ML20051S721
| ML20051S721 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 05/06/1982 |
| From: | Russell W Office of Nuclear Reactor Regulation |
| To: | Lainas G Office of Nuclear Reactor Regulation |
| References | |
| TASK-03-06, TASK-3-6, TASK-RR NUDOCS 8205170569 | |
| Download: ML20051S721 (15) | |
Text
./
c UN!TED STATES g
,j'y # y ;
NUCLEAR REGULATORY COMMISSION
- i f'
E WASHINGTON, D. C. 20555
'h((
May 05,1962 n..... ~j o
s f
RECEmyg HEMORANDUM FOR:
Gus C. Lainas, Assistant Director k
14 I98A -
a for Safety Assessment t uor Division of Licensing san a p %ac n
88
//
g FROM:
William T. Russell, Chief g
is Systematic' Evaluation Program Branch Division of Licensing
SUBJECT:
RECOMMENDATION FOR PROMPT 5% MAL NRC ACTION (ORDER OR ORDER TO SHOW CAUSE) TO REQUIRE CONTINUED SHUTDOWN OF SAN ONOFRE UNIT 1 PENDING RESOLUTION OF SEISMIC DESIGN DEFICIENCIES AND...-(Deleted) l On Monday, May 3,1982, Southern California Edison Company, presented new information on seismic design deficiencies at San Onc.fre Unit 1.
The licensee's seis,mic reanalysis of systems and components necessary to assure the integrity of the reactor coolant pressure boundary and systems necessary for safe shutdown predicts gross overstress in piping, piping supports and mechanical equipment anchorages and support structures.
For.a postulated Safe Shutdown Earthquake (SSE), represented by a 0.67 Housner Spectrum, the magnitude of the calculated overstresses are large enough to cause the licensee to predict structural failure of the overstressed c~omponents. A tabulation l
of the licensee's analysis results is contained in Enclosure 1.
The l
licensee's criteria for predicting structural failure are unrealistically l
high (e.g., piping failures are predi'cted at stresses greater than 5.2 i
times current ASME Code level D &1lowable stress and concrete anchor failures are predicted for loads greater than three times their pullout load) and exceed any seismic operability criteria previously used by the NRC (e.g.,
IE Bulletin 79-02,79-14). A lower criteria would identify many more modifications as necessary to insure structural integrity.
The staff does 1-not concur with some of the factors used by the licensee in developing their criteria.
Clearly, it is not appropriate to remove all design margins P-
-becau'se-it ' leaves"no room for error or uncertainty.
Staff connents an'd our basis for rejecting the licensee's criteria for structural integrity are contained in Enclosure 2.
These recent reanalysis results dc not confirm the licensee's bases for-l continued operation contained in their April 28, 1980 letter, nor do they confirm the staff's conclusions conta ned Jr. our November 16, 1981 Interim l
Safety Evaluation Report on the Seismic Design of San Onofre Unit 1. 'It remains the staff's judgement that structures, systems and components which I
are adeguately designed for a 0.5g Housner SSE should have sufficient design l
l l
t 82 0 517 0CG9
f i
Gu's C. Liainas margin to assure structural integrity at a 0.679 Housner SSE. This. general-conclusion is supported by several recognized experts (Newmark,. Hall, Cornell And others) and work done by Research (USI A-40, SSMRP, etc.).
- However, it appears that this assumption of " adequately designed for a 0.5g Housner SSE" is not valid for San Onofre Unit 1.
The staff also concluded that the reanalysis would identify modifications necessary to restore design margin.
However, in order to assure prompt notification in the event of significant seismic design deficiencies, the staff required the licensee to notify the staff if at any time reanalysis results indicated any structures, systems or components which would not maintain structural integrity for a postulated SSE (0.679 Housner). The I
licensee's comitment to notify the staff is contained in letters dated December 2,1981 and December 3,1981.
During the staff caucus, discussions with our consultants (LLNL, EG&G Idaho, W. Hall) and on the basis of further review several serious problems have been identified.
Our conclusions are as follows:
- 1) The piping, support and equipment problems identified by the licensee's reanalysis are significant and require implementation of modifications before start-up to assure integrity of the reactor coolant pressure-boundary and the ability to remove reactor decay heat.
- 2) The analysis methods and criteria used for seismic reanalysis must be reviewed and approved by the staff prior to start-up.
For example serious deficiencies exist in the methods of. analysis and assumptions with respect to branch lines connected to the main reactor coolant system particularly in the interface between the Westinghouse arid Bechtel analysis scope.
f 3)
Issues related to loose soils and their resolution through additional analysis (structures. floor spectra, etc.) or corrective measure to compact soil need to be resolved before start-up.
Modifications assumed in the masonry wall analysis must be implemented 4) prior to start-up to assure the analysis represents the "as-built"
.i -
condition.
Modifications shown to be necessary as a result of re-analysis must,also.be implemented prior to start-up.
~
The licensee has not adequately addressed the effects of spalling of 5) block walls or the effect of large displaceme.'ts on attached or nearby safety-related equipment.
While the staff has not completed its review of the recent analysis methods or results and has identified several potentially significant deficiencies in methodology, it is my judgement that significant seismic design deficien-cies exist at San Onofre Unit 1.
I strongly recommend formal NRC action,
Gus C. Lainas "
consistent with other'NP.C actions taken when significant s'eismic design deficiencies are identified.
I recomend either an order or an order to show cause to require the plant to remain shutdown until modifications are implemented which assure the integrity of the reactor coolant pressure boundary and assure the ability to safely shutdown (remove decay heat) in the event of a large earthquake (SSE at 0.679 Housner). Also, since the licensee has not provided a basis for delaying modifications to the' East Feedwater Heater Platform and South Turbine Building Extension and since the total number of modifications identified to pate for restoration of design margin in piping and components is large (over 800), I rec'omend that the licensee be required to show cause why all modifications necessary to restore design margin should not be implemented by January 1,1983, or prior to start-up which ever is later.
I believe that continuing to jawbone this licensee or issuing another 50.54(f) letter (we have already issued three (3) on this subject) will not ' achieve the desired and necessary results.
(Del eted )
I i
In conclusion, I recognize the serious nature of these recommendations.
I have individually spoken to our consultants, my' staff and others.
I conclude that these actions are necessary and are supported by the personnel involved in the technical. review.
I am willing to meet with you to discuss these issues, bpt I felt obligated to put these thoughts and recomendations in writing.
i
&$T William T. Russell, Chief w
Systematic. Evaluation Program Branch Division of Licensing
Enclosures:
l-As stated l
cc w/ enclosures:
D. Eisenhut J. Knight D. Crutchfield W. Hall (U of Illinois)
W. Paulson T. Nelson (LLNL)
R. Hermann T. Bridges (EG&G)
T. Ch,eng K. Morton (EG&G)
P. Y. Chen T. Tsai (NCT)
K. Herring C. Grimes E. McKenna
Li ', -
SUMMARY
Do Not Meet
. Total Modificap]ons Total-Reevaluation Do Not Meet Required Evaluation Criteria SI Criteria Component Piping Large 170 115 0
s Small and Supports 130 64 Not evaluated by 362 licensee large Piping Supports 690 179 22 444 Mechanical Equipment Extended Operator Valves 69 40 2
Other /, 3/
j4 13 7
2 TOTAL 1,076 411 31 1/ Either modification to existing support or new support 2/ Probably more, other similar equipment not counted
)/ Mostly anchorage or support frame Enclosure (l')
4 t
t) 4,.
Do Not Meet Do Not Meet The Reevaluation The Integrity Modifications System Total Evaluated Criteria Cri.teria (I.C. )
Required Lines or Auxi1iary Coolant (Component Items Scope Cooling, Residual lleat Removal, large Piping Spent Fuel Pit Cooling)
Piping 79 100 63 0
251 100 59 10 33 new.
Supp rts 147. modified Small Piping Piping and
~ 18 of Supports 26 70 13 7
23 supports Chemical Feed Large Piping I
Piping Supports Small Piping Pipingand Supports 8 of 8.
100 6
?
2 supports ~
1 1
i 1
Enclosure (1)
yj 4.-
Do Not Meet Do Not Hent The Reevaluation The Integrity Modifications System Total Evaluated Criteria Crite.ria (I.C. )
Required Circulating Lines or Wat'er Items Scope Large Piping Piping 11 100 6
0 Suppbrts 28 100 8
1 Add one modify 7' Small Piping Piping and Supports 8
100 3
?
17 supports Compressed Air Large Piping E
Piping 14 of 45 30 5
0 Supports 110 supports Small Piping Pipi,t.J and Supports 10 of 45 20 7
0 Enclosure (1)
l Do Not Meet Do Not Meet The Reevaluation The Integrity Modifications System Total Evaluated Criteria Criteria (I.C.)
Required Chemical and Lines or Volume Control Items Scope y
Large Piping Piping Supports Small Piping 5
Piping and Supports 51 of 57 90 8
?
106 supports Feedwater and
+
Condensate Large Piping Piping 15 100 11 Supports 78 100 19 2
3 new supports 50 existing supports-Small Piping Piping and Supports 3
100 3
?
30 supports 4
e k
Enclosure (1)
Do Not Meet Do Not Meet The Reevaluation The Integrity Modifications System Total Evaluated Criteria Criteria (I.C.)
Required Miscellaneous Lines or Water Items Scope
'Large Piping Piping 12 100 10 0
Supports 92 100 22 3
10 new supports
- 48. existing supports Small Piping Piping and Supports 9
100 6
?
6 supports Safety Injection Large Piping Piping 8
100 0
0 Supp' orts 100 100 28 3
4 new supports 47 existing supports Small Piping Piping and Supports 9
100 6
?
8 supports d
S Enclosure (1)
Do Not Meet Do Not Meet The Reevaluation The Integrity Modifications System' Total Evaluated Criteria Criteria (I.C.)
Required Lines or Reactor Cycle Items Scope Sampling Large Piping
+
Piping None Supports None Small Piping Piping and Supports 5
80 5
?
9 supports Reactor Coolant Large Piping
~
Piping 13 100 12 0
20 new supports-Supports 48 100 19 2
24 existing supports Small Piping Piping and Supports 7
100 7
?
19 supports e
as Enclosure (1)
Do Not Meet Do Not Meet The Reevaluation The Integrity Modifications System Total Evaluated Criteria Criteria (I.C.)
Required Steam Lines or Items
. Scope Large Piping Piping 18 100 8
0 Supports 93 100 24 1
3 new supports mods. to 47 existing Small Piping Piping and Supports 2
50 0
?
32 supports modified and/or added e
9 e
9 Enclosure (1)
MECHANICAL EQUIPMENT Meets Reevaluation Meets Structural Type Criteria Intggrity Criteria Comments Field-erected tanks (CST, RWST)
No Ye Bolt and hoop stresses Heat Exchanger (Regen)
No No A'nchor bolts (Seal Water)
No No Concrete anchor bolts supporting steel frame (CCW)
No No Anchor bolts and saddle bolts (Recirculation)
No No Anchor bolts to pedestal (Spent fuel pit, RHR)
No Yes Saddle and anchor bolts Tanks (VCT, CCW surge)
No Yes Anchor bolts (air receiver)
No Yes Base plates and anchor bolts Pumps (SI, CCW)
No No Holddown bolts (SWC)
No Yes Bolts (RHR, charging)
No No~
Motor mounting bolts holddown bolts Filter No No Support pipes and/or bolts base plate weldment After Coolers Yes Yes Air Compressors No Yes Foundation anchors Note:
Limiting elements identified 9
Enclosure (1)
J, LICENSEE REPORT TABLE 3-6
~
INTEGRITY CRITERIA ALLOL' ABLES Components Stress Criteria Large (>2")
5.2 x Code Allowable ( )-
Piping Small (<2")
Not Applicable Component Standard
- 5. 2 x Component Design Support Load Rating 3.x S,(2)
Pipe Supports Structural Steel Support Concrete Fasteners 3 x Ultimate Pullout All components except
'2.36 x Code Allowable (-}
' " ' '****"'#8 Mechanical Equipment Concrete fasteners 3 x Design Load Rating field Erected Tank Su (I)From Tables 3-1, 3-2 and 3-3 Ultimate Strength Staff comments and basis for rejecting Structural Integrity Criteria prepared by Southern California Edison Company.
Piping-large diadeter
~
. The application of a factor of 5.2 over the ASME Code Service Condition DE level stress allowables implies the pipe is capable of resisting applied,
loads 2.5 to 3.1 times the Reference Temperature Ultima'te Tensile Strength
~
of the respective materials (A106, A 304S) or in the case of 304SS over 11.C times the yield strength at temperature. Also, no consideration has-been gi/en to f'aws that are present in the material and/or weldments or degradation by cyclic behavior.
If A5PA l evel D 5ervice Limits
," permit gross general deformations with some. consequent loss of dimensional stability and damage requiring repair, which may require removal of the component from seryice."
Enclosure (2)
Pipino-small diamet' er
". Operating experience and elastic-plastic fracture analysis have shown that small diameter piping is more susceptible to unstable tearing (rupture) than large diameter piping (i.e., steam generator tubes and charging pump lines).
Past agency policy for plants designed to early editions of the B31.1 Power Piping Code (allowable of 1.8S ) has been to require that piping stresses h
be limited to 2.4Sh for short term structural integrity (i.d., system operability as defined in facility Technical Specifications).
Structural Steel Supports AStiE limits are approximately.75u for level D service.
This'does not represent buckling limits, brittle failure limits, or cyclic limits for certain type supports.
Concrete Fasteners Current NRC criteria require that margins of 4 to 5 exist for wedge and shell type concrete fasteners, respectively.
Past NRC requirements for short term operability have utilized a margin of 2.
Recent tests show that a factor of safety of 3 in highly stressed structures are marginal.
The integrity allowable represents a factor of 15 times over current requirements for shell anchors.
General Applying Factors of Conservatism to design code allowables it not appropriate in any instance.
These factors if they could be technically justified would apply to the seismic portion of loading only and not other loads which are combined with seismic loads and then compared to allowables.
The Factors of Conservatism identified in the various LLNL reports (NUREG/CR-1161 and UCID-18104,17965 and 18110) are according to discussions with the authors appropriate for identification of areas of further investigation l./ -
or research into seismic. design margins and are not appropriate for use without additional bases in, licensing decisions.
At the large levels of inelastic behavior implied by-these limits, deformation rather than stress criteria must be established.
/
LICENSEE REPORT TABLE 3-5 Practice Margin a) Broadening of Fl'oor Response Spectra 1.17 (UCID-18104) b) Use of Elastic Spectra-with Inelastic l.2
Response
c) Consideration of actual material 1.17 (UCID-17965) properties d) Consideration of coupling effects 1.44 (UCID-18110) e) Reserve plastic capacity 2.2 to 2.6 f)
Increased Damping 1.0 to 1.-25 Staff coments on Seismic De:ign Margins with respect to San Onofre Nuclear Generating Station Unit 1 Conclusion The staff considers that a f actor of.l.2,' where applicable, has been justified over f aulted allow >.bles only wh'ere increases in actual material properties would be expected.
In other cases, no increase is justified, and as indicated in the referenced report, they have not demonstrated that certain factors are not less than 1.0.
a) The 1.17 factor for floor spectra is based upon the ratio of B/A from UCID-
)_.
18104 defined as:
~?
A"- t$ean spectrum of those from 16 artificial time history analyses of a
" generic" structure, broadened + 157..
B - Mean 4 le spectrum of those f rom 500 analyses, randomly varying stiffness, l
damping, and the choice of time hist'ori.!s from the 16 used for A. Spectra were not peak broadened.
l The f actor as an average is 1.0 and varied from <1 to >3 as a function of spectral-f requency,1.2 is the average at the fundamental frequencies of the structure.
This factor is not applicable in the case of SONGS 1 where the spectra are based on one time history.
In f act, UCID-18104 indicates that without further study, the present SONGS 1 spectra may be nonconservative and that a factor of <1 may be appropriate.
Enclosure (2)
c..
.3 b) 1.2 for inelastic structure response NUREG/CR-1161 indicates that for snell ductilities of the structure -
(<1.3) elastic spectra generally tend to envelope inelastic spectra in the regions of the spectra below the " rigid" range.
The rigid range response can be increased.
Greater inelastic response can cause response spectra to increase above those for the inelastic case.
SONGS' 1 levels of inelastic building response have not been quantified.
Also, the ductility reductions when applied to MDOF systems have not been generally justified. Therefore, further studies ar,e needed to determine if even 1.0 is adequate.
c) 1.17 for actual material properties At least they have one f actor which makes sense.
d) The f actor of 1.44 for coupling is derived f rom UCID-18104.
It is based upon the response of 1 piping system, in 1 building, subject to 5 different artificial E.Q. time histories. The mean factor of con-servatism for all of the 5 analyses at all points at 50% or greater of the maximum stress was 1.44 However, the study concludes the co-ef ficient of variation was 0.39.
This large variation reflects large difference in factors of conservatism among different locations. At a particular location, the maximum coefficient of variation was 0.08,.
indicating that the response did not vary greatly for different systhetic earthquakes.
The f actors of conservatism ranged f rom 0.9 to 2.8.
Therefore, a factor of 1.44 is not conservative at all high stress points on the piping system.
e) Reserve Plastic Capactiy factor of 2.2 to 2.6 appears to be based upon assuming inelastic piping response ductilities (as defined in NUREG/CR-0098) in the range of 3 to 4.
The ductility reduction factors are not generally applicable to highly redundantly supported multi-degree-of-
' f reedom systems subject to building motions, especially at such high
~
levels of assumed ductilities.
In all. cases where they are applied, a detailed knowledge and engineering evaluation of the load / deformation
. characteristics of the item being considered is required.
Factors greater than 1.0 cannot be supported without these evaluations and studies.
d'~
-m-The Factor of Conservatism for damping of 1.25 'is based on a reduction f) in going from 3 to 10% damping.
Damping as high.as 10% is not demonstra-ted to be reliable and NUREG-0098 and NUREG/CR-1161 both recommend 3% cs the maximun lindt f or vital piping.
Therefore, the consensus opinion expressed in NUREG/CR-1161 is 3% for an upper limit. This mean" a factor of conservatism of 1 is generally appropriate.
Inelastic energy absorption is more apprcpriately accounted for elsewhere, other than in the damping term.
For example, NUREG/CR-0098 methods.
e e
Enclosure (2)
. - _.