ML16340C368
| ML16340C368 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 01/15/1982 |
| From: | Fair J, Herring K, Morrill P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV), NRC OFFICE OF INSPECTION & ENFORCEMENT (IE), Office of Nuclear Reactor Regulation |
| To: | Engelken R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| Shared Package | |
| ML16340C365 | List: |
| References | |
| NUDOCS 8203090089 | |
| Download: ML16340C368 (22) | |
Text
~@A AIQy~fp0
<a*++
UNITED STATES NUCLEAR REGULATORY COMMISSION REGION V 1460 MARIALANE. SUITE 260 WALNUT CREEK, CALII'ORNIASi696 January 15, 1982 MEMORANDUM FOR:
R.
H. Engelken, Regional Administrator THRU'ROM:
J.
L. Crews, Chief, Division of Resident, Reactor
- Projects, and fngineering Inspection P.
J
~ Morrill, Reactor Inspector, Region V
K. S. Herring, Systematic Evaluation Program Branch, Division of Licensing, NRR J.
R. Fair, Reactor Engineering
- Branch, IE
SUBJECT:
PRELIMINARY ASSESSMENT OF R. L.
CLOUD AND PG&E REVERIFICATION PROGRAM Based on recent inspection efforts of the R. L. Cloud/PG&E Reverification program, it appears, in our opinion, that there are serious concerns requiring prompt NRC Management/Staff evaluation and direction.
Without staff evaluation of the proposal submitted by PG&E on December 4, 1981, and direction and clarification to PG&E/Cloud as to what is acceptable or not acceptable, Cloud and PG&E will continue on their present "Program" with its various potential problems; while the inspection staff will be delayed in its effort by a lack of sufficient definition as to what constitutes a program acceptable to NRC.
Messrs.
K. Herring and J. Fair of Headquarters have worked during the period of January 13-15, 1982, to develop the following potential problem with ongoing work by R. L. Cloud Associates.
l.
Issues Identified b Cloud's Preliminar Seismic Reverification Pro ram are not Necessaril Bein Pursued b
Cloud's Current Work. e.
- Westinghouse adequacy to correctly use Blume's Seismic
Response
(incorrect use of Tau Filtered Vertical Spectra in at least two instances).
-'dequacy of Control Room equipment qualification based only on a
review of document transmittal
- dates, not the documents themselves.
2.
The Sco e of the Work Bein Done in Accordance with the December 4,
1981 Pro osed Seismic, Reverification Pro ram A
ears to be Less than that Re uired b
the NRC's Order of November 19 1981 e.
There is no provision in Cloud's work to verify PG&E's structural element evaluations given input from Blume.
Cloud has excluded Westinghouse and General Electric from further examination
"...because their work was performed in support of the sale and licensing of the NSSS and associated products.
8203090089 820208 PDR ADOCK 05000275 P
'E H. Engelken January 15, 1982 3.
The Criteria for Identification Si nificance Determination, Re ortin "Deviations" and Follow-u of Errors and Omissions Discovered ave not been Established or defined e.
The auxiliary building seismic responses documented in'he Hosgri report were not up-dated upon the receipt of the final Blume report to PG&E in October 1979.
The NRC was not informed when PG&E became aware of this deficiency in November 1981.
The intake structure seismic responses documented in the Hosgri report appear to have,similar problems which have not been reported previously to the NRC although PG&E and Cloud personnel are evaluating the problem.
- Further deficiencies in this area are discussed in Item 5, below.
4.
The Level of Reverification and Additional Sam lin A
ear Inade uate.
e.
- The independent reverifi'cation of the auxiliary building structure is being done by one individual with. hand calculations of the building masses and stiffnesses.
No independent computer run or examination of Blume's work is expected unless the masses or stiffness developed by PG&E are found to be greater than 10 to 15 percent in error. per the December 4, 1981 PG&E submittal.
Review indicates a problem in this area in the Blume QA/QC program.
- The seismic model for a fan cooler was found by Cloud to be in error.
The error turned out to be conservative (in this case),
consequently, no additional sampling appears to be scheduled for such cases.
5.
The Acce tance Criteria and Metkodolo of the Reverification Pro ram are not Ade uatel Defined.
e.
The methodology used for evaluation of structures (auxiliary building),
piping and equipment (in most cases) is the same as that reportedly used by PG&E or its contractors.
State of the art improvements are generally not being used.
On January 4, 1982, Cloud personnel documented "Criteria and Methodology for Independent Calculations" and "Criteria for Indepen-dent Evaluation" by an internal memo.
This material has not been incorporated into the December 4,
1981 program plan.
Both sets of criteria do not define in much detail the criteria for reanalysis, the acceptance and rejection criteria, and the criteria to be us'ed for considering sample expansion.
Discussions with Cloud's employees indicated that confusion and a lack of defined goals exist in these areas.
R.
H. Engelken January 15, 1982 These problems point out the need to promptly:
Accept or Reject Cloud Determine the Adequacy of the Reverification Program Develop Necessary Guidance (Scope, Criteria, Reporting, Methodology, etc. )
A convenient form to resolve these concerns is a meeting with PGEE and their contractors such as that previously scheduled for January 19,
- 1981, and subsequently postponed.
We propose that a meeting of appropriate NRC Staff be held as soon as possible to discuss further the concerns described
- above, to be followed up by a meeting with the licensee as previously planned.
Reactor Inspector, Region V
K. S. Herring Systematic Evaluation Program Branch Division of Licensing, NRR J.
R. Fair Reactbr Engineering
- Branch, IE
4' k
AUXILIARY BUILDING WEIGHT RECONCI LI ATION Commenta A computer program DYBOX was run by PG&E in December 1970 to calculate the building properties.
'The results of this program were given to URS/Blume for use in the development of the building dynamic model.
Another computer program SHERWAL was developed by PG&E and run in March 1977 to calculate the stresses and reinforcement of the building shear-walls.
A copy of the output of this run was given to URS/Blume.
Since SHERWAL also computes the building properties, URS/Blume compared the building weights calculated by SHERWAL versus the ones they had used in the dynamic model and reported a discrepancy to PG&E. It was found that the SHERWAL run performed in March 1977 was improperly distributing wall weights.
The error was corrected and SHERWAL was run again in July 1977.
This matter was discussed with URS/Blume at that time but the out-put of the July 1977 run. was not transmitted in 1977 to URS/Blume.
There-fore the. discrepancy reported remains an open item in their files'opies of the July 1977 SHERWAL output were recently transmitted to LES/Blume.
A new weight reconciliation has been performed and it was found that when the weights computed by SHERWAL are adjusted as shown on Table 1, they compare with those used in the URS/Blume model within 5%, except at Elevation 115 where it compares within 9
F 7~
OSCAR A, ROCHA
~
~
4'
TABLE 3 AUXI ARY BUILDING WEIGHT RECONCIL~ION Mass Points Eleva tion URS/Blume Model kips SHERWAL July 1977 ki s I
Remarks 288 2, 637.
FHB roof weight 363 11, 595 23, 612
/ Difference
= 0 140 58,079 62,897*
- Includes FHB roof weight in the total at this level
(.-)2,637
.58,079 60,260
/ Difference
= 4/
115 64, 292
'.64,292 55,733
(+)1,500
(+)
337
(+)
515 58,085 Weight of main steam and feed-water pipes and restraints not considered by the SHERWAL pro-gram.
4 Boric acid tanks 1/2 of concrete vaults outside building.
% Difference
= 9..7/
100 58,892 36,145
(+) 6, 630
+ 9,212 Weight of spent fuel pool water not considered by.SHERWAL Weight of slab resting on rock at El.
100 not considered by SHERWAL
+ 1,862 Live load on slab on grade not considered by SHERWAL
+
515 1/2 weight of 7 concrete vault5 outside building 58,892
+ 1,735 55,829 1/2 weight of liquid hold-up tanks
/ Difference=5%
4 1
UNITED STATES NUCLEAR R EG ULATORY COMMISSION WASHINGTON, D. C. 20555 PEB 8 1982 MEMORANDUM FOR:
Chairman Palladino Commissioner Gilinsky Commissioner Bradford Commissioner Ahearne Commissioner Roberts FROM:
SUBJECT:
Darrell G. Eisenhut, Director Division of Licensing Office of Nuclear Reactor Regulation INFORMATION ITEM - POTENTIAL DEFICIENCY IN THE SEISMIC ANALYSIS OF EQUIPMENT AND COMPONENTS IN THE CONTAINMENT ANNULUS OF DIABLO CANYON UNIT 1
(BOARD NOTIFICATION No. 82-11)
In accordance with present NRC procedures regarding Board Notifications the enclosed information is being provided to the Commission.
Enclosure:
Trip Report dtd. 2/3/82 cc:
ASLB ASLBP SECY OGC OPE EDO Service List
Contact:
B.
C. Buckley, X49-28379 Darrell I. Eisenhut, D rector Division of Licensing Office of Nuclear Reactor Regulation
UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, O. C. 20555 FiQ ljjp MEMORANDUM FOR:
Frank Miraglia, Chief Licensing Branch 3, Division of Licensing, NRR FROM:
SUBJECT:
Introduction K. S. Herring, Systematic Evaluation Program Branch, Division of Licensing, NRR TRIP-REPORT REVIEW OF URS/BLUME HOSGRI ANALYSES OF THE DIABLO CANYON AUXILIARYBUILDING, INTAKE STRUCTURE, CONTAINMENT POLAR CRANE, AND CONTAINMENT ANNULUS Enclosure 1 identified concerns regarding PG&E's apparent failure to properly reflect the results and findings. of final reports it received from URS/Blume in the docketed FSAR (Hosgri Report) and its piping, equipment, and component evaluations for the Diablo Canyon Auxiliary Building and Intake Structure.
Enclosure 1 also identified a concern over the depth of the PG&E proposed Seismic Reverification Program in the structural area with regard to the analyses performed by URS/Blume for PG&E.
To gain further insights into the analyses performed by URS/Blume and the adequacy of the related information contained in the docketed Hosgri Report, on January 25 and 26,
- 1982, I conducted a review of the URS/Blume calculation books for the Hosgri analyses of the Diablo Canyon Auxiliary Building, Intake Structure, Containment Polar Crane, and Containment Annulus.
The results of this review are summarized below.
Review Results Auxiliary Building:
1977 Weight Discrepancy - The Cloud verification effort identified that a significant weight discrepancy of 35K between original Blume and later PG&E analyses existed at elevation 140'.
The Blume calculation book indicated that discrepancies of about
+16%, +9.6X, and -34~ also existed at elevations 163',
115',
and 100', respectively.
The calculation book indicated that the original Blume weights were used in the analyses but gave no basis for this.
It appears that the use of these initial Blume weights in the Blume structural analyses was appropriate and not adequately documented in the initial calculations.
I
~
\\
I,
~,," nI j,
~
~
URS/Blume personnel stated that they had recently received a
written explanation from PGSE describing that the weight discrepancy was due to an erroneous March 1977 run of the SHERWAL computer program and the lack of consideration of certain weights by the SHERWAL program.
When a correct SHERWAL run was
- made, and the appropriate neglected weights were added to the SHERWAL weights, the weights compare well with those used by Blume in its Auxiliary Building analyses (see ).
This was confirmed by PG&E personnel who further indicated that SHERWAL computed weights are use'd in conjunction with acceleration profiles from Blume's structural analyses to perform structural evaluations of walls. Although there are differences between the masses used in the SHERWAL analysis, and those actually present, PGKE personnel contended that the SHERWAL analysis results would not be significantly affected')
Elevation 100'oil Spring Omission - Differences in spectra between the Hosgri Report N-S floor spectra for the Auxiliary Building and the final Blume report on the Auxiliary Building were identified by Cloud, and afterward found by the NRC to be attributed to an error in the incorporation of the soil spring at elevation 100'f the building model in the preliminary Blume analyses on which the FSAR (Hosgri Report) is based.
This was corrected and included in the analysis on which the final report was based.
An October, 1978 transmittal from Blume to PGLE regarding a
Blume design review reported this soil spring error but concluded that when the spring is included in the analysis,
~..the result leads to higher estimates of responses and is thus conservative."
Recent information indicates that this statement is not true with regard to certain of the N-S floor response spectra for the Auxiliary Building.
The Blume calculation book contains results of an evaluation conducted between 1/78 and 5/78 to determine the effects of neglecting the soil spring since this was erroneously omitted in earlier analyses.
The study concluded that the responses in the N-S direction were indeed affected by the presence of the soil spring.
Comparisons of floor spectra with and without inclusion of the soil spring were presented which indicated certain areas of the floor spectra were greater when the soil spring was included, especially for torsional
- response, and certain areas were lower.
The calculation book indicated with no stated basis that this analysis was not used.
The person at Blume who originated the 10/78 transmittal stated that
( 1) he was not aware of the 1/78-5/78 study and
0
'l k
~ ~ (2) that although his 10/78 conclusions addressed "responses",
he was referring only to building forces and peak structural responses not floor response spectra.
It appears that this error was caused by (1) inadequate personnel interfaces within
- Blume, and (2) lack of sufficient consideration, by Blume personnel, of PG&E use of floor spectra in their piping and equipment evaluations.
Intake Structure:
Use of Inappropriate Spectra for PG&E Evaluations - All preliminary and the final Blume reports regarding the Intake Structure indicated that the response spectra for the design of equipment at the roof level were similar to the ground spectra for most areas of the roof.
No spectra were supplied in these reports and PG&E has used the ground spectra for its evaluations of all areas of the Intake Structure, including the roof.
Spectra at several points at the roof were contained in the Blume calculation book for this structure.
One of these points was at the roof area above the Auxiliary Saltwater Pumps.
These spectra indicated significant spectral peaks in the 20-25 Hz range which are not present in the ground spectra.
Blume personnel indicated that PG&E had only recently requested these spectra and that Blume was now in the process of peak broadening them for transmittal to PG&E.
The PG&E review of the preliminary and final Blume reports on the Intake Structure was not sufficient to detect this issue earlier.
Containment Polar Crane:
1)
Polar Crane Analyses - The docketed FSAR (Hosgri Report) concerning the structural analyses and integrity of the Containment Polar Crane contains the results of the URS/Blume 2-D nonlinear and 3-D linear elastic analyses of the polar crane.
Results of these analyses (as relied upon by the HRC in Supplement 9
to its SER) indicated that the interaction ratios for stresses in the polar crane members are all less than 1.0; therefore, no overstresses are predicted.
A later 3-D nonlinear analysis of this crane was performed by URS/Blume and the results provided to PG&E by Blume in a report dated July, 1979.
This latter analysis indicated the potential for interaction ratios as high as 1.3 in the crane support columns, which is an overstress.
The Blume report concluded that this was acceptable since it was localized and there was only one peak load excursion.
~ ~
t
(
Blume calculations considered the time phasing of loads and used actual average material properties to evaluate these members in the absence of the normal AISC Code margins.
- Also, load time histories for these members contained in the Blume calculation book indicated that while there was only one peak load excursion, the potential existed for several in the range of 80-90K of the peak.
This analysis was disregarded and not docketed by PG&E regarding the structural integrity analyses.
PG&E personnel indicated that when they received the July 1979 Blume report, a comprehensive review of the report was not
,conducted since the Blume conclusion that no modifications to the crane were required was not changed from previous reports.
The PG&E review of the July 1979 Blume report was not sufficient to conclude that this report demonstrated that the analyses results presented in the FSAR were significantly less conservative, than those contained therein and,.therefore, may warrant further evaluation.
2)
Dome Service Crane - The dome service crane analyses being performed by PG&E incorporate undocketed 3-D nonlinear Polar Crane analyses results which have not been reviewed by the NRC (see above discussion of the Polar Crane Analyses).
Containment Annulus Structure:
2)
Reanalyses Models of the Annulus - The reanalyses of the Annulus that has been performed by Blume incorporate the original analysis model, with mass and stiffnesses. revised to reflect the "as-built" configuration.
Annulus Spectra for Use in PG&E Piping and Equipment Evaluations-Blume personnel indicated to PG&E that piping and equipment should be evaluated using spectra corresponding to the "frame" on which it is supported, even when they are located near the center line between connected frames with different responses.
This interconnection is not modeled in the Blume analyses.
Therefore, this approach is based on consistent application of modeling assumptions.
The adequacy of this model is under detailed f/RC review, using an independent WRC contractor's analysis of the Annulus structure.
t I
1
~
t~
Conclusion Based upon the results of my review, as summarized above and in Enclosure 1,
it is concluded that the information contained in the FSAR (Hosgri Report),
on which the HRC Safety Evaluation Reports and its supplements are based, is not accurate concerning the Auxiliary Building, Intake Structure, and Containment Polar Crane.
(The information for the Auxiliary Building and Intake Structure is based upon preliminary Blume reports and certain information was changed in the final Blume reports.
The information concerning the Containment Polar Crane was superceded by later analysis results.)
It appears that this has been caused primarily by a lack. of thoroughness in technical review by PGLE of final URS/Blume reports and the associated analyses.
Another contributor to the problem in the case of the Auxiliary Building appears to be that 1) inadequate personnel interfaces existed within Blume, and 2) Blume personnel focused considerations primarily on structure forces and peak responses, without sufficient consideration of PGEE use of floor spectra in their piping and equipment evaluations.
In addition, a secondary contributer to the overall problems is Blume's failure to clearly indicate substantative changes between preliminary and final reports.
Considering the implications of this review of the URS/Blume analyses, it should be determined to what extent similar problems may exist throughout the analyses and evaluations presented in the FSAR (Hosgri Report).
The implications of this problem can then be assessed.
See next page.
Kenneth S. Herring Systematic Evaluation Program Branch Division of Licensing
0'