ML20211D576
| ML20211D576 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon, 05000000 |
| Issue date: | 02/15/1983 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20209C072 | List: |
| References | |
| FOIA-86-151 NUDOCS 8702240037 | |
| Download: ML20211D576 (13) | |
Text
r
,* f Y '
n w STAFF REPORT EVALUATION l; J '
e' i REPORT TITLE: INTERIM TECHNICAL REPORT ON OIABLO CANYON UNIT 1 i INDEPENDENT VERIFICATION PROGRAM r
l PACIFIC GAS & ELECTRIC
'\ '
PUMPS ITR 32 REVISION 0 AND REVISION 1 IDVP DESIGNATION: P-105-4-839-032 ORIGINATOR: Robert L. Cloud and Associates, Inc.
Teledyne Engineering Services
. ] BMITTED BY:
INTRODUCTION Interim Technical Report 32 (ITR-32 Rev. 0) was prepared by Robert L.
Cloud and Associates (RCLA) for the Diablo Canyon Nuclear Power Plant Independent Design Verification Program (DCNPP-IDVP). It summarizes the independent analysis and verification of a sample of pumps from the DCNFP Unit 1. The results indicate that stresses for each of the pump samples met the allowables defined by the licensing criteria. However because of generic concerns noted, RLCA recommended that the design analyses of the two remaining pumps analyzed by Pacific Gas and Electric (PG&E) and its contractors and all seismic inputs be reviewed. This acditional review was reported in ITR-32, Rev. 1, which addresses two of the three concerns.
SUPMARY OF REPORT Three pumps were selected as the initial sample for the independent analysis and verification of pumps in the DCNPP-1. The sample consisted of the auxiliary feedwater pump, an auxiliary saltwater pump 'and a comconent cooling water pump. The additional sample consisted of the fuel oil transfer pump and a motor-driven auxiliary feedwater pump.
The turbine driven auxiliary feedwater pump is a six stage centrifugal
_ pump driven by a steam turbine which is connected to the pump through a flexible coupling. The pump and turbine pedestal supports are welded to a common base plate assembly located at the 100' elevation of the auxiliary building. The base plate assembly is attached to the concrete floor slab with six anchor bolts and backfill of grout. The weight of the pumo section is 7,250 lbs and the weight of the turbine section is 2,960 lbs.
Following field verification of dimensions, natural frequency estimates for the ;:sp and turbir,e alone and the composite pump-turbine assembly were made using single degree of freedom models and a composite spring-two mass model, Spring characteristics were chosen to provide conservative (i.e., low) estimates of the system natural frequencies.
Next corresponding to the lowest natural frequency estimates and 4".
8702240037 PDR FOIA 870212 HOLNESO6-151 PDR
. I damping, seismic accelerations for the pump and turbine were chosen from the Hosgri response spectra. Lastly, using an equivalent static method, the loads, forces and stresses produced in the units by the combined action of the seismic accelerations, dead weight and normal thermal induced nozzle loads were determined.
The computed stresses for key locations were all found to be below the allowable stress levels. A comparison of the results with the Pacific Gas & Electric (PG&E) design analysis results showed the independent analysis estimates for stress to be higher than the design analysis values. This outcome was expected as the independent analysis used conservative estimates for natural frequencies which resulted in higher seismic loadings.
Two auxiliary saltwater pumps are located in the Unit 1 portion of the intake structure. Each is approximately 33' long and weighs 10,440 lbs.
Each pump consists of an electric motor, a discharge head, pump column and pump bowl. The motor and discharge head are mounted at floor level while, the pump column and bowl extend down and are submerged in the water. A vertical shaft connects the motor to the impeller at the pump bowl. The pump is supported at the floor by a mounting plate and restrainted at the top of the motor and at the pump by horizontal constraints.
The pump was field verified and then modeled for analysis with the finite element code STARDYNE. The model consisted of four concentrated mass points connected together and to the ground with rigid and beam elements. In the model the mounting flange was considered a pinned connection to provide low estimates of the vertical natural frequencies.
Also, the mass of water in the pump was included at the model mass points. Using the STARDYNE code and the response spectra modal analysis method, the pump natural frequencies and seismic loads were determined.
The seismic loads were combined with the pressure, deadweight and nozzle reaction loads and the stresses calculated. All computed stresses were found to be within the allowables. The closest approach to the allowable was for the mounting plate tensile stress which was approximately 50% of the allowable.
A comparison of the verification results with the design analysis results shcw that they differ by more than 15% with the verification results being consistently greater. Review of the design analysis indicated that the Rayleigh energy method was applied incorrectly and the mass of water was not included, resulting in an erroneous estimate for the fundamental horizontal frequency of the pump. These errors and different boundary conditions assumed in the analysis contributed to the differences in results.
The component cooling water pumps are electric motor driven, single stage, centrifugal pumps and are located at elevation 73' in the Unit 1
portion of the auxiliary building. The electric motor driver and pump are connected by a flexible coupiing. The pump and motor are mounted on separate pedestals which are welded to a common base plate. The base plate is a raised, steel member supported platform, attached to the concrete floor slab with six cast in place anchor bolts and a grout backfill. The pump weighs 5,900 lbs., while, the motor weighs 4,300 lbs.
Following field verification separate finite element models were developed for the pump and motor. The pump was modeled as a concentrated mass supported by rigid and beam elements representing the pumo pedestal. The motor was modeled as a concentrated mass supported by quadrilateral and triangular plate elements. These models were used to estimate the natural frequencies of the components. For the pump, upper and lower bounding cases were run simulating the extremes of the adjustable pump feet to pedestal mountings. The natural frequencies of the both the pump and motor were found to exceed 33 Hz. The seismic accelerations for both components are therefore, equal and correspond to the free field, zero period accelerations.
Seismic accelerations, deadweight, torque and pump nozzle loads were used in equivalent static analysis to determine the loads and stresses at key locations. For the pump, an additional calculation was performed to determine the pump shaft bending and shear stress and deflections.
Lastly, the ccmbined loadings from tne pump and motor analyses were used to find the loads and stresses in the base plate. All stresses were founo to be below the allowable levels although the stresses in the base bolts were close (40,482 vs. 40,600 psi).
A ccmparison to the PG&E design analysis showed the results to differ by more than 15%. The differences were attributed to variations in the methods, weights and dimensions used in the analyses. It was noted, that incorrect values were used for the thickness of the pump pedestal and the weight of the pump impeller in the design analysis.
Six E0I's were issued as a result of the comparisor of verification and design analysis methods and results (E01's 1020, 1022, 1072, 1073, 1113 and 1114). These E01's noted the various errors or omissions uncovered in the design analyses as well as stress differences greater than 15%.
E01's 1020, 1072, 1073, 1113 were subsequentially closed when the spectra used were substantiated and the causes for result differences determined. E01 1022 was classed a type A/B error, as incorrect spectra were used-in the design analysis of the upper pump support. E0I 1114 was classed an error C and noted the omission of hydrodynamic mass as part of the mass of the submerged portion of the auxiliary saltwater pumo.
Three concerns were expressed by RCLA: a) use of correct Hosgri spectra, b) correct calculation of fundamental natural frequencies, and i
i l
i
_4_
c) inclusion of hydrodynamic mass where appropriate. They, therefore, recommended that an additional pump sample be verified. This sample consisted of the two remaining pumps, the fuel oil transfer pump and the motor driven auxiliary feedwater pumps.
RCLA did not perform full verification for these pumps as for the initial sample. The concern regarding the correct spectra will be resolved when the IDVP reviews the application of spectra in the DCP corrective action program, and will be reported in a subsequent interim technical report describing control and application of Hosgri spectra.
The two other concerns were resolved by RCLA by verifying that the fundamental frequencies for these pumps were calculated correctly, and that the hydrodynamic mass concern was inapplicable to these pumps since they are not immersed in fluid.
E'/ALUATION The procedures followed by RLCA in performing this independent analysis and verification are acceptable and did reveal deficiencies in the PG&E design analyses. For the most part, the models used by RLCA were designed to provide conservative results and were considered as sophisticated as the models used in the design analyses. All reported stresses were found to be below the allowables and therefore the pumps evaluated met the licensing criteria. However, for the auxiliary salt water pump, no results were presented for either the pump impeller shaft stresses and deflections, or impeller shaft bearing loads, if any.
Given the long span of this shaft, from motor to pump bowl, some estimate of these parameters should have been developed. In addition, no discharge head nozzle stresses were reported, nor compared to an allowable stress. These items wil.1 be addressed and reported by RCLA in the verification by the IDVP of the PG&E Corrective Action Program.
m . , _ . _ _
APPENDIX A
}
z I
i
FINAL RESOLUTION SHEET File No. 1020 1
Revision No.
(. 1. Classification Class (A, B, C or D)
Error
[I~ Deviation Closed Item l ,
e; .. . *
'2 .
Documentation Reviewed PGandE Auxiliary Saltwater Fu=p analysis File 116.31', Revision 0, 12/27/77. ,
F105-4-550-006 RLCA co=parison of spectra.
r Reported to ?GandE Transmittal Date 3 /17 /82~
c,, 3.
4 Description T.he Hosgri spectra attached to the PGandE analysis for the Auxiliary '
Saltwater Pump are preliminary spectra.
- 5. Final Resolution .
The preliminary spectra are identical to the H:sgri spectra. The accelerations used in the calculations are correct, therefore this ite= is a deviation.
t
/~ . M_ -m 3,!l7 bl Proj ect En.gineer /Dat e To Indiccte PLCA Final Resc'_ution
FRROR REPORT File No. 1022
. Class: A/B File Revision No. 5 A,B,C or 0 70093 PG&E Task No.
~
1.
Dates: Reported to Program Review Comnittee N/A Program Review Committee Action N/A Reported to PG&E and Originator 820910
- 2. Scheduled for TES Semimonthly Report No. Septemoer , ,
- 3. Structure (s), system (s), or component (s) involved:
Intake Structure.
4 Description of Error:
E01 1022 Rev. 4: The upper Auxiliary Salt Water Pump support -
is located ten feet above elevation *-2.1 ft. Spectra applicable at -2.1 feet were used as input for the upper pump support.
5
- 5. Significance of Error:
Scectra used in the design analysis may not be applicable to Auxiliary Salt Water Pump upper support.
- 6. Recommendation:
Based on the PGandE presentations on August 6 and September 1,1982, .,
the Intake Structure is in reanalysis in the PGandE Internal Technical
! Program. 'FiIes 967 and 9883 which also pertain to the Intake Structure, have been combined into this file,1022, which is issued as an Errer Class A or B.
- 7. Potential Error Report signed by Edward Denison (RLCA) on 820907 fype Name/Gro'nijz tien p2*n
- 8. Signatures: N/A pf g_ e20MO
- for Program Revie.. Committee Approvec//rogram Manager l
PROGRAM RESOLUTION REPORT File No. 1072 File Revision No. 2
- 1. Resolution of an: II Open Item: O Class Error
- 2. Incependent Design Verification Program Resolution is as:
- a. O Closed Item
- b. 23 Deviation
- c. O Open Item with future action by PG&E: Task
- 3. Date Reported to PG&E 4 Scheduled for TES Semimentnly Report No.
- 5. Resolution based on the folicwing documentation:
E0I 1072, Rev. 1 .
RLCA P105-4-550-010, Rev. 1 The spectra used in the qualification summary are identical to the Hosgri spectra except for the identification numbers.
A
- 5. 3r: gram Resoluticn is:
Since the spectra were not contained in a controlled document, this item is a deviation.
All other itens relative to tnis E0I may be closed.
- 7. ?0tentil' Pr0 gram Resolutien Recert signed oy Edward Denison (RLCA) on 920609
/ J-sy;e . Oa:e
- . Signature: 7.2 g g_ 'iame /Gegft.an o 7i oz sat i cn (Ac; roved / Program Manager)
J
ERROR REPORT File No. 1073 Class: C File Revision No. 2 A,B, r0 PG&E Task No.
- 1. Cates: Reported to Program Review Comittee N/A Program Review Committee Action wa Reported to PG&E and Originator ___ ._
- 2. Scheduled for TES Semimonthly Report No. 16 ,
- 3. Structure (s), system (s), or component (s) involved:
Aux. Salt Water Pump i 4. Description of Error:
The design analysis contains an error in the application of the Rayleigh-Ritz method. Approximate third mode displacements were assumed in PG&E design analysis whereas there are lower bending modes present in the system.
The bolt overstress cited in File Revision 0 does not exist for the correct shut-off pressure of 15.4K. Original value of 24.7K used by RLCA was determined to be overly conservative. .
- 5. Significance of Error:
All stresses are belew the allowable
- 5. Recc.mendation:
IDVP Completion Report can be issued if PG&E informs TES that there will be no physical modifications.
/. ,otential r Error Report signed by E. Denison/RLCA 820608 on anization Date S. Signatures: W Type Name/Org & & _ Szon e Fce Program Review Committee Approve:/ Program Manager
UFEN ITEM REPORT File no.
1113 File Revisien t:o. O 1.- h te reported to PG&E and TES February 1 1983
~
- c. Icheduled for .R(CA (OriginatorJ Semimonthly Reper: No. 'Feb rua ry 7 F.esponsive to FG&E Tecnnical Program: Task (if amplicacie)
- c. Frepared as a result of:
- a. O-CA Audit and Review. Report of
- c. O Field Inspection Deficiency
- c. .C Independent Calculation Deficiency
- c. O Seismic Input Deficiency
- e. O Design Methodology Deficiency-
- f. O 0:ner Deficiency -
- 5. Structure (s), system (s) or com;renent(s) involve :
Component Cooling Water Pump .
- 5. Description of Concern: "
Results of design and' independent verification analysis
. .= . differ by more than 15%. '
. .i s~
r-G- 7.
Significanca of Ccncern:
All stresses,3elow allowable. _
~ '
- 5. Rec er.rencaticn : .
Closed item. -- - -
< gi f.
0/ 8, -
[
WWNs U
_ _ _ _ _ _ _ _ .GW. .J. _ _ _ .L.I I.s.
/. .T.*. _]_ _ ~~'***'*I,- . t : : f ,' , '
File fio. i11a OPEN ITEM REPORT File Revision No. o
- 1. Date reported to PG&E and TES 2/15/83
- 2. Scheduled for RLCA (Originator) Semimonthly Report No. March
- 3. Responsive to PG&E Tecnnical Program: Task (if applicable)
- 4. Prepared as a result of:
- a. O QA Audit and Review Report of
- b. O Field Inspection Deficiency
- c. O Independent Calculation Deficiency
- d. O Seismic Input Deficiency
- e. 9 Design Methodology Deficiency-
- f. O Other Deficiency
- 5. Structure (s), system (s) or component (s) involved:
Auxiliary Saltwater Pump
- 6. Cescription of Concern:
The design analysis for the auxiliary salt'sater pump does not consider the virtual mass contribution of the ,
water surrounding on the sub=erged pump casing.
- 7. Significance of Concern:
The verificat, ion analysis considered the virtual water-
= ass contribution and found all stresses to be below the allowable.
- 3. Recc:mendaticn: .
Error Class C.
e
- 9. S igna ture : 2 fl5[73 ( or ig ina :o r /O rg an i z a: i cn >
- e. ..
- ~
1 , .
- h. e. 3 p j,f _
- L BROOKHAVEN NATIONAL LABORATORY
- - "(w..:
T a.
f [l -
' ASSOCIATED UNIVERSITIES, INC.
Department of Nuclear Energy Upton, Long Island, New York 11973 Building 129 (516)282s 244g FTS 666' Februa ry 16, 1983 , .
/,
7 7
& cLoc kl U [4 Dr. Mark Hartzman Room No. P-520-A Phillips Bldg.
US Nuclear Regulatory Commission 7920 Norfolk Avenue Bethesda, MD 20014
Dear Dr. Hartzman:
The modeling of PG&E Problem 8-118 and Westinghouse Problem RHR Loop 4 (6-4, 6-7) are progressing.
The submitted document package for PG&E Problem 8-118 is essentially complete. Moreover, the system dimensions and data as specified in.the PG&E computer input listing for this problem are generally consistent with the drawings and data in the document package. When the BNL modeling is complete a listing of the observed differences and data deficiencies will be provided.
Thus so far these differences have been small (i.e., dimensional dif ferences of a few inches have been noted).
The submitted document package for Westinghouse Problem RHR Loop 4 (6-4,6-7) is not sufficient to allow complete verification. In modeling this problem Westinghouse used sections of the RC Loop 4 hot and cold leg piping as well as models of the Reactor Vessel, RC Pump and Steam Generator. The submitted document package only provided data and drawings for the RHR Loop 4 pi pi ng. To proceed with the modeling we are referring to and accepting the data provided in the submitted computer input listing for the definition of the piping and components mentioned above. However, to allow complete verification we require an expansion of the document package to include all piping and components considered in the model of the RHR Loop 4 (6-4, 6-7).
For the major components (i.e, Reactor Vessel, etc.) the appropriate finite element or stick models would be adequate. A partial list of these needs are drawings and data describing:
[/,y/
~ ,
q
Dr. Mark Hartzman Februa ry 16, 1983
- ~
RC Loop 4 hot and cold leg piping Modeling details of Reactor Vessel RC Pump .
Supports and restraints on hot and cold leg of RC Loop 4.
It must be emphasized that required data should be assembled and augmented by someone knowledgable with the problem under review as our requests are based only on the limited data provided for RHR Loop 4 and is not necessarily inclusive.
Additional needs for this problem include:
Full sized Piping and Instrumentation drawings No. 102009 sheets 3 of 9 and 5 of 9 No. 102010 Weight specifications for regular and special insulation Plan view drawing showing containnent building and piping at elevation 107'.
Lastly, one large inconsistency'between the supplied RHR Loop 4 drawings and the computer input listing has been noted and should be resolved.
Specifically, the SD bend w;th a 6' angle shown on isometric 437985 was modeled by Westinghouse as a long radius elbow at a 16 angle. At this time we are using the Westinghouse definition in the BNL model.
Please forward these data requests to PG&E - Westinghouse. Until we receive new information we will continue to use the data in the computer input listing for the Westinghouse Problem. A prompt reply to the above questions, would really speed up the contenplated completion date for these problems, i
l Sin ere' yours, Morr i Reich, Head St uctura Analysis Division jm k i pc: P.T. Kuo H. Polk H. Schierling ;j/
- . _ - _ _ _