'Massie TVA BFN Site Licensing Don R. Denton TVA - CEB (Knoxville)
Thomas M. Cheng NRC i
Tom N.C..Tsai
.NCT Engineering /NRC Consultant Rick Cutsinger TVA/BFN Hank Nghiem.
GE.
l
I I
.
i
- -____--.-
,
-
_
.
- _ _.
_ _ _ _ - - _ _ _ _. - - _ _. - _ - - - _ _. - - _ _ _ _
-
1.,.*..:;
,
'
-
.,
,
...
'
ENCLOSURE 3-NRC REVIEW 0F GE/BFN RPV APRIL 28, 1989 NAME.
' ORGANIZATION
.,..
Wayne A. Massie BFN Site Licensing Thomas M. Cheng-NRC Tom N.C..Tsai NRC Consultant
. Rick-Cutsinger TVA/BFN
, Allen'R. Smith GE Licensing
,
James M. Ioakem.
GE Des. Engineering
,
Patrick W. Marriott GE Licensing John Wallach-GE-
'l
,
,
.
'l
.- -. -.- -.--. --- - -- - - -.. - -
. - - - - - - - -
_
.
_
_ _ _ _ _ - - _ _
'e f 4
.
}'
e
~~
.
..,
ENCLOSURE 4
-
,
,
EXIT MEETING JUNE 28, 1989 NAME ORGANIZATION Nancy Markisohn NRC/TVA/PD Wayne L.:Smathers TVA
.R ci k Cutsinger TVA John R. Rupert
.
'
. Wayne A. Massie TVA Thomas M. Cheng NRC/TVAPD John R.. Fair NRC/TVAPD David Terao NRC/TVAPD Gerry Gears NRC/TVAPD
!
r e
!
{
I
,
l
.
!
-
_ _ _ - _ _ _ _ _ - _ _-
-
-