ML19340D049
| ML19340D049 | |
| Person / Time | |
|---|---|
| Site: | McGuire, Mcguire  |
| Issue date: | 12/03/1980 |
| From: | Parker W DUKE POWER CO. |
| To: | James O'Reilly NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| References | |
| IEB-79-02, IEB-79-2, NUDOCS 8012180674 | |
| Download: ML19340D049 (59) | |
Text
_ _ - _ _ _ _ - _ _ _ _ _ _ _
7,q DUKE POWER COMPANY Powsm Dust.ntxo 37 3 pvg 422 SocTn Gnuncu Stater. Ciunt.oTTz. N. C. asaiaChii e e.
wi wa m o. ** a a r a. s a.
December 3, 1980
,3 beg 9 El 2 56 vice Persierav fetap=o=c: Aata 704 373-4cea seta= Paoovevion
, ; IiEC
, MTiOH SERVICE 5 c~,
~
Cl LCM Mr. J. P. O'Reilly, Director U. S. Nuclear Regulatory Commission Region II 101 Marietta Street, Suite 3100 Atlanta, Georgia 30303 Re: McGuire Nuclear Station Docket Nos. 50-369, 50-370 IE Bulletin 79-02
Dear Mr. O'Reilly:
Attached is Revision 3 of Duke Power Company's response to IE Bulletin 79-02.
This response includes the results of a statistical evaluation of anchor bolt safety factors which was discussed in our meeting with NRC representatives on August 8, 1980. With the submitfal of this information, all known outstanding issues ~have been addressed.
If you have additional questions regarding this matter, please advise.
Very truly yours, n
A d William O. Parker, Jr.i GAC:scs Attachment cc: Director, Office of Nuclear Reactor Research Director, Inspection and Enforcement Headquarters T. J. Donat, NRC Senior Resident Inspector, McGuire 8012180 Q
a i
MCGUIRE NUCLEAR STATION Responses to USNRC IE Bulletin 79-02,Revisfor 3, Original: July 2, 1979 Revision 1:
January 7, 1980 Revision 2:
July 24, 1980 Revision 3: December 1, 1980 McGuire Nuclear Station is in the later stages of construction and very near completion and fuel load of Unit #1.
Essentially all pipe supports have been erected in Unit #1 and a large number have been erected in Unit #2.
The l
following is a summary, by item, of the extent and manner in which Duke Power l
Company intends to satisfy Actions 1 through 9 of the IE Bulletin 79-02, Revision 2.
Response 1:
Duke Power Company will account for base plate flexibility in the calculation of expansion anchor bolt loads for all Seismic l
Category I pipe support base plates using either a conservative l
hand Calculation method which has been verified by non-linear finite ele ant analysis or a specific non-linear finite element analysis for a particular base plate. The models and boundary conditions, including appropriate load displacement character-istics of the anchors, used for the finite element analyses, are based on Duke studies and on work performed by Teledyne Engineering Services which was sponsored by a group of thirteen (13) utilities formed to respond to generic items of IE Bulletin 79-02. All expansion anchor support plates designed prior to implementing these analysis methods are being reanalyzed accord-ingly and will be modified if required to comply with allowable anchor bolt loadings.
I Response 2:
The minimum factors of safety, between the expansion anchor bolt design load and the bolt ultimate capacity determined from i
l M atic load test, used in Duke's design of pipe supports, are as follows:
4 Normal Conditions 4
Upset Conditions i
Faulted Conditions * -
4 These factors of safety are for wedge type and sleeve type expansion anchors.
Some shell type anchors were used in the early stages of McGuire construction.
Use of shell type anchors for Nuclear Safety Related applications was disconnected in February, 1975.
Duke Power Company has identified all pipe supports using shell type achors and the design of these sup-ports has been reviewed to assure that a minimum factor of safety of five (5) is maintained.
- This is based on a 95% confidence level that no more than 5% of the bolts on a nuclear safaty related system have a safety factor less than 4.
1 l
l
l McGuire Seismic Category I expansion anchor installations are restricted to normal weight structural concrete of varying strengths.
Expansion anchor bolt ultimate load capacities are based on manufacturer's test results and recommendations for normal weight concrete and installed concrete strengths.
McGuire Seismic Category I expansion anchor designs properly account for shear-tension interaction, minimum edge distances and bolt spacing in accordance with manufacturer's test results and recommendations.
Response 3:
Duke Power Company designs pipe supports to resist all appli-cable loading including seismic loads, hydro test loads, normal operating loads, thermal loads, etc.
A support is designed for a static or quasi-static load resulting from the most critical combination of the applicable loadings.
The safety factors used for the expansion anchors are as specified in Response 2.
Duke Power Company co-sponsored tests performed by Teledyne Engineering Services to demonstrate that expansion ar-hors installed at McGuire Nuclear Station will perform adequately under both low cycle /high amplitude loading (seitmic) and high cycle / low amplitude loading (operating). The final test report was generically submitted to USNRC for all Duke Power Company Nuclear Stations as described in Mr. W. O. Parker's (Duke) letter to Mr. J. P. O'Reilly (USNRC, RII) dated August 29, 1979 regarding McGuire Nuclear Station.
Response 4:
Duke Power Company has developed and is continuing to develop sufficient documentation to verify that expansion anchors used in Nuclear Safety Related pipe supports are the correct size l
and type and are properly installed in accordance with manufac-turer's recommendations.
The following is a summary of documen-tation developed:
In February 1977, Duke Power Company initiated some random testing of installed expansion anchors.
This testing was I
performed in response to concerns developing in the industry about improper installation practices.
Based on these tests, Duke decided that a formal inspection program for concrete l
expansion anchors would be implemented.
l In March 1977, Construction Procedure CP-503 was issued for wedge, sleeve and self-drilling type concrete expansion anchor l
inspection. There were four criteria to be met:
i
- 1) spacing, 2) perpendicularity, 3) torque, and 4) embedment depth, except for self-drilling anchors which had no specified torque.
In June 1977, inspection was initiated to check anchors installed prior to issuance of CP-503.
All anchors for all pipe supports not having documentation in accordance with CP-503 were inspec-ted in accordance with CP-503 and documented. A sampling of other types of attachments using expansion anchors was also i
made.
A total of 4357 anchors were inspected, 2072 of which.
were pipe supports.
This inspection was completed in September 1978.
2
l In August 1977, QA Procedure M-52 was issued which supersedes CP-503 as the tpplicable inspection procedure for concrete expansion anchors. 'M-52 stated that only anchors greater than 5/8" 9 required torque inspection and this was later revised to all anchors with a specified torque greatsr than 100 ft-lbs.
This reduction in torque inspection was based on the results of the testing performed in which only 2 of the 4357 anchors failed to meet torque requirements.
During site inspection visits the NRC has expressed concern l
regarding torque requirements for wedge and sleeve anchors.
This concern is based in part on observations documented in IE Inspection Reports 50-369/80-04 and 50-369/80-18. We have concluded that thg anchor bolts in question did receive an initial inspection torque sufficient to set the anchor. This i
conclusion is supported by subsequent inspection of a selected number of supports (reference Duke responses dated June 5.
l 1980; June 30, 1980; July 18, 1980; and September 5, 1980).
As stated in our September 5,1980 response, non-conforming Item Report No. 11,667 was initiated to obtain further management review of the anchor bolt torque criteria.
This NCI was cleared October 17, 1980 and reviewed at the site by the NRC on October 20, 1980. The conclusions are that no reportable item exists and no changes to erection or inspection procadures are warranted.
Some background on anchor bolt torque is necessary to put the basis for that procedure in perspective.
Torque is necessary to positively demonstrate that the anchor bolt has expanded fully.
Construction Procedure 308 specifies the installation torque to be applied.
The importance of any subsequent inspec-tion is not to assure a certain retained torque value but simply to identify any situation where a craftsman mistakenly failed to initially apply the installation torque. In order that inspectors would have a quantitative value for their check, CP-308 and Inspection Procedure M-52 specify an inspec-tion value which is 70% of the installation value. The 70%
value was established based on experience which indicates that post-torque relaxation is generally less than 30%.
Failure of the bolt to satisfy the torque inspection criteria does not imply that the bolt's load carrying capability is reduced, unless several turns are required to achieve acceptable torque.
j In recognition of the fact that the purpose of torquing the anchor bolt is simply a one time setting, Procedure M-52 speci-fies that if an inspector witnesses a craftsman applying the installation torque, then application of the inspection torque is waived. In further recognition of the one-time setting aspect, Procedure M-52 specifies that if an anchor bolt rotates when the inspection torque is applied, it shall be re-tightened to the installation torque and verified by the inspector.
CP-308 specifies that each anchor bolt will have an installa-tion torque applied.
Procedure M-52 specifies that each anchor bolt's torque will be verified.
In addition, Duke representa-tives described to NRC representatives on July 15, 1980, in the 3
l Region II offices, the 100% reinspection program for all safety related pipe supports.
For those supports with anchor bolts, this included anchor bolt torque.
Based on our inspection programs, reinspection programs and selected sample tests, we are reasonably confident that anchor bolts at McGuire Nuclear Station had sufficient torque applied to set the anchor during the erection / inspection process.
In April 1975, Construction Procedure CP-308 was issued to j
provide control over the installation of concrete expansion anchors.
This procedure has been updated periodically to reflect the experience gained by Duke through its inspection and testing programs.
Self-drilling shell type expansion anchors were installed in accordance with the manufacturer's recommended installation procedures. Adequate embedment depth and full expansion of the shell is assured since the anchor shell itself is used to drill its own hole and the shell is driven below the surface of the wall in the final installation step. Shell type anchors were inspected for size, type, perpendicularity, spacing and bolt snugness.
Response 2 indicated that Duke has identified all Seismic Category I pipe supports using shell type anchors.
Duke implemented a shell type anchor inspection program in accordance with IE Bulletin 79-02, latest revision, to supple-ment existing documentation. The parameters inspected were bolt thread engagement, shell shoulder to plug measurement, perpendicularity and bolt hole size, in addition to pull l
testing a 3% sample of visually acceptable shell type expansion anchors in each system to confirm that the visual inspection program is sufficiently regorous to identify any deficiency having a significant effect on load carrying capability of the This in' pection and testing program is outlined in anchor.
s McGuire Nuclear Station Specification MCS-1196.02-00-0003.
A total of 52 supports with 242 self-drilling anchors were inspected at McGuire Unit 1.
Four (4) supports were in the Diesel Generator Building, five (5) were in the Reactor Building and the remainder were in the Auxiliary Building. The 3% pull test sample was completed with 8 anchors being tested.
None of the sample anchors failed the pull test. Thirteen (13) of the 242 self-drilling anchors were found to have significant defi-ciencies which had the potential for degrading their ultimate load carrying capability.
Plate bolt hole size is specifically not inspe-ted as part of the Duke expansion anchor inspection program.
In response to Revision 1 of IE Bulletin 79-02, Duke inspected 331 bolt holes in 104 plates which utilized either sleeve or wedge anchors to confirm that p? Ce bolt hole sizing was not a problem.
Seven (7) plate bolt noles were found to be slightly undersized and 38 were found to be slightly oversized from design drawings.
All of the plate bolt hole sizes were acceptable. Duke has 4
concluded that this test sample provides reasonable and ade-quate assurance of proper plate bolt hole size for wedge and sleeve type expansion anchors.
The supplemental Self-Drill Inspection Program implemented under MCS-1196.02-00-0003 identified 55 of 191 plate holes inspected as oversized.
This oversizing was determined to be due to the self-drill anchor installation procedures.
All oversized holes have been reviewed and modification made where required.
In order to address the question of the relationship of cyclic /
load carrying capacity to installation procedure (anchor pre-load), the tests referred to in Response 3, performed by Teledyne Engineering Services and sponsored by the group of thirteen (13) utilities, have been performed on anchors installed in accordance with manufacturer's recommended installation proce-dures and have no more preload than is provided by the use of these procedures.
Based on Duke's understanding of the behavior of expansion anchors and on cyclic testing which has been performed, Duke Power ~ Company is confident that the anchors will perform adequately.
Some pipe supports with anchor bolts were physically inacessible for inspection under the provisions of this bulletin. These have been independently assessed to verify that they are not reasonably accessible.
Duke has concluded that there is reason-able and adequate assurance that these supports will perform adequately.
This conclusion is based on the results of inspec-tions of other supports and the long history of documentation associated with pipe supports.
Specifically, 115 supports with
~
and without anchor bolts out of a total of over 15,000 supports were not accessible.
In this group of 115, all of the supports which have anchor bolts were installed under Construction Procedure 308 and inspected under Construction Procedure 503 or QA Procedure M-52.
It is Duke's position that this documenta-tion assures the proper installation of the anchor bolts on inaccessible supports and satisfies the requirements of IEB 79-02.
Response 5:
Nuclear Safety Related/ seismic pipe supports are prohibited from being attached to block (masonry) walls using concrete expansion anchors.
In response to Revision 2 of IE Bulletin 79-02, Duke Power Company has conducted an on-site confirmatory review at McGuire Unit 1 of Nuclear Safety Related/ seismic pipe supports to assure that no such installations exist.
Results of this review have confirmed that there are no such installa-tions of this type at McGuire Nuclear Station Unit 1.
Response 6:
The expansion anchor installation and inspection procedures utilized at McGuire Nuclear Station and described in Response 4 apply to all expansion anchors installed in Nuclear Safety Related pipe supports.
Each expansion anchor is inspected regardless of the physical configuration of the steel members 5
being connected to the concrete.
These supports are included in the actions being performed by Duke Power Company to satisfy the requirements of IE Bulletin 79-02.
Response 7:
McGuire Nuclear Station is currently under construction, there-fore Bulletin Item 7 is not applicable.
Response 8:
McGuire Nuclear Station is currently under construction, there-fore Bulletin Item 8 is not applicable.
Response 9:
Those pipe supports which have not been installed are included in actions performed to meet the requirements of IE Bulletin 79-02 as outlined in Responses 1 through 6.
Revision 2 of Item 2 of the Bulletin requests verification by Duke Power Company that a uniform factor of safety was applied for all load combinations in the design of expansion anchors for McGuire Nuclear Station.
The expansion anchor design factors of safety utilized are outlined in Response 2 and the results of a detailed statistical analysis are presented in an enclosure to this response, A
l l
l l
6
I l
l 7
Duke Power Company McGUIRE NUCLEAR STATION ANCHOR BOLT SAFETY FACTOR ANALYSIS FOR USNRC I & E BULLETIN 79-02 AUGUST 8,1980 REVISED DECEMBER I,1980
)
raw g
LXJ l
g a
TABLE OF CONTENTS 1.0 HISTORICAL BACKGROUND
2.0 INTRODUCTION
3.0 MARGINS INHERENT IN ESTABLISHED PIPING SUPPORT / RESTRAINT DESIGN / ANALYSIS PROCESS 3.1 Energy Inputs to the Analysis Problem 3.2 Building Analysis 3.3 Piping Analysis 3.4 Suppor / Restraint Analysis 4.0 EXPANSION ANCHOR ALLOWABLES AND FACTOPS OF SAFETY 5.0 BASIS FOR MCGUIRE FAULTED FACTOR OF SAFETY 6.0 EFFECT OF TENSION ZONE INSTALLATION 7.0 DESIGN AND CONSTRUCTION INSTALLATION PRECAUTIONS 8.0 CASE-BY-CASE REV1EW OF SUPPORTS / RESTRAINTS 9.0 IMPACT OF FACTOR OF SAFETY RETROFIT 9.1 Backfit Process / Impact For Individual Support / Restraint Design, Erection And Inspection 9.1.1 Screening Process 9.1.2 Engineering Evaluation 9.1.3 Redesign 9.1.4 Construction Rework 9.1. 5 Reinspection -
TABLE OF CONTENTS (CON'T) l 4
3 9.2 Backfit Process / Impact for Pipe Stress Math Model Review 9.3 Backfit Process / Impact for QC Final Walkdown
10.0 CONCLUSION
S 11.0 SYSTEM BY SYSTEM ANALYSIS 12-01-80 11.1 Sumary of August 8,1980 Meeting' 11.2 Statistical Approach - Binomial Distribution 11.3 Samplina Method 11.4 Results 11.5 Impact on Fuel Loading and Full Power Operation 11.6 Plan of Corrective Action.~.
i ATTACHMENTS - USNRC IE Bulletin 79-02, March 8, 1979, pg. 2 of 3 - USNRC IE Bulletin 79-02, Revision 2, November 8,1979, pg. 3 of 7 - McGuire Nuclear Station FSAR, Figure 2E-20 - Ground Response Spectrum vs Synthetic Time History Spectra
' - McGuire Nuclear Station FSAR, Figure 3.7.2-31 - Graph of Faulted Factors of Safety, FSp, Above Which Faulted Load Combination Will Govern - Expansion Anchor Factor of Safety Review - Histogram and Cumulative Distribution Function of Minimum Faulted Factors of Safety For Sample of Supports / Restraints Failing Screen.
- - Safety Factor Backfit Process - Individual Support / Restraint Implementation Process. 0- Safety Factor Backfit Process - Piping Stress Analysis Math Model Review 1-Safety Factor Backfit Process - Construction Q C Final Walk-down. 2 - Minimum Sample Size 3 - Results
- Rigorous Systems and Alternate Analysis Systems I
-iii-
l i
1.0 HISTORICAL BACKGROUND On March 8,1979, USNRC submitted I & E Bulletin 79-02 to Duke Power Company's McGuire Nuclear Station requiring a written response in 120 days.
Item 2 of the Bulletin, shown as Attachment 1, requested verification that wedge and sleeve type expansion anchors have a minimum design factor of safety equal to four (4) and that shell type expansion anchors have a minimum design factor of i
safety equal to five (5). The load combinations, for which these factors of safety are applicable, were not specified. Duke did not consider the lack of load combination specificity to be unusual because it is the general industry practice when utilizing the working stress design method to establish allowa-bles for service load conditions and then to appropriately increase these al-lowables for factored load conditions. This is the general approach followed by USNRC Standard Review Plan Sections 3.8.3, 3.8.4, 3.9.3 and standard codes of Engineering practice such as AISC, ACI, ASME, and the McGuire Nuclear Sta-tion FSAR. On July 2,1979, Duke submitted its response to the Bulletin for McGuire Nuclear Station. The response to Item 2 verified that the design cri-teria for McGuire required a factor of safety of four (4) for wedge and sleeve expansion anchors in the service load condition with appropriate revision of allowables for factored load conditions. The response also included a comit-ment to retrofit all supports / restraints necessary to comply with the higher facter fo safety of five (5) for shell type cxpansion anchors required by this Bulletin.
On November 8,1979, USNRC issued Revision 2 to the Bulletin, shown as Attach-l ment 2, which revised Item 2.
The revision to Item 2 indicated that the fac-tors of safety issued in the original bulletin were intended to apply to all load combinations regardless of load condition. Duke responded to this revi-sion of the Bulletin on January 7,1980. Duke response to Item 9 of the Bulle-tin identified the revision to Item 2 as a change in the Bulletin with which the design criteria for McGuire Nuclear Station was at variance. A brief out-line of the bases on which Duke established its design criteria for expansion anchors was provided.._.
I
2.0 INTRODUCTION
Section 1 provides the historical development in identifying the apparent con-flict between the original design criteria utilized by Duke Power Company for designing support / restraint expansion anchors and the requirements of Item 2 of USNRC I & E Bulletin 79-02.
In an effort to resolve this conflict Duke Power Company has conducted a review of the McGuire Nuclear Station expansion anchor design criteria and its implementation for piping support / restraint designs.
The design factor of safety (FS) of an expansion anchor is defined as the an-
{
chor ultimate load capacity (Pu) detennined by static load tests, which simu-late the actual conditions of installation, divided by the anchor design load (P). Pu is generally well defined for a given facility based on actual tests and FS is similarly well-defined as set down in pertinent design criteria. P is less well defined in the context of Nuclear piping support / restraint design because of the number of variables and assumptions which are introduced into the complex analysis process utilized to establish it. P is, however, assured to be conservative since each variable and assumption employed in the analysis process is etudied thoroughly and conservatively established. Unfortunately, because of the complexities involved, the setting of conservative assumptions is often done by isolated studies which can not consider the design / analysis l
process as a whole and often results in a final accumulated margin of safety in excess of that which is reasonable because of compounding margins.
The Duke review of pipe support / restraint expansion anchor factors of safety carefully reviewed the margins present in establishing Pu, P and FS specific-ally for McGuire Nuclear Station. The following sections outline the results of this review and conclusions regarding adequacy of the McGuire piping support /
restraint expansion anchor designs in view of the criteria required by Revision 2 of USMC I & E Bulletin 79-02.
l -
3.0 MARGINS INHERENT IN ESTABLISHED PIPING SUPPORT / RESTRAINT DESIGN / ANALYSIS PROCESS The expansion anchor design load, P is established through an extensive analy-sis process outlined in Section 2 and 3 of the McGuire FSAR. Conservative as-stsnptions are employed in four (4) broad categories of analysis including energy inputs to the problem, building analysis, piping analysis and piping support /re-straint analysis. Since the factor of safety in the factored load condition is of prime consideration in this review, the nost intensive review effort centered on margins present in the analysis for Safe St.utdown Earthquake loads.
When a margin is introduced at one step of the seismic analysis process, it is important to note that this does not imply a unifonn margin is maintained throughout the prot,lem or that the same margin will have a unifom effect on the outcome of another problem, similarly analyzed. This is the nature of the dynamic analysis model whose behavior is a function of the mass distribution, stiffness, eigenvalues and eigenvectors of the system. Therefore, some of the margin inherent in the design process must be discussed qualitatively since the resultant margin present in the expansion anchor load is dependent on the trans-j missibility of the particular dynamic model and would require a case-by-case j
analysis to quantify. Many of the margins, however, can be quantified and will be discussed in this context, when possible.
3.1 ENERGY INPUTS TO THE ANALYSIS PROBLEM Appendix 2E of the McGuire FSAR thoroughly outlines the conservative seismol-ogical bases upon which the site design ground response spectrum is developed.
As described in Section 3 of the McGuire FSAR it is necessary in the analysis process to generate "in-structure" response spectra as input to the piping analysis. This process requires the generation of a synthetic earthquake motion time history to which the building is subjected. The only requirement is that the response spectra of the synthetic time history envelope the site design j
ground response spectrum. graphically demonstrates the relationship between these curves i
for McGuire Nuclear Station. Attachment 4 lists the input energy margins intro-duced at this step in the analysis process...
3.2 BUILDING ANALY3IS The building structural damping, for generation of "in-structure" response, is taken as 5% of critical for the Safe Shutdown Earthquake in accordance with Section 3.7.1.3 of the McGuire FSAR. USNRC Regulatory Guide 1.61 would re-quire damping equal to 7% of critical as reasonable and conservativa.
The "in-structure" response spectra are peak broadened in accordance with USNRC Regulatory Guide 1.122. However, the peak is also amplified 10% which is in ex-cess of this guide. Attachment 5 is an example of a typical broadened and am-plified "in-struc%ure" response spectra used as input to the piping analysis problem.
3.3 PIPING ANALYSIS The margins discussed in Sections 3.1 and 3.2 are introduced into the rigorous piping analysis problem through the "in-structure" response spect:a inputs at the piping support / restraint locations. Rigorously analyzed piping systems are generally flexible and possess many modal frequencies below the rigid range.
This generally leads to the occurrence of multiple piping frequencies coinciding with the peak frequencies of the input spectra due to the peak broadening dis-cussed in Section 3.2.
This would imply that multiple frequencies of the pipe system are in reasonance with one building frequency, which is physically im-possible. A method to prevent the introduction of excess conservatism, in this I
situation, is discussed by Hadjian. Additional margin was introduced at this step of the analysis since McGuire piping analysis permitted multiple piping frequency excitation by the broadened input response spectra peak. This margin is particularly pronounced because the response spectra peak to which each coin-cident piping mode is excited was developed with conservative structural damping and was then amplified 10% as discussed in Section 3.2.
If these modes are closely spaced, in addition, then modal responses were combined absolutely in lieu of square root of the sum of the squares.
I Hadjian, A H, "Some problems with the Calculation of Seismic Forces on Equip-ment". proceedings of the Specialty Conference on Structural Design of Nuclear Plant Facilities, Volume II, December 17-18, 1973.
_-_=_--__ -
Some of the expansion anchors which are anticipated to possess faulted factors of safety less than four (4) are installed in support / restraints whose loads are derived from analysis load combinations requiring the absolute sumation of the peaks of several dynamic events, assumed to occur simultaneously. Considerable effort is now underway within the industry and USNRC to reduce the conservatism inherent in this approach, if possible. This question is currently under study by USNRC under NRR Task Action Plan B-6, " Loads, Load Combinations, and Stress l
l Limits", the Reactor Safety Research Seismic Safety Margins Research and Load Combination Programs. The recently issued Revision 1 to NUREG-0484 is a major j
sap forward in this regard.
i McGuire piping thermal analysis was generally conducted using the design temperature condition in lieu of the lower normal operating temperature for establishing support / restraint loads. This results in the thermal component of the support / restraint load representing a conservative valve. This provides some flexibility for future revision to operating temperatures without signifi-cant plant impact.
Approximately 40". of the support / restraints for small bore piping have been designed using loads generated by alternate analysis procedures. This is an enveloping seismic analysis technique whose advantage is speed, simplicity and application using hand calculations. The offsetting penalty is the enveloping i
conservative assumptions which must be employed in develeping the procedure such that it has a reasonably convenient general application with minimum complexity.
j i
Numerous case studies have been conducted by Duke to establish the overdesign resulting from the application of this technique as compared with typical com-I puterized rigorous analysis. The results of these studies show the alternate approach to require significant additional supports / restraints on a given pipe run with the load occurring on each of these support / restraints to be con-siderably higher than that obtained from rigorous analysis.
3.4 SUPPORT / RESTRAINT ANALYSIS Review of the support / restraint analysis techniques and s 'itiera utilized on McGuire Nuclear Station identified several items which introduced considerable i
margin into the design load, P, established for the expansion anchor. Most of these were simplifications in design office procedures which enhanced the
design process at the expense of overestimating the expansion anchor load.
Baseplate shear was assumed to be taken equally by the expansion anchors as if the baseplate was installed on a " frictionless" surface. Generally, shear is counteracted by plate / concrete friction developed as a result of the interface compressive force between plate and concrete. This compressive force exists 2
due to expansion anchor installation preload and/or the requirement of equili-brium accompanying baseplate moment.
If the applied shear is in excess of the friction force which can be developed, the excess shear would be expected to be taken by the anchors located in the compression zone beneath the plate based on the relative stiffnesses of the assembled paits.
Inclusion of the full shear force applied to the expansion anchor in the shear-tension interaction con-servatively reduces the computed factor of safety.
The hand calculational technique used to incorporate the flexibility considera-tions of USNRC IE Bulletin 79-02 was developed with considerable margin so as to render it simple and general enough for design office use. This technique was based on the finite element calculational procedure developed by the Teledyne Utility Gcoup on USNRC IE Bulletin 79-02. The experimental work con-ducted by this group similarly confimed the finite element approach to be amply conservative.
Many McGuire support / restraint expansion anchor loads have been established based on an " allowable load" which represents the maximum loading to which the support / restraint has been qualified. The final design loads for the support /
restraint are often less than the " allowable load" values and therefore re-calculation of the expansion anchor factors of. safety was not necessary. Un-fortunately, this is not evident when tabulating current expansion anchor factors of safety.
Friction forces applied to the support as a result of thermal movement of the piping were assumed conservatively to occur in all directions, irrespective of the direction or magnitude of pipe movement. The friction force was established using the applicable coefficient of friction. However, compatibility of support /
restraint and pipe displacement were conservatively neglected for simplicity.
i !
.. _. _. _, - ~. - -. _._
The shear-tension interaction utilized for McGuire expansion anchors was linear.
Based on shear-tension interaction work performed by the aforementioned utility group, a less severe elliptical shear-tension interaction relationship has been shown to be technically well-founded.
[
l i
l I
i t
i.--
4.0 EXPANSION ANCHOR ALLONABLES AND FACTORS OF SAFETY Expansion anchor ultimate loads for McGuire were established by direct re-duction of manufacturer's published ultimate loads by the ratio of 3000/3500, to account for the lowest nominal concrete strength specified for the plant site. Allowables were computed utilizing a factor of safety of four (4) for nomal and upset load conditions and two and one-eight (21/8) for factored load conditions. The load combinations of prime concern for the bulk of McGuire support / restraint expansion anchors are as follows:
D+L+.6SSE1[P Nomal/ Upset l
P l
Faulted D + L + SSE 1 h At this point, it is instructive to observe that application of these load com-o l
binations and stated factors of safety will result in the nomal/ upset load combination governing the expansion anchor design, in all cases. This can be shown mathematically by setting the faulted load factor of safety, FS, equal p
to a variable and equating the load combinations, i.e.
)[P Faul ted D + L + SSE < (
4.0 Therefore
FSp=
(Eq.1)
.4SSE j, D + L +.655E Equation 1 defines the faulted factor of safety above which the nomal/ upset load combination will govern and below which the faulted load combination governs.
FS will vary depending upon the total percentage of faulted load which is re-p l
presented by the term (D + L). The higher this percentage, the lower the change in nomal/ upset vs faulted load and therefore a higher specified FSp would be necessary to govern the design. A graph of this relationship is shown in At-tachment 6.
Review of Attachment 6 indicates that the minimum FSp found at McGuire is actually 2.4 as opposed to the 2.125 delineated in the design criteria. This minimum would occur only where the load is composed solely of earthquake forces such as a seismic snubber. The more typical.
I support / restraint would include several contributions to D + L in addition to seismic, thereby assuring a higher factor of safety present in the faulted load combination simply based on the factor of safety design criteria and load com-binations employed.
Actual strength of Nuclear Safety Related Category 1 concrete was established by test during construction and thoroughly documented.
Duke reviewed these concrete strength tests and randomly selected 50 samples for each designated mix used within each building, for statistical reduction. Utilizing acceptance l
criteria delineated in ACI 318-71 and statistical analysis methodology and sampling size established by ACI 214-65, the actual design concrete strength was computed at 90 days. Concrete ageing of two (2) years was similarly incor-porated using test results available in the literature and described by 2
Nilsen's analytical expression. The following actual concrete strengths re-sulted:
Building Mix F' (Psi)
AB A
4348 AB B
5886 AB C
7375 RB C
7210 l
1 2Nilson, A H, Desion of Prestressed Concrete, John Wiley and Sons, Inc,1978.
9 u
=
e I
5.0 BASIS FOR MCGUIRE FAULTED FACTOR OF SAFETY Duke Design Engineering Department conducted a study in 1974 to establish a uni-form criteria for design of expansion anchors at McGuire Nuclear Station. This study reviewed current industry practices and available test data on the perfor-mance of the particular anchors specified for McGuire Nuclear Station. The con-clusion of this study, as applied to piping supports / restraints, was that use of a factor of safety of four (4) for normal / upset loadings and two (21/8) for faulted loading was reasonable and conservative. Considering the margins al-ready introduced at various steps of the analysis process necessary to estab-lish governing anchor loads and all the experimentation, testing and study of anchors which has taken place since 1974, our conclusion remains unchanged.
The service load condition factor of safety of four (4) has gained wide accep-tance as a reliable design value where proper installation controls are exer-cised. Adjustment of service load condition allowables to obtain the faulted load condition allowable is also well established and accepted with near unan-imity amongst the industry, national code comittees and appropriate government regulatory agencies. Examples of this agreement can be seen in ACI 349-76,
" Code Requirements For Nuclear Safety Related Concrete Structures", USNRC Stan-dard Review Plans 3.8.3, 3.8.4, 3.9.3 and others, USNRC Regulatory Guide 1.142 which reviews Chapters 1 thru 19 of ACI 349-76 and supports this work with min-or exceptions, the " Proposed (Draft 1, September 12, 1978) Specification for the Design, Fabrication and Erection of Steel Safety Related Structures for Nuclear Facilities" by AISC Comittee and the McGuire FSAR. Duke is unaware of any technical basis either analytical, experimental, or based on experience to substantiate a deviation from this design philosophy as proposed in USNRC IE Bulletin 79-02, which requires an equal factor of safety for all load con-di tions. -
l 6.0 EFFECT OF TENSION ZONE INSTALLATION Some concern has been expressed within the regulatory agency regarding the ef-fect of tension zone concrete cracking on expansion anchor capability. Duke is aware of only limited experimental work in this regard. Some testing of this effect has been performed on the LIEBIG Safety Bolt by the Technical Uni-versity Darmstadt, West Germany. Although this anchor functions somewhat dif-ferently than typical McGuire wedge and sleeve anchors, limited infonnation j
may be gained from this testing. Some degradation of ultimate anchor load capability did occur but in no case would it have rendered a bolt incapable of handling its load if the McGuire design criteria had been utilized, i.e. any I
degradation which occurred fell well within the factor of safetys utilized at McGuire. Tension zone cracking of the magnitude of concern would be expected to occur in isolated and local regions, if at all, within a nuclear facility during the Safe Shutdown Earthquake. Several combinations of conditions must be present to structurally challenge an anchor located in one of these regions.
The anchor design load must have been seriously underestimated, structural re-view of the support restraint discussed in Section 7.0 overlooked and anchor load occurring in phase with crack opening of sufficient size to reduce the anchor ultimate capability. We believe the likelihood of such a combination to be remote. However, we underscore the need for the responsible support /
restraint designer to be aware of this possibility and exercise good engineer-ing practice in factoring this into his design. This is the basis for Duke's requirement of Structural Engineering review and approval as described in Sec-tion 7.0.
f l.
7.0 DESIGN AND CONSTRUCTION INSTALLATION PRECAUTIONS It is generally established engi aering practice to utilize safety factors as at allowance for uncertainty in analytical methods, materials, construction techniques, and/or field conditions. Some or all of these considerations may apply to a given engineering endeavor, and industry standards attempt to provide guidance for specific cases.
G11 dance for anchor bolts must cover a wide spec-trtsn of projects ranging fr m crinercial buildings to nuclear power plants.
The controls which govern design and construction practices for nuclear power plants are substantially more stringent than for other types of heavy con-struction. For this reason, much of the uncertainty which safety factors ac-count for does not exist in nuclear power plants. Similarly, much of the per-fonnance history which provides a major part of the knowledge upon which factors of safety are based was established in the commercial industry. At Duke Power Company nuclear facilities, the support / restraint designer's are prohibited by procedure from installing anchor bolts in the lower 1/3 of structural concrete beams (tension zone), unless no alternative exists.
In this case, the support /
restraint design must be independently reviewed by the Structural Section of the Civil-Environmental Division for potential degradation of anchor capacity and effect on the beam. This review is documented on the support sketch prior to release for construction, if installation is appmved.
Installation torques specified for the wedge and sleeve anchors at McGuire are sufficient to preload the anchor assembly to loads in excess of P /2.125 (except for 3/8" sleeve u
anchors). This provides verification that the anchor assemblage is capable of carrying loads in excess of design.
l Design control and quality control are extremely important aspects of Duke's work. During the entire history of McGuire construction, anchor bolt in-stallation has been closely monitored in the field by Construction Technical Support personnel, Quality Control personnel, and Design Engineering personnel.
Item 4 of the Duke response to USNRC IE Bulletin 79-02 provides a brief history I
l of the controls used to document proper installation of expansion anchors.
In addition to the procedures identified above, Design Engineering initiated a reinspection program in the spring of 1979 to assure that construction tol-erances were being correctly applied by construction and quality control l -
personnel. This reinspection program covered the support as a whole and not just the anchor bolts. Such extensive controls and the verificatior of ccmpleted designs is almost without precedence. The net result is a much higher level of confidence that the support can perform its intended safety function under postulated loading conditions.
- L, e
r I
i l
l l., _ _ _~., _ _.. _
8.0 CASE-BY-CASE REVIEW OF SUPPORTS / RESTRAINTS McGuire Support / Restraint Design Group personnel conducted several reviews of supports with anchor bolts to verify compliance with IE Bulletin 79-02. Prior to receipt of the Bulletin, base plate flexibility was not considered in the design. Supports at McGuire are generally classed as rigorous or alternate depending on the type of piping analysis employed.
~
Alternate analysis supports comprise approximately 6,513 of 15,276 safety re-l lated supports in Unit 1.
A screening of these alternate supports identified 549 which potentially would not have the Bulletin design safety factor in the normal / upset load combination if base plate flexibility were considered. A detailed review of each of the support / restraints was conducted accounting for base plate flexibility. Modifications were specified as required to upgrade the support to meet this design safety factor criteria. This work has been completed.
Rigorous analysis supports comprise approximately 8,763 of 15,276 safety re-lated supports in Unit 1.
These were also screened to account for base plate flexibility considerations, and 1,952 supports potentially did not have the required design safety factor. A detailed review of each of these support /
restraints was conducted accounting for base plate flexibility. Modifications were specified as required to upgrade the support / restraint to meet this design safety factor criteria. This work has been completed.
In response to concerns expressed by U S Nuclear Regulatory Comission l
Region II inspectors regarding the McGuire design criteria of FSp = 2.125, McGuire Support / Restraint Design Group reviewed a sample of 1,410 rigorous supports to estimate the overall impact of upgrading Unit I to a faulted safety factor of 4, holding all else constant. The importance of concrete strength was examined by reviewing a 3000 psi nominal strength case and a 5000 psi nominal strength case. presents the projections derived from this analysis. Note that 604 rigorous and 98 alternate supports, in Unit 1, were identified as having a strong potential for not complying with a faulted condition safety factor of 4 (5000 psi concrete). A detailed review of standard design practices at -.
McGuire identified some very significant conservatisms which were not required, and which if removed would likely result in safety factors very close to satis-fying a minimum requirement of 4 for the faulted loading condition (see Sections 3.0and4.0).
To verify this, a sample of 65 rigorous supports was randomly selected from those which failed both the aforementioned screening cases (59 failed both cases and 6 failed only 1). The McGuire Support / Restraint Design Group re-computed the safety factor with conservatisms such as those described in Sections 3.0 and 4.0 removed, if applicable, while all else was held constant.
The reanalysis did not attempt to remove any margins associated with the piping or seismic analysis of the building. contains three graphs which stannarize the results of this analysis.
The first graph is a histogram showing the current safety factor frequency dis-tribution within selected intervals. Note that most are clustered in the 2.4 to 4.0 safety factor range. The second graph is a similar histogram for re-calculated safety factors. As expected, the distribution has shifted to cluster around the 3.6 to 5.0 range. The third graph is a cumulative distribution curve for the reanalyzed safety factors. Note that in the 3.80 - 3.99 range, the cumulative distribution is 18.5%; or stated another way, approximately 81.5%
f of the reanalyzed safety factors are 4.0 or above. Considering the sample was taken from the group which had failed the screening process, it is reasonable to conclude that very few of the total expansion anchor population at McGuire would fail to cogly with a FSp = 4.0, if an elaborate and intensive analysis review were conducted for each. In fact, if this finding were used to extra-(
polate the impact on the total population of support / restraints in McGuire Unit 1, it would represent less than 1%.
9.0 IMPACT OF FACTOR OF SAFETY RETROFIT The prccess and impact of attempting to backfit a safety factor of 4.0 for faulted condition loadings have major proportions relative to Design, Con-struction and Plant Operation. Many of the individual support / restraints that will require Design and Construction rework to meet FSp = 4.0 are large and structurally complex. The base plates contain a large number of expansion anchors and most have already been significantly revised from original designs to meec conservative base plate flexibility criteria and revised loadings.
In-sta11ation of these support / restraints required many iterations between Design and Construction over a period of several years to achieve Design-acceptable, erected and inspected configurations. As will be discussed in later paragraphs, 1
i it is probable that further Design-acceptable modification of these support /
restraints, to meet FSp = 4.0, will not be possible in many cases and will re-quire pipe routing changes. This consequence will reopen other major areas of plant qualification and preclude plant operation during modification.
A backfit program will directly reopen, at a minimum, the following three Design / Construction activities:
Individual support / restraint design, erection, and inspection Pipe stress math model review for acceptable st:oport/ restraint location i
and load capability Construction QC final walkdown of completed math models j
Other activities with a high probability of being reopened or significantly impacted are as follows:
Piping stress analysis Equipment qualification analysis (nozzles)
Piping layout design Civil structure analysis Plant operation during backfit program l
Qualification and installation of miscellaneous components which require physical change to permit backfit modifications to pipe support / restraints, pipe routing or both to meet FSp = 4.0.
In subsequent discussion of the three activities directly affected, it will be l
stated or apparent how these other activities can be drawn into a backfit program.
I I
9.1 BACKFIT PROCESS / IMPACT FOR INDIVIDUAL SUPPORT / RESTRAINTS DESIGN, ERECTION, AND INSPECTION The process of backfitting FSp = 4.0 into individual support / restraints would consist of activities shown on Attachment 9.
These activities can be concisely stated as follows:
Screening to identify affected supports / restraints
- Engineering evaluation to identify necessary redesign support / restraints Redesign Construction rework Construction QC reinspection 9.1.1 Screening Process l
A screening process will be required to identify all of the safety related support / restraints which utilize expansion anchors that do not meet FSp = 4.0 under design loadings. The process of identifying safety related support /
restraints which utilize expansion anchors is largely clerical in nature. The l
process of determining if FSp = 4.0 is met will require engineering calculations in almost every case because normal / upset conditions govern design for essentially is not available in existing docu-all support / restraints and therefore FSp menta tion. A calculation of FSp would be required for well over 90-percent of the support / restraints. The method for calculation would be the same as used to qualify the expansion anch' ors initially (hand or computer). All support /
restraints which do not meet FSp = 4.0 by this check would be identified.
I Results of the screening process are shown as output Item 3 on Attachment 9.
9.1. 2 Engineering Evaluation The support / restraints identified as not meeting FSp = 4.0 from the screening process would be evaluated analytically using the most sophisticated analysis techniques available (STRUDL, ANSYS computerized analysis) and removing all obvious conservatisms in the original analysis, such as those described in.
Section 3.4 This process will consume substantial stress analysis manhours and computer resources for a period of months. Support / restraints which still t
l do not meet FSp = 4.0 after this evaluation will require redesign and addi-tional calculations (output Item 4 on Attachment 9),
9.1. 3 Redesign The redesign process will reopen the cycle between Design and Construction since any revision to a design will require, at a minimum, a reinspection.
For these particular support / restraints, craft rework to some degree will be re-quired also. As noted earlier, support / restraints falling into this redesign i
scope will include many of the largest and most structurally complex in the i cerations to plant, which have already required many Design and Construction achieve acceptable installations.
Redesign would encompass the following pro-bable methods which are listcJ in order of ascending Complexity:
(1) Use of larger anchors Where it is feasible and will achieve FSp = 4.0, larger anchors would re-place existing anchors. This would increase the FSp with minimum new drilling in base plates and concrete.
Due to rebar density and undesirable or impossible conditions to cut rebar, any new location drilling would be avoided. Use of larger anchors will be a problem in most cases due to violation of edge distance or bolt spacing on plates and concrete.
l l
(2) Use of larger number of anchors of same (or larger) size For support / restraints where existing anchor location or space permit, ad-ditional anchors of the same or larger size would be designed for attach-i ment. The two options available are to add (1) additional anchors to the existing plate or (2) wing plate (s) to the existing plate to acconnodate more anchors. Adding anchors to the existing plate would be feasible for only the simplest support / restraints configurations which are typically lightly loaded. Generally, the heavily loaded support / restraints already utilize all available space on the existing plate and frequently wing plates have already been added to some of these. Addition of wing plates to acconnodate more anchors will reopen base plate flexibility considerations which, from experience, has led to significant redesign (gussets, etc.).
l Obviously, this option is less desirable than simple addition of more i
l.
i expansion anchors and is more appropriately defined as a total redesign.
(3) Total redesign - same location Where anchor changeout or additi$n will not provide FSp = 4.0, redesign at the same location would be attempted. A change in location would im-pact piping analysis among other areas as noted in (4) below. Redesign at the same location is significantly affected by existing, related or unre-l lated items such as piping, other support / restraints, equipment, instru-mentation, etc. The ability to go bacx and reinstall the same support /
restraint at that location is impossible in many %nstances due to items erected later which wou,1d interfere with craft work and/or inspection.
Although this is a Construction impact,. such situations are reviewed be-fore redesign and factored into design. For the more structurally complex support / restraints, maximum utilization of existing space has already been made.
(4) Total redesign - different location Redesign at a different location.is the least preferable option and would cause the most significant overall impact on Design activities. Available space for additions to the buildings is at a minimum due to congestion, a i
large amount of which is created by support / restraints. Location of an available space itself constitutes a major problem. Beyond this, the following impacts are certain:
Review of f@act of new location on existing piping stress analysis -
a significant amount of reanalysis has already been identified and l
completed due to inability to install support / restraints at specified locations (reopens consideration of all support / restraints in the af-fected math model).
Drilling at a new location reopens the iterative process of Design and 1
Construction to achieve an acceptable installation without rebar or l
other interference pmblems.
Total rewsk of Design calculations on the support / restraints moved.
The impact of moving a support / restraint to a different location is clearly a major concern. The possible impact on piping stress analysis has rami-fications well beyond simple review, as is noted in later discussion. As -
can be seen from the above probable courses of action, the redesign process cannot be considered lightly in a backfit program. Considerations noted above apply principally to the Design process but are in every case tied directly to erection.
It is impossible to grasp the significance of re-design without considering erection problems in parallel.
9.1. 4 Construction Rework I
Construction rework resulting from redesign would represent a major block c' the work, time, and cost required to backfit to FSp = 4.0.
The following are examples of impacts / problems that are certain to occur:
(1) Rework at existing locations would be impossible in most cases without removal of part of the existing support / restraints or items unrelated to the support / restraints. This can create major rework on items totally un-related to FSp and opens the possibility for damage to these items.
(2) Rework at existing locations would require, in many cases, special equip-ment to reach inaccessible areas. Currently there are over 300 support /
restraints which are known to be inaccessible for reinspection under the l
open.50.55(e) related to support / restraints in Unit 1.
(3) Removal of existing anchors is expected to result in concrete damage at some locations which would preclude installation of other anchors at that same location. The possibility of compromising the integrity of the con-l crete structure at that location is a concern. At a minimum, Design evaluation of damage would be required (reopen structure qualification analysis).
(4) Any new drilling for installation of new anchors has high probability of rebar interference which has historically resulted in many Design and Construction iterations to produce acceptable installations.
(5) pooval of existing support / restraints where the redesigned support /
restraint had to be moved elsewhere is expected to require, in most cases, removal of unrelated items. This would occur because the most prevalent reason for being unable to modify the existing installed support / restraint, will be interference preventing further craft work. These same inter-ferences will impact removal of the support / restraint.
The above impacts cannot be quantified, scheduled, manned, or costed readily because of the indeterminant nature of the number of Design and Construction interactions required. This has been verified by actual experience at McGuire.
Both the Design and Construction organizations are placed in responding modes to the situation and utilization of personnel and other resources is typically grossly inefficient and difficult to estimate in time and cost.
l 9.1. 5 Reinspection Reinspection of reworked support / restraints would be conducted under the same program as currently in place. As in the case of Design and Construction craft work, this activity will be in a responding mode to craft completion of instal-lation which would be a drawn out process.
9.2 BACKFIT PROCESS / IMPACT FOR PIPE STRESS MATH MODEL REVIEW As a result of the 50.55(e) filed on Unit 1 support / restraints, Design and Construction reinspected safety related support / restraint locations in McGuire Unit 1.
To ensure compliance with IEB 79-14, all support / restraint locations (and pipe routing) have been compared to latest pipe stress analysis math models. Discrepancies, if any, are evaluated nd, if necessary, support /
restraints have been moved or reanalysis performed, or both. At present this work has been conpleted and only a small amount of followup stress analysis re-mains (is in progress).
Support / restraint revisions from this effort are ex-pected to be minor.
If, in a backfit program, support / restraints have to be moved, the math model effort would have to be reopened and the possibility of reanalysis exists. Because of congestion in the plant it is likely that this effort would extend to support / restraints and piping unrelated to the FSp back-fi t. 0 shows the overall process involved, beginning with math model review and extending through the activities involved.
Math model review is initiated after Construction is complete or verified for all piping and support / restraints in a math model.
If the review indicates that all support / restraints and piping are located within tolerance of analyzed.
locations, or, if not, are acceptable by engineering judgement, there is no impact and documentation is provided for piping analysis closecut and N-stamp signoff.
If there is impact, then typically an attempt is made to relocate the support / restraint to the required location (piping deviations are typically al-ways basis for reanalysis). Since the only support / restraints that would be mislocated in this review are those that were relocated from analysis locations intentionally (could not achieve FSp = 4.0 at required location), it is clear that relocation to the analyzed location is not an opticn, therefore, it is not shown on Attachment 10.
Revised piping analysis may require other support / restraints to be relocated in order to qualify the piping for the one or more support / restraints, deliberately moved. After analysis, all support / restraints in the math model would be re-viewed for acceptable location.
If all are acceptable, individual support /
restraint review and/or redesign would be perfonned (see latter part of At-tachment9).
If one or more locations is unacceptable, new locations must be identified or piping changed until the analysis is acceptable using installable support / restraints. For a backfit program, this path could be divergent, i.e.,
for many large support / restraints, there is no Design and Construction ac-ceptable alternative location that would be acceptable to piping analysis.
Once a pipe reroute is mandated to achieve acceptable support / restraint location geometry, the following activities in addition to piping analysis are reopened:
piping design Equipment qualification Concrete / steel structure qualification Installation and qualification of miscellaneous impacted items.
The ensuing activities result in a repeat of a majority of the initial Design' activities for the affected section of piping subject to congestion already in the plant.
In the case of the large, complex support / restraints, backfit of one support / restraint could have this cascading impact.
Because there is no way to determine the total process required for a given support / restraint backfit, the resources and time frame cannot be estimated reliably. The above process, if required in total, would extent for several months for each affected piping run.
9.3 BACKFIT PROCESS / IMPACT FOR OC FINAL WALKDOWN Construction OC final walkdown is generally performed in parallel with math model review.
The flow path for this process is shown on Attachment II. Un-acceptable results indicates that discrepancies are identified requiring Con-struction or Design review. Typically, discrepancies are contact with adjacent piping, support / restraints, or other items. On occasion, these discrepancies can only be resolved by revision to support / restraint designs, which would re-vert back to the redesign / rework process described in Section 9.1.
Final walkdown is generally completed within the envelop of time required for math model review and would not constitute a critical path function is no re-design was required as 'a result.
l _
10.0 CONCLUSION
S Section 3.0 has been presented to provide a brief summary of some of the major design margins intmduced at various steps of the analysis problem, upstream of expansion anchor load outputs.
It is not intended to be all inclusive and it is recognized that the prudent analyst could significantly augment this sumary.
In setting the McGuire design criteria, it was considered essential to closely monitor accumulated design margin in the solution process to provide a basis for setting downstream margins. We conclude that the anchor design load, P, has been substantially overestimated.
Section 4.0 demonstrates that the normal / upset load combination generally would show values governed expansion anchors on McGuire and calculation of FSp of 2.4 minimum with the preponderance of values well above 2.4.
Analysis of actual concrete strengths demonstrates that mix proportioning was conservatively designed, as is comon for r,3 clear class concrete, and strengths exceed nominal requirements.
Section 5.0 outlines the process by which Duke established its expansion anchor factor of safety criteria in 1974. We believe the design philosophy established
(
at that time incorporates good engineering practice and is soundly based.
Section 6.0 concludes that tension zone installation of expansion anchors could degrade an isolated anchor, under extreme conditions, however not to an extent sufficient to jeopardize the anchor when designed to the McGuire criteria. Duke supplementally reviews installations of this type as a prudent measure during the design process. This condition is no more prevalent or severe than the random presence of an installation deficiency for which a proportion of the factor of safety is allocated.
Section 7.0 outlines the extensive inspection program for expansion anchors and support / restraints at McGuire Nuclear Station. As recently as 1979, a complete reinspection was initiated to provide final confimation of the adequacy of these installations. The need for that portion of the factor of safety at-tributable to field conditions was undoubtably been diminished by this intensive inspection effort.
l
! l I
Section 8.0 provides the results of a sample reanalysis of support / restraint with more rigorous attention to providing a less conservative estimate of the anchor load by removing some of the simplifying assumptions present in the McGuire support / restraint standard analysis approach. This reanalysis gave no consi-deration to margins remaining in the analysis problem upstream of the support /
restraint loads. Based on the results of this work, we conclude that it is likely that nearly all plant expansion anchors would actually possess FS 's in p
excess of 4 if subjected to an intensive effort to estimate the actual anchor loads more accurately.
i l
Section 9.0 outlines the impact of retrofitting a fac. tor of safety change at this time on McGuire.
Unfo,rtunately, because of the design time and magnitude of effort involved in an extensive reanalysis effort, such as conducted in Section 8.0 and incorporating applicable items in Sections 3.0 and 4.0, would not be possible or attempted at this stage of McGuire's construction program.
A conservative screening would have to be implemented with resulting retrofit.
Impact on the plant schedule and plant physical facility would be extensive.
USNRC IE Bulletin 79-02 requires that justification be provided by the licensee when the minimum Bulletin factors of safety cannot be fully verified for a plant site. We believe this review of the McGuire expansion anchor factors l
of safety provides justification that full and reliable long term service can be expected from these insta lations in both the service and faulted load con-di tions.
Duke does not believe the Bulletin FSp and McGuire's design criteria are substantially at variance for McGuire Nuclear Station, however they are based on dissimilar design philosophy.
l l
l 1,...
11.0 SYSTEM BY SYSTEM ANALYSIS 12-01-80 This section has been added to present information developed after the meeting with USNRC on August 8, 1980.
The conclusions of Section 10.0 remain unchanged, and this additional information simply substantiates the contention of Section 8.0 that it is likely that nearly all expansion anchors actually possess a safety factor of 4 or greater.
11.1
SUMMARY
OF AUGUST 8, 1980 MEETING Duke Power Cosapany representatives presented the first ten sections of this report to members of the USNRC on August 8, ic80 in the Atlanta offices of Region II. Those in attendance concluded that Duke Power Company has a strong case concerning the quality of anchor 1,olt design and installation at McGuire helear Station.
So that conclusions could be drawn on a system basis, the NRC requested that Duke Power Company analyze the safety factor issue for each nuclear safety related/sd smic system.
The NRC stated that IE Bulletin 79-02 requirements would be fully resolved if Duke Power Company could demonstrate that there is a 95% confidence level that no more than 5% of the anchor bolts, on each nuclear safety related system, have a safety factor less than 4.
11.2 STATISTICAL APPROACH - BINOMIAL DISTRIBUTION The object of our statistical approach is to design a sample on a System-by-system basis to verify with 95% confidence that less than 5% of all anchor bolts in the system do not meet the specified minimum safety factor of 4.
It 26
is assumed that the event to be detected (an anchor bolt not meeting the specifiad minimum safety factor - a defect) follows a binomial distribution.
This event is expected to occur rarely (less than 10% of the time).
Therefore, normal sampling procedures would require prohibitively large sample sizes; whereas inverse sampling techniques offer some efficiencies.
The method used involves randomly selecting anchor bolts until:
- 1) a predeter-mined sample size has been reached without an anchor bolt with a FS<4 being detected, or 2) one is detected.
If the predetermined sample size is reached j
without an anchor bolt with a FS<4 being detected, the incidence of occurence is significantly less than 5%.
However, if one is detected, further sampling is necessary.
A revised sample size is determined and sampling continues until this new sample size is reached (without another anchor bolt with a FS<4 being identified), or until a second FS<4 is detected.
If another FS<4 is found, the process is repeated until further sampling and analysis are impracti-cal. 2 gives the minimum sample size required for a given number of anchor bolts with a FS<4.
11.3~
SAMPLING METHOD Duke compiled a considerable volume of information in preparation for the August 8 meeting, and this was used as a starting point for extending the analysis.
This initial data was taken from a sample of 1410 supports, as discussed in the August 8 meeting.
Duke selected additional supports using a table of random numbers whan it was necessary to extend the sample size.
27
The minimum sample size required in using a binomial distribution assumption is 74.
Large sample sizes are permitted and have correspondingly larger allowable number of anchor bolts whith FS<4 to satisfy the hypothesis.
These values are readily obtainable from a table of confidence limits for propor-tions applicable to Binomial, Poisson, and Hypergeometric Distributions.
The Binomial distribution function and inverse sampling techniques assume a large population (N) compared with the sample (n).
Some systems have fewer anchor bolts than the minimum sample size and will be analyzed absolutely, i.e., the total population (N) analyzed.
11.4 RESULTS The results for the 29 Nuclear Safety Related systems are presented in Attach-ment 13.
All rigorously analyzed systems passed the acceptance criteria.
Alternate Analysis systems which did not meet the criteria are NF (ice conden-ser refrigeration), VE (annulus ventilation), VG (diesel generator starting air), WS (solid waste), WZ (ground water drainage), and YM (demiaeralized water).
The small number of systems which did not meet the criteria and the fact that time and cost trade-offs did not permit analytical credit for the full range of conservatisms existing in the original design demonstrate clearly that anchor bolt safety factors are generally far in excess of that required by the plant design criteria employed for McGuire Nuclear Station.
Many of the anchor bolts in the sample which did not have a minimum safety factor of 4 had calculated safety factors of 3.5 or greater. As demonstrated by Attachment 6, there are no anchor bolts with a safety factor less than 2.4.
j 28
f l
1 11.5 IMPACT ON FUEL LOADING AND FULL POWER OPERATION Four of the six systems identified in Section 11.4 as not meeting the safety l
factor criteria are non-essential systems for fuel loading and low power l
operation.
A detailed discussion of these systems is covered by the McGuire FSAR, which has been reviewed by the NRC.
Even in the unlikely event of an l
earthquake after fuel loading but prior co Full Power Operation, there is no unacceptable increase in risk to members of the public while four of the 1
l systems are being upgraded. Only two of the six systems are part of the plant Enginnered Safety Features designed to mitigate the potential for an accidental release of radioactive r.aterial or to permit safe shutdown of the reactor.
These are VE (annulus ventilation) and VG (diesel generator starting air);
both of which will be upgraded to meet the acceptance criteria prior to fuel l
loading.
Also, because of limited accessibility after fuel loading, all anchor bolts inside containment which were identified in the sampling program or during the corrective action phase as having a safety factor less than 4 will be upgraded to a minimum FS of 4 prior to fuel loading. The remaining l
four systems (identified in Section 11.4) will be upgraded to meet the accep-tance criteria prior to receipt of a full power operating license.
In addi-tion, all anchor bolts outside containment which were identified in the samp-ling program or during the corrective action phase as having a safety factor l
less than 4 will be upgraded to a minimum FS of 4 prior to the end of the l
first refueling outage.
l l
This schedule committment for upgrading goes beyond the requriements of IE Bulletin 79-02, which permits interim operation at full power as long as no anchor bolts have a safety factor less than 2.
The committment to repair 29
anchor bolts even though the statistical acceptance criteria was met goes well beyond the requirements agreed to in our August 8, 1980 meeting.
- However, Duke believes this does comply with supplemental USNRC requests in this regard.
11.6 PLAN OF CORRECTIVE ACTION i
All anchor bolts on each of the systems which did not meet the criteria will be screened, analyzed in detail, and repaired as necessary to achieve a condi-tion which meets the acceptance criteria, i.e. no more than 5% of the anchor bolts on that system with a safety factor less than 4.
This program will be similar to the hanger baseplate review which was done to consider baseplate flexibility. Also, we will expeditiously correct all bolts identified as having a safety factor less than 4.
l The schedule for our phased approach to corrective action is stated in detail in Section 11.5.
Although the last step in upgrading all bolts in the sampling program to a minimum FS of 4 is scheduled for the end of the first refueling, Duke expects that this work will be completed much sooner.
Duke will advise NRC inspectors of the progress of repairs during site inspection visits and will submit a written confirmation when all repairs are complete. A summary l
of the time iranc for upgrading to meet the 95% confidence level acceptance criteria and to correct all bolts identified as having a safety factor less than 4 is presented below. Where possible, Duke plans to schedule work on those with the lowest safety factors first.
l 30
-. ~.
ACTIVITY TIME FRAME Upgrade to Meet 95% C. L. - System VE & VG Prior to Fuel Loading Correct Bolts Identified w/ FS<4 - All Inside Prior to Fuel Loading Contair. ment Upgrade to Meet 95% C. L. - System NF, WS,
Prior to Full Power License
,WZ & YM Correct Bolts w/ FS<4 - All Outside Containment Prior to End of 1st Refueling 31
ATTACHMENT 1
y IE Bulletin No. 79-02 March S, 1979 should be considered ficxibic if the unstif fened distance betvcen the member vcided to the plate and the cdae of the hace piste in creater than twice the thickness of the plate.
If the base piece i= determined to be facxible, then recalculate the bolt loads using an appropriate analysis which vill eccount for the effects of chear - tencion interaction, minieu t edge distance and proper bolt spacing.
This is to.be donc prior to testint of enchor bolts. These calculated bolt loads are referred to he enf ter as the bolt design loadc.
2.
Verify that the concrete c7pensims anchor bolts have the folleving nininuta fcctor of ecfety betvcen the bolt design lond and the bolt ultire.tc capacity determined front ctatic load tests (e.g. anchor bolt T..nufacturer's) whic:t simulate the actual corAitions of
~
inctallation (i.e., type of concrete and its strength proporcies):
c.
Four - For wedge and sleeve type anchor bolts, b.
Five - For chcIl type anchor belts..
l 3.
Deacribe the design requirements if applicabic for anchor belt to t.'ithetend cyclic leeds (e.g. sels: tic 1 cads and high cycle operatin3 loeds).
4.
Verify from existing QC doctn.cntation that design requirenents have been met for coch anchor bolt in the following areas:
(a)
Cyclic leads beve been considered (e.g. enchor bolt preleed is equel to or greater than bolt design load).
In the cice of the shell tfpe, ccoure that it ic ' tot in contact with the beck of the support pinte prior to preload testin;;.
l (b)
Specified design size and type is correctly installed (e.g. procer l
c bedcent depth).
If cufficient docurent: tion does not exist, theit initiate a tectiva procrnm that trill casure that einimum design requircrents hcvc been ret with rcepect to sub-ite=s (a) and (b) chove. A sempling technic.uc to cceepteblo. One neceptchic techt'ique is to randonly celcet and test ene anchor belt in cach besc plate (i.e. some supports c: y have core than ene base plate). The test chould provide verification of sub-itens (c) cud (b) above. If the rest failn, all other bolts on thct bacc pinto should be similarly tested. In any event, the test progra t chould accure th t each Scicmic Category 1 system will perform its intended function.
J 2 of 3 ATTACHMENT 2 IE Bulletin No. 79-02 November 8, 1979 Revision 2 Pase 3 of 7
'~S V
It has been noted that the schedule for analytical work on base plate R1 flexibility for some facilities extends beyond the Bulletin reporting time R1 frame of July 6, 1979.
For those facilities for which an anchor bolt R1 testing program is required (i.e., sufficient QC documentation does not R1 exist), the anchor bolt testing program should not be delayed.
R1 2.
Verify that the concrete expansion anchor bolts have the following minimum factor of safety beLWeen the bolt design load and the bolt ultimate capa-city determined from ctatic load tests (e.g. anchor bolt manufacturer's) which simulate the actual conditons of installation (i.e., type of con-crete and its strength properties):
Four - For wedge and sleeve type anchor bolts, a.
j b.
Five - For siell type anchor bolts.
1 The bolt ultimate capacity should account for the effects of shear-tension R1 interaction, minin um edge distance and proper bolt spacing.
R1 If the minimum fat tor of safety of four for wedge type anchor bolts and R1 1
five for shell tyl e anchors can not be shown, then justification must be R1 provided. The Bulletin factors of safety were intended for the maximum R2 support load incitding the SSE. The NRC has not yet been provided adequate R2 justification that lower factors of safety are acceptable on a long term R2 basis.
Lower factors of safety are allowed on an interim basis by the R2 provisions of Sup;lement Mc. I to IE Bulletin No. 79-02. The use of R2 N
reduced factors of safe'
'te factored load approach of ACI 349-76 has R2 not yet been accested L
..H C.
R2 Describe the desi a requirements if applicable for anchor bolts to with-3.
t stand cyclic load (e.g. seismic loads and high cycle operating loads).
4.
Verify f rom existing QC documentation that design requirements have been
. met for each ancht r bolt in the following areas:
Cyclic loads have been considered (e.g. anchor bolt preload is equal
)
(a) to or greater than bolt design load).
In the case of the shell type, i
assure that it is not in contact with the back of the support plate prior to preload testing.
j l
(b), Specified design size and type is correctly installed (e.g. proper embedment depth).
~
If suf ficient documentation does not exist, then initiate a testing program that will assure that minimum design requirements have been met with respect to sub-items (a) and (b) above. A sampling technique is acceptable. One acceptable technique is to randomly select and test one anchor bolt in each base plate (i.e. some supports may have more than one base plate). The test should provide verification of sub~ items (a) and (b) above.
If the test fails, all other bolts on that base plate should be similarly tested.
In any event, the test program should assure that each Seismic Category I v
system will perform its intended function.
ATTACHMENT 3 30,-
7 v
//'>(
b$45AkMdA / 5% kWW r' fx?</xX'>XS888 MCK <'xSt>ild
$u [whk>O
['3@) O3 x / / x / x ^ nns s Nu x
/we v m xx w i x
ev/ xv ^ ^wm xN
>wux xx 8 s YsiNN X'N /X
/N LXfNYNW
^'#/Nv'x' XN ANX% '9 X/N /N A N/N'N>CiN///VM/ P4 h I'sX'h'KMN X A X vMsxv/'>^x'i X: A l
i X'49 W s# /M'xf N % ( M >
4
? %8I>'d2%$$h^>{')8%%W *xx 3
l1
%x/X XNxNsy yN x//h'X X'9sN'sX A W'>(Y MM
'@/d(%Mb2 i
,i"fx0W M'h?8.s66Cx /'s>M'Mdr y
_/ N
/N
~
x
=
'N
's
( '.,
~
y b
bb
)[
M
/[
N
^.
I
.9
>JN x ; / W/ N //x / x (A x NA AN Pl V w//A/ N/x _
\\
- ' N ^ ^ W W I' e
N'AT(y#
X/ /v' / X. VN K ixWQg' \\. X/V'x' V X VN _
"8
/
s
.6
-N'^' N'K6'<//^x'l X
^s X 'iDM%9 FX 7Ni /N
! Nkh s/
%k N/
o g@ IGD &'%d *xOx(#wx%@l)O
~66
/
l 3 'y Ex'y* /x ~
I x w x w
/x xYb (YY ff l
g' ',/
l
/ )
/N
?x
/ I LAW ENGINEERING TESTING COMPANY.
yh MIR{9N C A
S W !ON
/ /
o' o,
i, s
Ad l
.i
/
.. s v
!^!.A i,m x>
j 005 0.07 0.1 0.2 0.3
_3.0. 5 0.7 1.0 2.0 3.0 Undamped Period (Sec.)
ATTACHMENT 4 MCGUIRE NUCLEAR STATION UNITS 1 AND 2 GROUND RESPONSE SPECTRUM VS SYNTHETIC IIME HISTORY SPECTRA
REFERENCE:
MCGUIRE FSAR, FIGURE 2E-2D PEAK GROUND MOTION =.08s, e-5%
FSAR SYN HETIC I-H "fk$"
[ks$
kksI^
kk}"
i
.3
.034
.058 70.6
.5
.05
.08 60.0 1.0
.10
.16 60.0 2.0
.20
.29 45.0 3.3
.20
.26 30.0 5.0
.20
.27 35.0 10.0
.16
.18 12.5 20.0
.11
.11 0.0 l
l l
I l
l
n Iyh.Zt n l
J cwErEwGooG%E'_.od
- G "z g i oo!
o "'
o.e o
t'o
- 9o
- oo o $"
b o
05 u
/
W I
?
'j
!//,
1;i l
'8 i;;
0 l
Il 0
C J
Il 1
'g Il I l,
1r A
L
~
C l
t
\\
U
\\
\\
Ii L
\\
0 T
' \\
A 2
E
\\
0 D
\\
AD y
A-J S
U P
S E
T C
b-E P
T D
D E
R R
U v O
.0 M
N I
S 5
P D 0 N
E v%
s-C I
N T
v\\
F.
S mN U
E C
'\\s M
O 1
O Ys\\-
N 0 0
D S
s
\\2 s-Il i,
i i
2 i
0 c
0 g
%=
5 3
0 0
1 0
0 0
m0a. 51" 4Uw$gpE SMrwE-o5a5e n o4 e5bou 5ewm5' z8,4 8 $* o5 S p 5 '5>ap 5 O
n5.N
$o5mwZ3oage 5Q9z I8Le M.s 4M
.Ae
ATTACHMENT 6
GRAPH OF FAULTED FACTORS OF
- SAFETY, FS,
ABOvE g
WHICH FA_ ULT. ED LOAD COMBINATION WILL GOVFRN t
i k Faulted Load Combina-
,McGuire Normal / Upset Lo 4 l Combination, FS bas d l
l tion Governs if FSp ton original desYhN, crit 1hnaa4 3.
in this Range 40 s
a s
a s
l
~
3.8 l
l l
4.0 ASYMP. l i
u.,
34 l
\\
1 3.2
- o. 3 3.0
/1 Nomal/ Upset Load Com-
- bination with FSgjn =in 2.8 i
y 4.0 Governs with FS NcGuire Faulted Load Con-
/
p j
2.6
- i.his range.
j bination, FSp, Based on
/
Orioinal Design Criteria.
I m
~
2.4 2.2_ _
m_
2.0 l
l
____l_
.51.01.52.07.53.03.54.04.55.05.56.06.57.07.58.0 (D+L)/SSE i
NOTES:
- 1. Nomal/ Upset Load Co%ination PL +.6SSE <. [P l
- 2. Faulted Load Cortination D+L + SSE < (pf )
1 4.0 F
p>L1+P(f6SSE Faulted Governs Design
- 3. FS 4.0 j
- 4. FSF*
I + Pb SSE Nomal/ Upset Governs Design p
m.-
u Dead Load Contribution to Expansion Anchor Load l
- 5. D
=
I Operatino Load Contribution to Expansion Anchor Load
- 6. L
=
.L L_
Safe Shutdown Earthquake Contribution to Expansion Anchor Load u
- 7. SSE
=
l ATTACHMENT 7
MCGUIRE NUCLEAR STATION UNIT 1 EXPANSION ANCHOR FACTOR OF SAFETY REVIEW AUGUST 8, 1980 NSR SUPPORT / RESTRAINT
SUMMARY
Y Cl
- 'Sb' BUILDING RIGOROUS S/R
)
ALTERNATE S/R AB 5477 4989 10466 RB 3286 1524 4810 TOTALS 8763 6513 15276 f
NSR SUPPORT / RESTRAINT SAFETY FACTOR REVIEW (21/8 vS 4)
FAULTED LOAD COMBINATION l
S REEN F=4 S REEN W/ F = 4
$' =
' = 500 PSI r
PSI C
C SUPPOSTS/ ESTRAINTS BUILDING W/tXP NCHORS PASS FAIL PASS FAIL RIGOROUS (1,2) 3680 2926 754 3259 421 ALTERNATE (3) 1614 1453 (5) 161 1524 (4) 90 RIGOROUS (1,2) 2208 1750 458 2025 183 l
ALTERNATE (3) 793 774 (5) 19 785 (4) 8 RIGOROUS 5888 4676 1212 5284 604 l
Sun-TOTALS ALTERNATE 2407 2227 180 2309 98 TOTALS 8295 6903 1392 7593 702
ATTACHMENT 7
CONTINUED 4
MCGUIRE NUCLEAR STATION UNIT 1 EXPANSION ANCHOR FACTOR OF SAFETY REVIEW AUGUST 8, 1980 (1)
BUILDING BREAKDOWN BASED ON 3-5 RELATIONSHIP OF RB/AB RIGOROUS S/R'S.
(2)
RIGOROUS PROJECTIONS BASED ON 5888 TOTAL RIGOROUS S/R'S WITH EXPANSION ANCHORS, AND FAILURE RATES BASED ON A STUDY OF 1410 SUPPORTS.
(3)
ACTUAL FOR TOTAL ALTERNATE ANALYSIS S/R'S, AS OF SUMMER 1979.
(4)
PROJECTION BASED ON RIGOROUS S/R FAILURE RATE.
(5)
BASED ON ALTERNATE ANALYSIS SCREENING RESULTS.
1 l
9 1
I - _ -.
CL N Star m W1 ATTACHUENT g HsTOGRAMs E Meeuw ETPAN*. TON ANCHOR FAULTED FACTDAS li g h rrTV FcR Luptr g SupaceTs/prsTRAiNTs FA!L!NG ti.
gcpggg I USNAC tE 9le LETIN 79-02 84 DM D D
la ;',
m1
,4 13.
i b:
- U='
ia.
w G)h v
~
il.
10.
L htETOGftAM & CUEIEENT iperWIT/Rf3T1 TANT WMARA Fb_
L 7.
b S=
E 4.
3.
2.
i.
I
~
n s
a a
i a
e -
La 22 28 20 24 34 4.2 48 SD 54 SS &2 SA 10 74 78 12 as to 9418102 De itD IL4 fhfATED IDCTOftS Or SMITt F5r bN
.5.
A
~
1 IL mm.9 & stepoftT/RESTItANT taNIWJu rse AFTER IEANALYSIS I
14.
l 4a" 10.
L 7.
8
~
u S.
4 2.
l
,L E.
ll-o m u 2s sa 24.ss u 4s so u se a2 n m m m u u so u u na im aa ns ag am as i
m m W M F5r
_n.
3 T
4 N
1 2
Mt IA n,
R 6
T s
S f, I E
R 0
9
/
H T
6 R
8 O
S 1_
F 2
S P
I S-F 8
P Y
.I U
L 8
7 S
A Y
N J-4 1
7 t
F A
F O
E o.
R A
7 S
N s.
O s
F R
IT E
2 O
C T
s R
- i N
F 8
O U
A f
T s
F F
4 CA s
S r
F N
F o.
O s
D IT J
s.
E U
M T
4 B
U 2
LU I
D R
I 8
F M
4 A
E T
N
%h U
S K
3 I
N D
M I
4 I
T 3
N E
0 O
~
3 V
C a
I T
s.
A 2
8' L
2
-}
9 U
2 T
M 8
N U
i 1
.4 3
2 i.
o.
I E
C 3
s 9
8 F0
~
o M
1 HCA J.44. o vU $ m I.
T Z O F D @ oFm 5
)1L.2 F TA l
i ATTACHMENT 9
3AFETY FACTOR BACKFIT PROCESS-INDIVIDUAL l
SUPPORT / RESTRAINT IMPLEMENTATION PROCESS l
STARTING SCOPE: ALL SUPPORT / RESTRAINTS TASK OUTPUT l
SCREEN ALL S/RS TO
= @ SAFETY RELATED LIST OF IDENTIFY SAFETY-RELATED S/R'S SUPPORT RESTRAINTS Y
SCREEN @ TO IDENTIFY S/R'S
- @ LIST OF SAFETY-WITH EXPANSION ANCHORS RELATED S/R'S WITH EXR ANCHORS U
LIST OF SAFETY-SCREEN @TO IDENTIFY S/R'S RELATED S/R'S l
WITH FSr e 4.0 WITH EXR ANCHORS l
WITH FSr 24.0 u
PERFORM DETAILED EVALUATION OF @ TO IDENTIFY-*@ LIST OF SAFETY S/R'S REQUIRING REDESIGN RELATED S/R'S l
REQUIRING REDESIGN U
p+
PERFORM REDESIGN ON r @ REVISED DESIGN DRAWING /CALCS.
9 L _
PERFORM CONSTRUCTION r-
- REWORK ON r @ REWORK S/R'S l
9 L--
REINSPECT r @ REINSPECT S/R'S MATH MODEL REVIEW (ATTACHMENT 10)
FINAL WALKDOWN (ATTACHMENT II) _. _. _ _ __. _
l ATTACHMENT 10 SAFETY FACTOR BACKFIT PROCESS-PIPING STRESS ANALYSIS MATH MODEL REVIEW STARTING SCOPE: ALL STRESS MATH MODELS WITH ONE OR MORE REDESIGNED S/R'S (FSr OROTHER)
REVIEW STRESS MATH MODEL FOR LOCATION CHANGES u
u NO IMPACT I MPACT U
y S
REVISE PIPING ALYSI
. STRESS ANALYSIS DOCUMENTATION CLOSEOUT / N-STAMP n
y M
ACCEPTABLE S/R UNACCEPTABLE S/R LOCATION LOCATION y
U REVIEW / REDESIGN REVIEW AFFECTED S/R'S FOR LOAD S/R'S FOR LOCATION CHANGES CHANGE I
I f i t U
ISSUE REDESIGN TO NEW LOCATIONS NEW LOCATIONS CONSTRUCTION IDENTIFIED NOT POSSIBLE (SEE ATTACHMENT 9
{
FOR FOLLOW ON REvlSE ACTIVITES)
PIPING LAYOUT I
IDENTIFY NEW S/R LOCATIONS r.
ATTACHMENT l'
SAFETY FACTOR BACKFIT PROCESS CONSTRUCTION Q.C. FINAL WALKDOWN STARTING SCOPE: ALL STRESS MATH MODELS WITH ONE OR MORE REDESIGNED S/R's (FSr OR -
COMFL)ETE.ALL ERECTION /INSPECTI OTHER PERFORM FINAL WALKDOWN IfIf u
' ACCEPTABLE UNACCEPTABLE
_.RESULTS RESULTS o
'I 1r
~COMPLE'TE DESIGN REVIEW TJOCUMENTATION
. DISCREPANCIES CLOSECUT 1r y
EXISTIN ' DESIJNS EXISTlH9 LESl?NS UN. CCEPTAFLE ACCEPTAELE y
PERFORM REDESIC-N OF AFFECTED S/R's if ISSUE REDESIGN TO CONSTRUCTION SEE ATTACHMENT 9 FOLLCW ON ACTIVITl:S i
l i,
ATTACHMENT 12 Minimum Sample Size Allowable Defects t
74 0
t-115 1
1 156 2
183 3
e 198 4
255 5
300 6
335 7
368 8
392 9
416 10 432 11 448 12 465 13 476 14 l
486 15 500 16 m.... -.
, ATTACHMENT 13 Rigorous Systems Actual Allowed #
Hypothesis System Name w/FS<4 w/FS<4 Accepted
- BB Steam Generator Blowdown 2
4 Yes CA Auxiliary Feedwater 3
0 Yes l
FW Refueling Water 0
0 Yes II Incore Instrumentation 0
0 Yes KC Component Cool 4
16 Yes KD Diesel Generator Engine Cooling 0
0 Yes Water KF Spent Fuel Cooling 2
2 Yes i
LD Diesel Generator Engine Lubricating 0
0 Yes Oil NB Boron Recycle 0
7 Yes l
NC Reactor Coolant 0
2 Yes ND Residual Heat Remo;al 0
0 Yes NI Safety Injection 12 12 Yes NM Nuclear Sampling 0
0 Yes NR Boron Thermal Regeneration 0
0 Yes NS Containment Spray 0
0 Yes l
ATTACHMENT 13 (Cont'd)
Rigorous Systems Actual Allowed #
Hypothesis System Name w/FS<4 w/FS<4 Accepted
- NV Chemical & Volume Control 0
0 Yes RN Nuclear Service Water 4
6 Yes RV Containment Ventilation Cooling 1
9 Yes Water SA Auxiliary Steam 0
0 Yes SM Main Steam 0
0 Yes SV Main Steam Vent 0
0 Yes VQ Containment Pressure Control 0
0 Yes VX Containment Air Return Exchange 0
0 Yes WL Liquid Waste Recycle 2
5 Yes YC Chilled Water 0
36 Yes l
l l
1 l
1 l
i
ATTACHMENT 13 (Cont'd)
Alternate Analysis Systems l
l l
Actual Allowed #
Hypothesis System Name w/FS<4 w/FS<4 Accepted
- l FD Diesel Generator Engine Fuel Oil 0
0 Yes l
NF Ice Condenser Regrigeration 13 1
No RF Fire Protection 0
0 Yes VB Breathing Air 1
1 Yes VE Annulus Ventilation 6
0 No VG Diesel Generator 16 4
No I
l VI Instrument Air 0
0 Yes VS Station Air 0
2 Yes WE Equipment Decontamination 0
1 Yes WG Waste Gar 0
0 Yes WS Nuclear Solid Waste Disposal 8
0 No WZ Groundwater Drainage 40 1
No YM Demineralized Water 6
1 No 0 The acceptance criteria requires a 95% confidence level that no more than 5%
l of the anchor bolts on a system have a safety factor less than 4.
I j i
.