ML20054J659
| ML20054J659 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 06/24/1982 |
| From: | Withers B PORTLAND GENERAL ELECTRIC CO. |
| To: | Clark R Office of Nuclear Reactor Regulation |
| References | |
| TAC-12369, NUDOCS 8206290393 | |
| Download: ML20054J659 (9) | |
Text
_. _ _ _ _ _ _ _. _ _ _ ___
1 l
numsms r
-- Ell PortlandGeneralElectricCorigxuly
.m
. o.. c.,
e p. w i
June 24, 1982 Trojan Nuclear Plant Docket 50-344 License NPF-1 l
Director of Nuclear Reactor Regulation ATTN:
Mr. Robert A. Clark, Chief Operating Reactors Branch No. 3 Division of Licensing U. S. Nuclear Regulatory Commission Washington, DC 20555
Dear Mr. Clark:
Evaluation Criteria for Masonry Walls The attachment to this letter completes our response to the request for information contained in your letter of May 8,1981. Responses to items 6 through 10 are attached. Responses to questions 1 through 5 were previously submitted by letter dated June 19, 1981.
Sincerely, Bart D. Withers Vice President Nuclear Attachments c:
Mr. R. H. Engelken, Director U. S. Nuclear Regulatory Commission Region V g/
Mr. Lynn Frank, Director State of Oregon Department of Energy 8206290393 820624 PDR ADOCK 05000344 P
PDR 121 S.W Sa' mon Street, Portlarx1 Oregon 97204
Trojan Nuclsar Plant Robsrt A. Clark Docket 50-344 June 24, 1982 License NPF-1 Attachment Page 1 of 8 Question 6:
The licensee has not addressed the criteria for determining the over-all structural response from the missile impact. Either these criteria should be provided, or justification for not considering the response should be provided.
In addition, during the recent phone conversation with the licensee it was indicated that the licensee may not have con-sidered the combined effects of tornado wind load and pressure drop loads as indicated in Attachment 1 to the October 23, 1980 letter. Our position is that all of the above three effects are to be combined (this is in conformance with the FSAR; earlier SERs relied on the compliance with FSAR provisions). The licensee should comply with the above cri-teria or provide justification for not doing so.
Answer:
As described in Attachment 2 to our October 23, 1980 letter, tornado load design criteria as now included in the NRC Standard Review Plan (SRP) and Regulatory Guide 1.76 had not yet been established during the period in which the Trojan Plant structures were designed.
- However, pursuant to AEC and ACRS interests in tornado load effects during the licensing phase for Trojan and notwithstanding the extremely low prob-ability of occurrence of a major tornado in the vicinity of the Plant, safety-related structures were evaluated to determine their threshold capacity to withstand tornado loads. As a result of these evaluations, at appropriate locations new shield walls were added and existing walls were strengthened during construction. Following such modifications, Plant structures containing systems needed to achieve and maintain safe shutdown were determined to be capable of resisting 200-mph tornado loads, and many of these structures were determined to be further capa-ble of resisting 300-mph tornado loads (Ref FSAR Section 3.3.2).
The evaluations of tornado effects were performed in accordance with Bechtel analytical guidelines available at that time which included consideration of tornado wind loads, differential pressures, and pene-tration and structural stability under missile impact. The individual effects were not, however, combined in the manner that is currently recommended in the SRP. Exterior wall and roof slab barriers intended to provide resistance to tornado loads were evaluated for their capa-bility to withstand tornado loads in the following manner:
l 1.
Velocity Pressure - Loads caused by velocity pressures are less than those caused by atmospheric differential pressures.
2.
Atmospheric Differential Pressure - Loads caused by l
atmospheric differential pressures were considered to l
act alone.
If velocity pressure loads are combined with differential pressure loads per the SRP, the resulting combined loads are less than the differential pressure loads acting alone.
4 l
Robert A. Clark L
Trojan Nuclear Plant Decket 50-344 June 24, 1982 l
License NPF-1 Attachment Page 2 of 8 Missile Penetration - The capabi)Ny of, barriers to 3.
resist missile penetration was evaluated using the modified Petri Formula.
(Ref FSAR Sections 3.3.2 and 3.5.3.4.)
Of the Trojan criteris missiles, the I
4-in. x 12-in. wood plank and the 3-in.-diameter f
Schedule 40 pipe were determined to have the., largest penetrating potential and, hence, were used'in the evaluations. Since the modified Petri Formula. is independent of coexistent gross loads, use of the.
SRP load combinations would have no effect on the evaluation conclusions for missile penetration.,
4.
Missile Impact Overall Structural Response - The 108-lb wood plank and the 76-lb pipe were excluded from evaluations for overall structural response of missile impacts because of their relatively lowmass compared to that of the barrier. Structural response calculations were performed for impact' of' the 4000-1b car. The evaluations used an energy balance, technique available at the time, but different from methods presently used for such analyses.
The combination of velocity pressure loads with mis-sile impact effects was not a specific criteria requirement in the original evaluations, and missile structural response was considered separately.
The ductility demand for a barrier is, of ' course, pre-dominantly caused by the govern (ng. missile impact, and the additional demand due to' velocity pressure is relatively small.
Pursuant to the NRC Structural Branch question, a reanalysis of over-all structural response under combined tornado / loads has been performed using load combinations in accordance with'SRP Section 3.3.2.
Con-ventional analytical techniques were used to determine overall s
structural response. An equivalent dynamic mooel was developed for
~
each wall panel, and structural response was calculated using a time-history approach (response chart method).
Impact force time-histories were derived from the. available experimental data as provided in " Full
,/
Scale Tornado Missile Impact Tests", Stephenson; A. E...EPRI Report No. NP-440, Sandia Laboratories, July 1977; and'" Missile Impact Testing of Reinforced Concrete Panels", Calspan Report No. HC-5609-D-1, Calspan Corporation, Buffalo, New York, January 1975. Resulting maximum deflec-tions calculated from the time-history analyses were used to determine l'
the maximum ductility demands. The maximum allowable ductility ratio was limited to 10.
The reanalysis has shown that tornado protection at Trojan is adequate to withstand combined-load structural response demands in accordance with SRP Section 3.3.2, except as described in,the answer to Question 8.
/
Trojan Nuclear Plant Robert A. Clark Docket 50-344 June 24, 1982 License NPF-1 Attachment Page 3 of 8 Question 7:
The FSAR spectra of tornado-generated missiles consists of only three missiles. The licensee should assess the impact of consid-ering the missile spectra given in Standard Review Plan Section 3.5.1.4 and provide the information for staff's review.
Answer:
In response to the Branch request, an example analysis has been performed to illustrate flexural ductility demands, using load combinations in accordance with SRP Section 3.3.2., for both the Tr ojan FSAR tornado missile spectrum and the missile spectrum recommended in SRP Section 3.5.1.4.
The barrier considered in the example is a 16-in.-thick, reinforced double-wythe masonry wall panel, 15-ft-high and 19-ft 3-in. wide, having continuous top and bottom supports and free edges. This barrier is representative of typical wall panels of the north wall of the Auxiliary Building (Column Line 46, El. 45 ft 0 in, to El. 77 ft 0 in.).
The wall panel was analyzed for concurrent missile impact and wind load effects corresponding to a 200-mph tornado.
Results of the example analysis are summarized in Table 7-1 (attached).
Ductility ratios would, of course, be numerically different for other wall panel geometry and conditions of reinforcing steel continuity at boundaries (ie, one-way simple, one-way simple-fixed, plate action, etc), but a general comparison of SRP 3.5.1.4 and Trojan FSAR tornado missile structural response effects can be made on the basis of the results shown in Table 7-1.
For missiles that are essentially common to both criteria (wood plank, 3-in.-diameter steel pipe, and automobile), it is apparent that no difference in structural response conclusions for any wall panel configuration should result because the automobile missile governs ductility demands and the parameters that determine energy are identical.
For the SRP utility pole missile, the ductility demand signifi-cantly exceeds that for the automobile missile. Hence, regardless of geometry and boundary conditions, some of the Trojan tornado missile barriers, which are within the postulated height trajectory of the utility pole and automobile (taken as 25 ft above adjacent j
grade for both),-will not withstand the utility pole missile at a ductility ratio not exceeding 10. Above the elevation where the SRP utility pole governs, similar conclusions would be reached with respect to the SRP 12-in.-diameter steel pipe.
Conclusions with regard to the Trojan FSAR tornado missile spectrum based on structural response analysis of all tornado missile barriers are presented in the answer to Question 8.
Trojan Nuclear Plant Robert A. Clark Docket 50-344 June 24, 1982 License NPF-1 Attachment Page 4 of 8 TABLE 7-1 EXAMPLE COMPARISON OF STRUCTURAL RESPONSE FOR STANDARD REVIEW PLAN
- AND TROJAN FSAR TORNADO MISSILES; TOTAL TORNADO VELOCITY = 200 MPH SRP TROJAN FSAR MISSILE Wt Vel Wt Vel (1b)
(ft/sec)
(lb)
(ft/sec)
A.
Wood Plank, 4" x 12" x 12'-0" 200 235 1.6 108 293 1.0 B.
Steel Pipe, 3" dia, Sch. 40, 10'-0" 78 117 0.9 76 110 0.7 C.
Steel Rod, 1" dia, 3'-0" 8
176 0.1 D.
Steel Pipe, 6" dia, Sch. 40, 15'-0" 285 117 1.4 E.
Steel Pipe, 12" dia, Sch. 40, 15'-0" 743 117 5
F.
Utility Pole,13-1/2" dia, 35'-0" 1490 117 13.5 2
G.
Automobile, frontal area 20 ft 4000 58.7 8.1 4000 58.7 8.1
'4
- SRP Section 3.5.1.4, Rev 2, Missile Spectrum A.
- p-ductility demand in flexure.
Trojan Nuclear Plant Rob:rt A. Clark Docket 50 !*4 June 24,1982 License e'>
Attachmen't Page 5 of 8 Question 8:
The licensee should clarify, with regard to missile protection, whether the masonry missile barriers protecting single trains of safety-related systems were evaluated to determine their design for adequacy.
If this was not done, provide the basis for justification.
Answer:
All areas containing single trains of Plant safety-related systems needed to achieve and maintain safe shutdown of the plant are also protected by a minimum of a 16-in.-thick, double-wythe masonry wall missile barrier, or the equivalent (see the answer to Question 9 below with regard to the exceptions noted in Attachment 2 to Licens-ees October 23, 1980 letter). For these areas, tornado missile bar-riers were reanalyzed for the Trojan FSAR tornado missiles, using the SRP Section 3.3.2 load combinations. The reanalysis has shown that all barriers can resist combined tornado loads at ductility ratios not exceeding 10, except for barriers at two locations described below.
One exception is in the vicinity of the spent fuel pool heat exchanger room (Fuel Building, El. 61 ft 0 in.) where portions of 14-in.-thick reinforced double-wythe masonry wall, which protects a section of 4-in. service water system piping, have been determined to have ductility ratio demands exceeding 50. Modifications will be performed to provide additional tornado missile protection at this location. These modifications are currently scheduled to be completed by January 1,1983.
The other exception is in the area of the component cooling water (CCW) heat exchangers (Fuel Building, El. 45 ft 0 in.) where, because of the presence of wall openings for CCW heat exchanger maintenance (the openings have steel plate missile shields), two 16-in.-thick reinforced double-wythe masonry wall panels have been determined to have ductility demands of 15.5 to resist impact of the 40-mph, 4,000-1b automobile missile.
(It is to be noted that no credit has been taken for the existing precast concrete panels along Column Lines 41, A, and a portion of D because it cannot be shown that these panels will resist the wind and missile loads per SRP 3.3.2.)
We consider the combined probability of a tornado in the region of Trojan, generation of an automobile missile, and missile impact at this specific location to be extremely low. As a result of this low probability, the present wall design in the CCW heat exchanger area is considered acceptable.
Trojan Nuclear Plant Robert A. Clark Dockst 50-344 June 24, 1982 License NPF-1 Attachment Page 6 of 8 Question 9:
The licensee should upgrade the missile protection at locations identified in Attachment 2 to its October 23, 1980 letter. The upgrading should be based on agreed-upon criteria between the NRC staff and the licensee. As general guidance, the SRP Section 3.5.3 criteria should suffice for the upgrading work.
Answer:
Additional missile protection, in the form of structural steel plate barriers, was provided at the locations identified in Attachment 2 to licensee's letter dated October 23,1980 (field work completed in July,1981). Load combinations used for the missile protection design were in accordance with SRP Section 3.3.2.
Barrier design for local effects was performed in accordance with BC-TOP-9A, Rev 2, using methods which are consistent with those referenced in SRP Section 3.5.3.
Trojan Nuclear Plant Robert A. Clark Docket 50-344 June 24, 1982 License NPF-1 Attachment Page 7 cf 8 Question 10:
With regard to the use of expansion anchor bolts and local load capacity, the licensee has made references to currently ongoing test programs by other organizations. The licensee should submit these test results, along with the justification for their applica-bility to situations at Trojan, to confinn the conservatism of its criteria.
Answer:
As background information, a brief description of expansion anchors used at Trojan and the IE Bulletin (IEB) 79-02 results will first be provided.
Expansion anchors used in the original construction of the Trojan Nuclear Plant were the ITT Phillips self-drilling type.
These were used in both concrete and concrete masonry. During the period from 1975 to 1979, a small number of ITT Phillips wedge-type anchers were used in Plant modifications.
The PGE response to IEB 79-02, dated November 21, 1979, contained Table 1-1, which listed results of a complete reanalysis of all large pipe supports. As a result of this analysis, pipe support base plate expansion anchor bolts found to have a factor of safety of less than 5.0 were modified.
(As described in licensee's response to IEB 79-02, the allowable values for expansion anchors in masonry were obtained using a coefficient of 0.60 applied to equivalent values for concrete based on data from ICB0 Standard 1372, February 1975.) After accounting for these modifications, the data in Table 1-1 can now be summarized as follows:
Shear-Tension Interaction Ratio Range Factor of Safety
- Number of Supports Percent of Total 0 < 0.25
>10.0 1,086 75 0 < 0.50
>7.1 1,264 87 0 < 0.75
>5.8 1,354 94 0 < 1.0
>5.0 1,445 100
- Factors of safety are based on the most highly loaded bolt in a pipe support base plate.
1
Trojan Nuclear Plant Robert A. Clark Docket 50-344 June 24, 1982 License NPF-1 Attachment Page 8 of 8 l
In addition to pipe support modifications performed pursuant to IEB 79-02, a considerable number of supports have also been modi-l fled more recently as part of IEB 80-11 (masonry wall) field work and Control Building modification field work. The results of those additional pipe support anchor bolt modifications are not reflected in the above table, but would represent a further increase in the factor of safety for some of the pipe support anchor bolts, as well as a decrease in the total number of pipe supports which utilize expansion anchors.
The table shows that only a small percentage of pipe supports at Trojan have expansion anchors with factors of safety near 5.0, and the majority have factors of safety significantly greater than 5.0, based on the IEB 79-02 allowable values (greater than 10 for 75 per-cent of the installations, etc). This small percentage includes both supports in masonry and concrete and a number of supports which subsequently have been modified.
Expansion anchor bolt testing that has been conducted recently by others has been performed mostly in concrete (TVA, HEDL survey, etc).
One test program has been performed in concrete masonry of the type used at Trojan. Unfortunately, the test report is proprietary to the organization that sponsored the testing.
In summary, we believe that the factors of safety in the Trojan pipe support base plate expansion anchor bolts are sufficiently large to account for variables in design and construction in accordance with the intent of IEB 79-02.
TEB/nem A-24.13B4