ML19242A088
| ML19242A088 | |
| Person / Time | |
|---|---|
| Site: | Brunswick  |
| Issue date: | 06/12/1979 |
| From: | Bevan R Office of Nuclear Reactor Regulation |
| To: | Ippolito T Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 7907310218 | |
| Download: ML19242A088 (13) | |
Text
.
- arc
+
()
UNITED STATES 8'
I,j NUCLEAR REGULATORY COMMISSION y
g C
W ASHINGTON, D. C. 20555
\\
- /
JUNE 1 2 1979
.,g Docket Nos. 50-325 and 50-324 MEMORANDUM FOR:
Thomas A. Ippolito, Chief, ORB #3, D0R FROM:
Roby B. Bevan, Project Manager, ORB #3, D0R
SUBJECT:
MEETING
SUMMARY
A meeting was held with representatives of Carolina Power and Light Company (CP&L) and their contractors in Bethesda, Maryland on May 30, 1979.
The purpose of the meeting was to provide the staff with the current status of the seismic ipe stress reanalysis being conducted rcr Brunswick Units 1 and 2 required for IE Bulletin 79-07, and to review the licensees program of identifying and modifying overloaded pipe supports and hangers fcund incident to the reanalysis.
A list of attendees at the meeting is enclosed.
Also enclosed 's a CP&L letter dated May 29, 1979, with attachments, addressing the it'as to be dis-cussed at the meeting.
On May 25, 1979, CP&L infomed DOR staff by telephone that, in the course of their reanalysis for IE Bulletin 79-07, they had found six rapports stressed beyond code allowable.
On learning of this situation, they shut down both units and infomed the D0R staff and I&E of their action.
By telephone on May 28, 1979, CP&L requested this May 30, 1979 meeting.
At the meeting, CP&L explained that the six supports had not been adequately designed to account for the torsion stress on I-beams, a condition that might exist for sone other supports in the plant.
They had, therefore, initiated a program to investigate all pipe supports on all safety related systems.
This program would be completed and all modifications made, before the two units would bepn a return to power operation.
CP&L also described modifications in their reanalysis program, both for already-analyzed and for to-be-analyzed cases, to accmmodate the staff position regarding the BSEP Final Safety Analysis Report commitment to use absolute summation techniques with the two dimensional model.
The Staff agreed to review their proposal, and infomed them by phone on June 1,1979 that their method is an acceptable one.
CP&L has previously committed (letter dated May 22, 1979) to verify as-built dimensions fcr all safety related piping in and support systems.
They reported that this program of " walking the lines" is in proaress. During the current shutdnwn, all lines inside containment will be verified.
The reanalysis of the lines is being done concurrent with the as-built verification.
CP&L described procedures for handling deviation of as-built frm as-analyzed.
7907 31 cms 42 269
Thomas A. Ippolito JutiE 1 2 r379 In response to a previous staff inquiry regarding the location of the postulated LOCA pipe rupture relative to the highest stress point, CP&L informed the staff that the piping design did not locate the break at the highest stress point, but instead analyzed for a double ended or a longi-tudinal break to occur at any point on the line, both inside and outside containment.
In response to a previous staff request, CP&L discussed their program of management controls and reporting criteria to assure appropriate licensee action when problems are identified in the continuing program of pipe and support reanalysis.
CP&L expressed their intention to return to operation in a few days.
They therefore req Jested (and we agreed) to meet with us again on June 4,1979 to review their status.
Specifically, they would at that time provide *.he current status of their pipe and supports analy;es, identifying modifications yet to be completed (if any), and verifying the as-built condition of all lines procedures for handling deviations found in their ongoing as-built verification arogram.
0!
P w
Roby B.
evan, Project Manager Operating Reactors Branch #3 Division of Operating Reactors Enclosure _.
As stated e 6,'l
'l i U r
ATTENDEES MEETING WITH CAROLINA POWER & LIGHT COMPAfV_
BRUNSWICK 1 AND 2 PIPE & SUPPORTS REANALYSIS May 30, 1979 CP&L M. Connor, J r.
W. Kincaid B. Furr P. Howe D. Bensinger GE N. Shirley UE&C L. Kreider G. Rigamonti B. Huselton liRC R. Bevan K. Wichman A. Lee L. Modenos, Region II J. Fair T. Ippolito H. Wong J. Glynn n ' 'g y
,1 !
~
,} cj Qy 9/O a Dt u iI L p.e. g
.c.'y./ % c.
, G cr
# I E f
' L~m..
s/f).
'z
/
3 C a r o h na p ov. e r a ag*.t c u, c.
File:
NC-3514(B)
Ma y 2 9, 1979 SERLAL:
CD-79-1401 Office of Nuclear Reactor Regulation ATTEhIION:
Mr. T.
A. I p po li t o, Chief Operating Reactors Branch No. 3 United States Nuclear Regulatory Commission Washington, D.
C.
20555 BRUNSWICK STEAM ELECTRIC PLANT, UNIT NOS. 1 AND 2 DOCKET NOS.
50-325 and 50-324 LICENSE NOS. DPR-71 AND DPR-62 SEISMIC ANALYSIS OF SAFETY-RELATED PIPING
Dear Mr. Ip po li to :
At our meeting on May 21, 1979, Carolina Power and Light Company crueltted to provide the NRC Staf f additional inf ormation concerning our re s po nse to IE Eulletin 79-C7 on seismic pipe stress analysis.
)n May 23 and 2 i, 1979, the Staff identified to us, by telephone, and to a representative of Uaited Engineers and Constructors, our architect engineer for the Brunswick Steam Electric Plant, several additional items taat should be addressed in our r e s po n s e.
The remainder of this letter and attachments respond to those requests.
1.
The analysis of the loads for the pipe supports for the first ten (10) lines reanalyzed for pipe stresses has shown that there were ten cases where the load exceeded allowable.
Table 1-1 s umma rize s the data on the 98 pip 2 supports on the se te n (10) lines.
Table 1-2 presents the details of the ten (10) supports that were overstressed.
While evaluating these ten pipe supports, it was cetermined that the supports had been underdesigned initially.
1r no case was the overstressed condition a result of the.:w lord from the seismic reanalysis.
As shctm on Table 1-2, the new load a:tually decreased in five cases, increased less than 2.5% in four cases, and increased 19% in only one case (which was already over capacity by 14.5%).
These ten supports were analyzed to determine if their struc tural integrity would be maintained under the identified loads.
Four of these supports were found to maintain stresses less than yield and thus would maintain structural integrity.
Khen it was determined that structural integrity would be compromised f or the six supports under the calculated loads, Carolina Power & Light Company decided to shut down both units and sake aecessary modifications to ther.e supports to reduce stresses to less than allowable.
These modifications have been initiated and the new capacity is <:hown on Table 1-2.
pA9
),,
eom FE "." U*i. NT O I""_*U ~ *.O
'*EvdJ
2-SERIAL:
CD-79-1401 File:
NG-3514(B)
~I)uring this evaluation, it was noted that the overloaded pipe, supports f ailed in two ways:
either concrete anchors or in
~
'ttWETmi-~YIhvestigation was befun to 1olk at all_ pipe supMs on'Bf6ty reldreid Tystems_to%erming if similai overloaded ~coiditions may exist under the original load. T e results7f d!Js ' investigation will be availaI> lion June.f, and all necessary modifications vill be made prior to returning the g k f[0.a dfe% 4-2. lye Ef/Jw (
units to operation.
d 2.
On May 24, 1979, the Staff informed us by telephone that the seisnic stress analysis should be based on absolute sum if a two-dimensional seismic analysis was used, and that the square root of the su= of the squares (SRSS) was acceptable if a three-dimensional seismic analysis was made.
The Staff further stated that a stress f rom a two-dimensional analy sis calculated using SRSS and multiplied by a factor of 1.38 would be acceptable.
At the time that BSEP was licensed, two-dimensional h'
N SRSS seismic analysis was acceptable criteria, and it is not a pparent to us that the back-fit of a two-dimensional absolute sum seismic analysis has undergone the necessary requirements r
of 10CFR50.109.
Although CP&L does not accept the Staff's
)
po11 tion, we have prepared a revision to Table 2 of our letter h
o f May 21 demonstrating the ef f ect of multiplying the N1 tao-dimensional analysis results by the 1.38 factor.
We have h#
also taken credit for conservatism that exists in the relationship between the OBE and the DBE.
The results of this exercise show that only one line of the first thirty-nine reanalyzed lines exceeds total allowable stress by 2%.
For this line, the total stress is still less than 0.9s.
y
~sW For the unreanalyzed lines _shown in Attachment 3 of our May 21
,p le tter (GD-7 9-1342), we have used the 1.38 f actor to establish criteria for priority of lines to be reanalyzed.
We do not plan v
to base our conclusions of acceptability on the use of the 1.38 factor, since it is not the appropriate criteria for BSEP.
In determining the criteria for priority of reanalysis of the remaining lines, SRSS stresses were estimated on the basis of a factor of 1.5 increase, and this resultant was then multiplied by 1.38.
Credit for the conservatism of the OBE/DBE relationship was taken into account prior to applying the 1.5 increase.
When this was applied to the 411 lines that have not been reanalyzed, 39 of the 411 exceeded allowable stress, and are tabulated in Attachment 2 to this letter.
Our reanalysis priorities have been changed to include these 39 lines in those to be reanalyzed the week of May 28, and the results of *.his reanalysis should be available on June 1, 1979.
We still anticipate completing the total reanalysis in accordance with our previously stated completion date of July 21, 1979.
U2 2-4
Ippolito !by 29, 1979 Mr. T.
4.
3.
As a 'e s ul t of an I & E inspection at the Brunswick Steam Electric Plant to verify that the as-built dimensions were the same as the as-designed (as-analyzed) system, four deviations were noted.
These are discussed in Attachment 3.
As stated in the meeting on May 21, 1979, and confirmed in our letter of May 22, 19/9, Carolina Power & Light Company will verify ae-built dimensions for all safety related systems at BSEP.
This verification is currently in progress for those lines outside containment.
The lines inside containment will be verified at the next scheduled outage.
Due to the time constraints on reanalysis, the reanalysis is being condected concurrently with the as-built verification.
If any discrepancies are identified between the as-built /as-analyzed configurations, an evaluation by a stress analyst vill be made to determine if the line should be reanalyzed. Th' evaluation will be based on evaluating the magnitude of the amputed stresses for the area in question, and the impa.. (increase or decrease) on the stresses expected f or such deviation.
If it in determined that the line needs to be reanalyzed to determine tne new stress level, we will promptly reanalyze the line.
4.
During our recent meetings, the relationship of IE Bulletins 79-02 and 79-07 has been discussed.
Some of the pipe supports analyzed in the first ten lines are anchored using concrete expansion anchors discussed in Bulletin 79-02.
In the 79-07 support reanalysis, these base plates were and will continue to be analyzed using IE Bulletin 79-02 as a guide.
The capacity established for the concrete anchors is 20% of the manufacturer's rated capacity, Using this criteria, two supports on the first ten lines had to be redesigned and now have sufficient capacity.
As stated in item 1 above, the remaining supports using concrete expansion anchors are being investigated to determine their adequacy and will be reported on June 1.
A final report on all of our analyses and testing related to the concrete expansion anchors and IE Bulletin 79-02 will be submitted in compliance with the bulletin schedule.
5.
The Staff requested information on the location of the postulated pipe rupture for a LOCA relative to the point of highest stress.
The BSEP piping design did not use the mechanistic approach of locating the pipe break at the point of The postulated break for doubled-ended highest stress.
guillotine or longitudinal split was analyzed for the pipe break to occur at any point on the pipe, inside or outside c ontainme n t.
6.
We have been informed that during a meeting between NRC, another licensee and United Engineers & Constructors (UE&C),
some questions were raised by the NRC staff about the subject kh2
4-lby 29, 1979
!!r. T.
A. Ip po li to of valve operability.
In the event the staff may have any questions concerning this topic as it may apply to BSEP, we will be prepared to address this issue.
7.
Carolina Power & Light Company's c jritea for determining if an overstressed condition is rep 6ftable is set forth below:
a.
Lines Yet To Be Reanalyzed The stress using new Geismic data and revised analytical criteria are estimated for the lines that are yet to be r eanaly ze d.
As stated previously, those with high estimated stresses are being analyzed first in the reanalysis program.
We will not use estimated stress at a basis for determining overstressed conditions which are r epo rtable.
b.
Reanalyzed Lines Those lines which have been reanalyzed and which show an apparent overstress condition will be evaluated in detail to determine if it is a reportable item.
First, the known ccnse atisms will be removed from the analysis.
The line will be analyzed to determine if the stress at any single modal point exceeds FSAR criteria of 0.9S or 1.8 S e y
h whichever is the higher.
If the pipe remains overstressed, this will then be considered a reportable item and the NRC will be informed within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
c.
Reanalyzed Pipe Supports When the reanalyzed pipe data is available, the pipe supports will be reanalyzed f or the revised load.
If the load exceeds the apparent support capacity, the specific support will be analyzed in detail to determine if the stated capacity is the actual capacity without exceeding 0.9 S.
If the load still exceeds the capacity, a y
determination is made if the support will maintain s tructural integrity even if the allowable is exceeded.
If structural integrity is maintained, this is not considered reportable.
If structur il integrity is not maintained, the support is taken c.t of the computer piping configuration, and the line is reanalyzed.
The results of this reanalysis are evaluated to determine if other supports and the pipe can take the additional load without exceeding their structural integrity.
If the system maintains integrity, the item is not re po rtable.
If the system does not maintain structural integrity, the item will be reported within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
- c n
May 29, 1979 M. T.
A.
Ippolito In sum =ary, CP&L has evaluated the data f rom the lines reanalyzed to to be reanalyzed, date, and the estimates for revised stresses for lines yet and it is our "onclusion that the continued operation of the Brunswick Steam 1 & 2 is warranted without undue risk to the public Electric Plant, Units The health aad safety, while the reanalyses of seismic design continues.
problem associated with those supports ?. hat were found to be overstressed is a of initial underdesign of those supports, and is not related to the use r es ult sum of the square, or absolute summation of seismic square root of algebraic,The modifications of those supports which were originally stresses.
under-de;igned will be completed in early June, and at that time, both units will be returned to power.
As stateu in our letter of May 15,1979, and in 24-hour reporting criteria have been established if any item 7 of this letter, piping or supports are determined to be overstressed during the reanalyses.
If you have any questions concerning this information, please do not hesitate to contact our staff.
Yours very truly, d M (( (4 E. E. Utley
/
Executive Vice President Power Supply DLB/sg bec:
Messrs.
D. L.
Bensinger C.
S. Bohanan D. B. Waters / File NG/3514(B)
J. M. Johnson W. B. Kincaid S. McManus A.
C. Tollison, Jr.
C. W. Woods (LIS)
File:
BC/A-4 File:
B-X-0274 97h
,- S hDL L'
GD-79-1401 ATTACllMENT 1 PIPE SUPPORT A"ALYSIS An evaluation was performed on the pipe supports of the first ten lines that were reanalyzed in the seismic pipe stress rea talysis program. There are 98 pipe supports made up of snubbers, vendor catalog pipe supports, and fabricated supports.
The recalculated loads comparea to tne original load and support structural capacity are tabulated on Table 1-1.
As can be seen on Table 1-1, the load did not increase appreciably due to the seismic stress reanalyses and recalculation of loads.
The load decreased for 30% of t.e supports and increased less than 25% for 607l ef the supports.
The load increased greater than 25%
for only nine supports, but the new loads were less than 75; of capacity for these supports.
However, ten supports were found whure the load exceeded the applicable allowable.
Further investigation revealed that these ten supports were underdesigned init iall:..
For these ten supports, the new loads were less than the old loads in five cases, increased less than 2.5*.
in four cases, and in only one case, the increase was 19%.
These ten supports were analyzed in detail to determine if they would maintain their structural integrity under the specified loads even it they exceeded allowable.
This is summarized en Table 1-2.
In four cases, including the one where the new lond was 19; higher than the old load, the supports retained their structural integrity.
Six supports would fail.
The six supports that would fail under the specified load (old or new) were redesigned to have their stresses less than allowable. The new design loads for these pipe supports are shown on Table 1-2.
It has been concluded f rom the analysis of 98 pipe supports that the seismic stress reanalysis does not contribute to overloaded pipe supports.
However, it has been recognized that there is a potential for certain
^
supports to be overloaded due to an error in the initial design.
These errors have been found to be with concrete expansion anchors and with torsion of the beam support.
An investigation has begun to examine the pipe supports of the ether safetv-related piping for similar problems.
The results will be reported at a later date.
4 6L2 277
ATTACHMENT 2 SEISMIC PIPE STRESS ANALYSIS CRITERIA As stated previously, the original seismic analysis for pipe stress used algebraic summation within each mode. A reanalysis effort was undertaken for all safety-related lines usi..
the UELC - ADLPIPE-2 Computer Code which empicys the square root - sum-of-the-squares (SRSS) load combination within each mode.
The results of the rear.alyses, given to the NRC Staf f in our re-sponses to IE Bulletin 79-07, in letters dated April 24, May 15, and May 21, 1979, used the SRSS nethod.
On May 24, 1979, the NRC Staff notified CP&L that the use of SRSS with a three-dimensional seismic analysis was acceptabic, but for a two-dimensioral seismic analysis the absolute sum method should be employed within each mode. The analysis for Brunswick uses a two-dimensional seismic input approach.
At the time BSEP was licensed, the two-dimensional SRSS analysis was the accepttble criteria.
Therefore, the acceptability of stress levels shot.1d not be based on absolute sum.
However, to use the most conservative case for comparison purposes unly, the stresses calculated using UE&C - ADLPIPE-2 were multiplied by 1.38 (a number acceptable to the NRC Staff) to obtain stresses for the Operating Basis Earthquake (OBE).
As discussed in Attachnent 7 of our letter to the NRC GD-79-1342, dated
>by 21. 1979, the previous seismic analysis used a most conversative approach of relating stresses for an OBE to that for a Design Basis Earthquake (DBE), known today as a Safe Shutdown Earthquake 'SSE).
The stress's computed in the OBE were mult plied by 2 and used as the stresses for a DBE.
As discussed on May 21, 1979 with the NRC St.*f, our reevaluation of the OBE and DBE Amplified Response Spectra (ARS) indicates that the relationship between the two ARS in the frequency range that affects pipe stress is less than 1.2, and frequently less than 1.0.
However, a value of 1.2 has been selected for use to convert OBE stresses to DBE stresses.
For the thirty-nine lines already reanalyc e d, the conversative stresses for comparison purp;ses for the DBE and total are shown onJ able 2-1.
The DBE stresses in this table are calculated as follows: U DBE =
(7~ OBE x 1. 38 x 1. 2, whe r; CI~0BE is obtained using the UE&C - ADLPIPE-2.
For the lines yet to be reanalyzed, the stress for a DBE was estimated as explained in Attachment 7 to our May 21, 1979 letter using a factor of 1.5 to account for the highest expected increase in stress due tc the reanalysis for SRSS (within each mode) in lieu of algebraic sum (withfn each mode) and which is based on the data from the reanalyzed lines.
Por those lines identified on Attachment 3 to our May 21, 1979 letter, the stress for a DBE were estimated as follows:
(I OBE x 1.38 x 1.2 x 1.5
()DBE
=
Est.
Crig.
where CI OBE was computed la tae original analysis.
Those lines whose Orig.
estir.ated otresses exceeded allowable are tabulated on Table 2-2.
9'^
,f n
')
-- EVALUATION As can be sten from Table 2-1, one lina (RllR-60, fcsidual Heat Removal) exceeds the allowable (1.8 S ) by 1.7 perc2nt.
However, this stress is less than the stress equal to 0.9 Sy (32,400).
The BSEP FSAR allows the use of 0.9 Sy or 1.8 S whichever is greater, as the allowable g,
stress during emergency cc:idition (DBE).
Therefore, the stresses are acceptable for all lines reanalyzed.
Table 2-2 shows that 39 of 411 lines yet to be reanalyzed exceed allow-able (1.8 S ).
These stress values are not necessarily based on coinci-h dent point naximums, but rather the su==ation of maximum stresses for each individual loading.
It should be restated that these.ctresses are es'inated and that they were derived using a conservative factor of 1.5 to cover the maxicum increase expected for the reanlaysis (old algebraic. to new SRSS combination within each mode).
As discussed in Attachment 7 and shown on Attachment 8 of our May 21, 1979 letter, in over 58% of the lines already reanalyzed, the new seismic stress was.less than the original seismic stress.
In over 87% of the cases, the new stresses were less than 1.25 of the original stresses.
As discussed previously in our letter, Carolina Power & Light Company commits to placing these lines in the highest reanalysis priority categor'/, regardless of the priority category previously established on a furcticn and size basis.
It should also be pointed out that of the 39 lines estimated to be overstressed, 27 are 2" or less in diameter.
l D
u <-
< / 9
ATTACILT;T 3 AS-BUILT DRNJINGS As a result o f the NRC-I&i walk-through of approximately 67 pipe supports on safety relsted Lines, four discrepancies were identified:
1.
Iso netric 17 High Pressure Coolant Injection main pu=p discharge line above el. 18'-9" data point 45 does not agree with piping dra41ng.
Actual lccation of support is 9'-2" from valve F006 in lieu of 7.0 ' as shown on the analysis isometric.
Comment:
The analy:cd location has been reviewed by streas analysist and confirmed that the actual placemant of the support will have little or no effect on the results of analysis for the following reasons:
1.
The total maximum stress of the line is less than 507. of the code allowable stress, see attachment 2 2.
The placement of the support within approximately two pipe diameters of its analyzed position on this 14" Sch.120 pipe will not adversely effect the analysis.
2.
Iscnetric 20 Core Spray dats point 101 is located approxt=ately l'-3" closer to valve F015A thsn shown on the isometric.
Comment:
Review by stress analysist confirms that since the data po:nt is a snubber placing it closer to the valve is better than the original placement.
In addition the new placement will have no adverse effect on the stress analysis since the nra placement is in the same plane as analyzed.
b' GO t -
3.
1scr:etric 12 Reactor Core Isolation Cooling pu=p suction lines data point 272 vertical snubber is located on the opposite side of un elbow than is shown on the analysis isometric.
C o= men t : Review by stress analysist confirm that placement has no effect on the stress analysis.
The analysit program treats the elbow as a point in the model, therefore transfer from one side of an elbow to the other has no effect on the analysis re-sults as long as the snubber acts in the required direction.
Field check of the installation has verified that the snubber is acting in the correct (vertical) direction.
4.
isometric 18 Core Spray Pump suction line 2B, data point 236, is eleven inches closer to pump than shown en the analysis isome.tric.
Ccm=ent:
Review by stress analysist confirms that the location of the support within one pipe diameter will not adversely effect the stress analysis.
In addition the support is a sliding dead weight support which has no effect for seismic support.
It should be noted that in the above cases it has been determined by a stress analysist that there is no adverse impact on the pipe stresses. However, Carolina Power and Light has co=mitted to perform an as-built verification on all lines included in the reanalysis to increase the confidence that the as analyzed condition is consistent with the as-built condition.
Thiref additional supports have been checked by field personnel and no additional problems have been found. /43 0n fUL
/_ U