ML20203C243
| ML20203C243 | |
| Person / Time | |
|---|---|
| Site: | Monticello  |
| Issue date: | 02/11/1998 |
| From: | Schibonski C NORTHERN STATES POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| REF-GTECI-A-46, REF-GTECI-SC, TASK-A-46, TASK-OR GL-87-02, GL-87-2, TAC-M69460, NUDOCS 9802250124 | |
| Download: ML20203C243 (18) | |
Text
- o Northem States Power Company Monticello Nuclear Generating Plant 2807 West County Road 75 Monticello, MN 55362 February 11,1998 U S Nuclear Regulatory Commission GL 87-02 Attn: Document Control Desk USl A-46
\Vashington, DC 20555 MONTICELLO NUCLEAR GENERATING PLANT Docket No. 50-263 License No. DPR-22 Questions 7 and 8 in Response to Request for Additionalinformation on the Resolution of Unresolved Safety issue A-46 (TAC NO. M69460)
By letters dated October 8,1997 and November 12,1997, the NRC requested additional information on Resolution of Unresolved Safety issue (USI) A-46, Seismic Evaluation Report (TAC NO. M69460). By letter dated January 15,1998, Northern States Power provided responses to all but two of the questions. This letter provides Monticello's response to the remaining two questions and completes our reply to the referenced NRC letters.
This submittal contains no now NRC commitments, nor does it inodify any prior commitments.
Plje se conjact Sam Shirey, Sr Licensing Engineer, at (612) 295-1449 if you require additional formatioryrela d to est.
,I '/
Craig >Schibonski Acting P nt Manager Mont' llo Nuclear Generating Plant c: Regional Administrator-Ill, NRC NRR Project Manager, NRC Resident inspector, NRC State of Minnesota Attn: Kris Sanda j J Silberg i j
Attachments: Affidavit to the US Nuclear Regulatory Commission.
Exhibit A: Response to NRC Request for AdditionalInformation
/
9802250124 980211 '
_ _ _,_ _._ . llI.lill
.. ll 1.I!l.l!I ll
UNITED STATES NUCLEAR REGULATORY COMMISSION NORTHERN t TATES POWER COMPANY MONTICELLO NUCLEAR GENERATING PLANT DOCKET NO. 50-263 Questions 7 and 8 in Response to Request for Additional Information on the Resolution of Unresolved Safety Issue A-46 (TAC NO. M69460)
Northern States Power Company, a Minnesota corporation, hereby provides responses to questions 7 and 8 as requested by the USNRC in letters dated October 8,1997 and November 12,1997, titled " Request for Additional Information on the Resolution of Unresolved Safety issue A-46, Seismic Evaluation Report. (TAC No. M69460)*
This letter contains no restricted or other defense information.
NORTHER TATES P PANY By -
Craig A S'chibonski ctin lant Manager icello Nuclear Generating Plant On this // ay of AC7 W //,9f before me a notary public in and for said County, personally appeared Craig N. Schibonski, Acting Plant Manager, Monticello Nuclear Generating Plant, and being first duly sworn acknowledged that he is authorized to execute this document on behalf of Northern States Power Company, that he knows the contents thereof, and that to the best of his knowledge, information, and belief the statements made in it are true and that it is not interposed for delay.
STEPHEN R. BLEGEN
..- NOT'AYPutuc tmIGIE80EA Stt(phen R. Bleger( .nsy c m s o Jon at sees Notary Public - Minnesota -
Sherburne County My Commission Expires January 31,2000 2/5/98 MHV JtlCENSOSQUGRAl298 DOC U
RESPONSE TO NRC REQUEST FOR ADDITIONAL INFORMATION PERTAINING TO USl A-46, SEISMIC EVALUATION REPORT s
MONTICELLO NUCLEAR GENERATING STATION February 11,1998 Exhibit A
- 7. In Reference A, the response to Question 21(b) indicated that for Monticello, application of Generic Implementation Procedure (GlP) method A has been completed in accordance with the requirement that the safe shutdown earthquake (SSE)is defined at the ground surface. As requested in Reference B, please provide a technicaljustification for not using the in structure response spectra provided in Attachment 2 to Reference A.
Response
Monticello's SSE ground motion response spectra is defined at the free field ground surface which is consistent with SQUG Bounding Spectra.
The approved Generic Implementation Procedure (GIP) contains two optional methods for estimating seismic demand: Method A which estimates demand as 1.5 x Ground Response Spectrum (GRS) for equipment located at elevations under about 40 feet above the effective grade with fundamental frequencies greater than 8 Hz, and Method B which permits use ofin-structure response spectra (ISRS). Method B defines two types ofISRS,
" conservative, design" and " realistic, median-centered," and requires higher safety factors for the median-centered spectra.
NSP's understanding of the NRC's concern with the use of Method A is based on NRC correspondence with SQUG (Reference C) on this issue. Our understanding of the NRC p' .; ions in Reference C concerning use of Method A is that the utilities are to meet the following two positions: (1) when SQUG utilities provided information in their 120-day response letter (as requested by the staffin Supplement No. I to GL 87-02 D and Transmits Supplemental Safety Evaluation Report No. 2 (SSER No. 2), Reference D) on the methods used to calculate ISRS, licensees committed to using those spectra, and only those spectra, for all GIP evaluations and (2) that a precautionary statement in the GIP precludes use of Method A any time the plant's ISRS exceed the Method A spectra. These NRC positions are inconsistent with the approved GIP, and differ from the understanding among SQUG, the independent expert review panel, Senior Seismic Review and Advisory Panel (SSRAP),
and the NRC stafTat the time the GIP was developed. Our description of these two issues is provided below:
Purpose for Submitting ISRS to NRC In the 120 day response letter to the NRC (Reference E) NSP submitted information on methods used to calculate ISRS so the NRC could make their own determination of whether they found it to be " conservative, design" or " realistic, median-centered."
Neither the information request, the response, nor the NRC evaluation (Reference F) of the 120 day letter in any way committed to using either method as is now being inferred by the staff. Further, no changes were made to the GIP provisions which give utilities the option to use either Method A or Method B.
1
Response to NRC Request for AdditionalInformation February 11,1998 Exhibit A There is nothing in Generic implementation Procedurc Rev 2 (GIP 2) or SSER No. 2 wh!ch limits use of the GIP to hiethod 11. GlP 2 clearly states,in Part 11, Section 4.2, that either hiethod A or hiethod 11 may be used to compare seismic capacity to seism lc demand:
"The seismic capacity of an item of equipment can be compared to a seismic demand which is defined in terms of either a ground response spectrum or an in-structure response spectrum. Table 4-1 outlines these types of comparisons as either hiethod A or 11. hiethod A is for making a comparison with a ground r:.sponse snectrum: Section 4.2.3 discusses this type of comparison. hiethod f . : a comparison with an in structure rest onse srectrum: Section 4.2.4 discusses this type of comparison, hiethod A comparisons are Senerally easier to apply than hiethod 11 comparisons."
(GlP-2, pg. 4-8)
SSER No. 2 does not take exceptiori to this provision in the GIP but concludes that the j screening procedures and criteria are adequate and acceptable:
"The staff has reviewed the screening procedures ano criteria, llased on the evaluations and findings described in Sectic,ns 11.4.2,11.4.3, and 11.4.4 belou , the storf concludes that the screening procedures and criteria are adequate and acceptable only for verifying seismic adequacy ofequipment in USl A-46 plants, subject to the staff clarifications, laterpretations, exceptions and positions described in the sections that follow."
(SSER No. 2, pg.12)
There are no staff clarifications, interpretations, exceptions, or positions in SSER No. 2 which contradict oitr conclusion that we may use either Method A or Method 11.
The above interpretation of NRC's position on GIP 2, as contained in SSER No. 2, is supported by the NRC position in GL 87-02 which clearly allows use of 1.5 times ground response spectra (later called GIP hiethod A)in lieu of plant floor response spectra (also called ISRS):
"When horizolaal floor spectra exist, these spectra may be used to obtain the equipment spectral acceleration. Alternatively, for equipment mounted less than about 40 feet above grade, one and a half times the free-field horizontal design ground spectrum may be used to conservatively estimate the equipment spectral acceleration. For equipment mounted more than about 40 feet above grade, floor spectra must be used."
(GL 87-02, Enciasure, Section 3, page 5) 2
. v . -
R:sponse to NRC Reau st for Addition:l Inform: tion Fcbruary 11,1998 Exhibit A 1herefore, we conclude that the NRC's request for information in GL 87 02, Supplement No. I and SSER No. 2, and their review of these ISRS, was for the purpose of evaluating the adequacy of these spectra for use in GIP Method B and for deciding whether to classify them as either " conservative, design" or " realistic, median-centered." The licensee submittal of the ISRS in the 120 day response does not imply a commitment to use these ISRS cxclusive of Method A. GIP Method A is clearly an alternative to Method B, and no requirement to use the more conservative of the two methods is ctated or haplied.
Restrictions and I. imitations on Use of GlP Method A ,
in Reference C, the NRC quotes a precautionary note in the GlP that Method A is not intended for structures where amplification is expected to be more than about 1.5 times the free field ground response spectrum, and then the NRC concludes that if the plant's ISRS show calculated amplification greater than 1.5, use of Method A is precluded. The NRC has taken this precaution out of context. The full text of the precaution makes clear that it is intended to alert revievers not to use Method A for atypical structures, ud clearly indicates that Method A is intended for use on reinforced concrete frame and shear well structures and heavily braced steel frame structures which are typically used in coramercial nuclear plants. The remainder of the text in GIP 2 reads as follows:
"Scismic Capability Engineers should be alert for unusual, piant specific situations which couM cause the amplification factor to be greater than that of typical nuclear plant structures. The 1.5 amplification factor is only applicable to reinforced concrete frame and shear wall structures and to heavily braced steel frame s'.ructures."
(GIP 2, pg. 4-16)
The criteria for deciding whe;her this condition is met are not whether the calculated ISRS for a structure are more than 1.5 x GRS, but whether the structu e is similar to commercial nuclear plant structures, i.e., reinforced concrete frame and shear wall structures and heavily braced steel frame structures for which Method A was specifically intended. <
This interpretation of the GIP is consistent with the SSRAP report (GIP-2, Reference 5), upon which the GlP is based, and the NRC's review of this report in Generic Letter (GL) 87-02. SSRAP's position in this report is that calculated ISRS (1) are known to have excessive conservatism (2) are expected to exceed 1.5 x GRS (GIP Method A spectra), and (3) should not be used v hen assessing seismic ruggedness of floor-
-- mounted equipment. Their position on this issue is as follows:
"Oflen floor spectra for nuclear power plants are very conservatively computed. In such cases an amplification greatcr than 1.5 may be found even at elevations below 3
R:sponse to NRC R:qu:st for Addition:1 h Arm: tion Fcbru:ry 11,1998 Exhibit A 40 feet above grade. Ilowever, when more median centered analyses are perfonned that use reasonable damping levels for the structure and account for embedment and wave-scattering effects, these high amplifications are not observed with most canhquake ground motion records. The Seismic Safety Margin Research Program (SSMRP)(References 31 and 32)' has demonstrated the large conservatism which exists in traditionally computed floor spectra versus median floor spectra. Further, floor spectra measured less than 40 feet above the grade on moderately stiff portions of the Pleasant Valley Pump Station, the ilumboldt Bay Nuclear Power Plant (Reference 33), and Fukushima Nuclear Power Plant do not show amplifications over the ground spectra of more than 1.5 for frequencies above about 6 Ilz. In fact, on floors corresponding to near grade level, the flooi ;putra are less than or about equal to the ground spectra at frequencies above about 6 liz. Tims, it is SSRAP's judgment that amplifications greater than a factor of 1.5 are unlikely in stiff structures at elevations less than 40 feet above grade, except possibly at the ftmdamental frequency of the building where higher amplifications might occur when such a frequency is less than about 6 liz. Tims, for equipment with fundamental frequencle.t greater than about 8 llz in their as anchored condition, it wasjudged that floor spectral amplifications within 40 feet of grade would be less than 1.5 when reasonably computed using more median-centered approaches."
"It wasjudged by SSRAP that the use of ve'y conservative floor spectra should be avoided when assessing the seismic ruggedness of floor mounted equipment, it was also the opinion of SSRAP that many of the operating plants may only have these very conservatively computed floor spectra available. To avoid the burden of having to compete more realistic floor spectra, SSRAP decided to nachor its conclusions to ground spectra at the nuclear plant sites in those cases where this was judged to be reasonable."
[SSRAP lleport (OlP-2 Reference 5), pgs.101 103]
The NRC came to a similar conclusion in OL 87-02 where they state the following:
"The comparison of these seismic bounds with the design horizontal ground response spectrum [OIP Method A] isjudged to be acceptable for equipment mounted less than about 40 feet above grade (the top of the ground surrounding the building) and for moderately stiff structures."
(GL 87-02, Delosure, pg. 9)
The NRC also endorsed SSRAP's approach in their Regulatory Analysis of the USI A-46 program where NUREG 1211 (Reference G) states the following:
'NRC contractor, LLNL, presents the results of their work in NUREG/CR-1489 in which ISRS are estimated to have factors of conservatism ranging from 1.5 to 8.
4
R:sponso to NRC Request for Addition:l Inform: tion Fcbruary 11,1998 Exhibit A "The NitC staff has closely followed the SSitAP work and is in broad agreement with its conclusions. "
(NUlti G 1211,Section V.1, pg.17)
In summary, NSP believes that we have corn .i plenented the GIP and that the recent interpretation on the use of GIP Method A ..,. resents a new position which is inconsistent with documented safety evaluations.
1:. The NitC staff has concerns about the way the issue regarding the A-46 cable trays conduit raceway was being disposed of by licensees. The stafThad issued requests for additional information (ItAI) to several licensees on the issue. SQUG responded instead of the licensees i>ccause SQUG considered the ilAl to be generic in nature. The stafrissued a subsequent ilAl to SQUG as a follow up to SQUG's res,rwise, llowever, the staff found that the correspondence with SQUG did not achieve the ir.teaded results in that it did not address the technical concerns of the staff. "I herefore, the stafris directing the following question to each afTected licensee.
The GlP procedure recommended performing a limited analytic evaluation for selected raceways and cable trays. The procedure further recommended that when a ce.tain cable tray system can bejudged to be ductile and if the vertical load capacity of the anchorage can be established by a load cl.eck using three timm the dead weight, no further evaluation is needed to demonstrate lateral resistance to vibration from carthquakes.
The GIP procedure eliminates horizontal force evaluations by invoking ductility. Ilowever, the staffi :lieves that "non ductile cable trays" would eventually become ductile by inelastic deformation, buckling or failure of the non-ductile cable tray supports and members. The GIP procedure is a departure f;om conventionel methods of engineering evaluation and the GIP does not provide an adequate bases for dealing with those cable trays that initially are judged to be non ductile but are eventually called ductile by postulating failure of the lateral supports, if this procedure was followed for eliminating cable trays for further assessment at Monticello, then all the cable trays could conceivably be screened out from A-46 evaluation. The following questions are to elicit infonnation that would support the staffs safety evaluation of cable trays at Monticello, llesponse The GlP methodology for cable and conduit raceways is based on the criteria that (1) the electrical rables centinue to function and (2) the raceway systems continue to maintain overhead support:
" Cable and conduit raceway systems are considered seismically adequate if, during and following a safe shutdown earthquake, the electrical cables being supported by the 5
R:spons3 to NRC Requ:st for Addition llnform tion Fcbruary 11,1998 Exhibit A raceway systems can continae to function and the raceway system.: ontinue to maintain overhead support as denned in this section. Minor damage, such as member buckling or connectio: yielding,is considered acceptable behavior."
(GIP-2, Section 11.8.0, Page 8 2)
Using this criterie, SQUG, the NP.C, and the Senior Seismic Review and Advisory Panel (FSRAP) agreed on a two part evaluation technique as defined in the GlP consisting of(l) a seismic walkdown using eriteria and guidelines to verify that the raceway systems are bounded by successful earthquake experience and shake table test data, and (2) a limited analytical review of worst-case, representative samples of raceway support systems to verify they are at least as rugged under seismic loadings as those in the carthquake experience and shake table test data base that perfonned well.
These two evaluation techniques are based on the results of s.iccessful performance of many cable and conduit raceway systems in commer:ial and industrial facilities which were subjected to significant earthquakes or shake table test loadings. Many of the overhead mounted supports were inherently ductile, i.e., they allowed significant lateral deflection with ductile yielding in certain support members. Despite this ductile behavior, no visible signs of distress were observed and none of the support systems failed if these systems met the criteria contained in the GIP. For this reason, the NRC accepted the GIP methodology including the condition where lateral bracing fails to provide lateral support.
As defined in GlP-2, Section 11.8.3.3, on page 8 26, the primary support members may be considered ductile if the " brace buckles or has a connection failure before the primary support anchor capacity is reached ...". In addition, these vertical support members must also satisfy the one times and the three times the dead weight checks in the GIP as illustrated in Figure 8.6 of the GIP.
Therefore, we conclude that the GIP is an acceptable engineering method to evaluate the seismic adequacy of cable and conduit raceway systems even thcugh it is different from current methods of engineering evaluation. The NRC staff came to the same conclusion in SSER No. 2 where it states the following on pages 30 and 31:
" Evaluation and Conclusion "The staff has reviewed the guidelines proposed by the SQUG for evaluating the seismic adequacy of cable and conduit raceway systems. The main objective of the proposed guidelines was to develop a cost-effective means of verifying the seismic adequacy of raceway supports in USI A-46 plants. These guidelines were developed on the bases of analytical studies, shake-table experimental model tests, and assessment of the perfonnance of cable and conduit suppo;t systems in past earthquakes."
"The staff considers that the plant walkdown guidelines represent au acceptable approach for evaluating the seismic adequacy of existing cable and conduit raceways in USI A-46 plants. Alse, the stafragrees that the proposed analytical procedure is a 6
ww ,
R:sponsa to NRC Rcquest for Addition l Inform: tion Fcbru;ry 11,1998 Exhibit A reawnable approach to ensure that the cable and conduit raceways and supports in USI ;
A-46 plants, when all the guidelines are satisfied, are as rugged as those observed in the ;
past carthquake experience data. Although the proposed guidelines would not require detailed analyses and, therefore, would not predict the structural response of the raceway support systems, they should provide the needed rationale tojudge the seismic adequacy of the raceway support systems with a reasonable factor of safety. Therefore, the stafTconcludes that the proposed guidelines for evaluation of seismic adequacy of cable and conduit :aceways and their supports are acceptable subject to the staff evaluations described in this supplement."
(SSER No. 2, pgs. 30,31)
Nowhere in other parts of SSER No. 2 are there " staff evaluations" which contradict or limit the evaluation and conclusion given above.
De above RAI question also states the following:
"The GlP procedure eliminates horizontal force evaluations by invoicing ductility.
Ilowever, the staff believes that "non-ductile cable trays" would eventually beame ductile by inelastic deformation, buckling w failure of the non ductile cable tray supports and members. The GIP procedure is a departure from conventional methods of engineering evaluation and the GIP does not provide an adequate bases for dealing with those cable trays that initially are judged to be non-ductile but are eventually called ductile by postulating failure of the lateral supports. If this procedure was followed for climinating a cable trays for further assessment at Monticello, then all the enble trays could conceivably be screened out from A-46 evaluation."
This statement does not accurately reflect the procedure contained in the GIP. Ilorizontal force evaluations are to be performed on non-ductile raceway supports, even when a lateral brace buckles or fails, so that the associated reaction loads are determined in other parts of
. the support structure, particularly for overhead anchorage. Section 8.3.4 of Part 11 of GlP-2 states the folloaing:
"For diagonally-braced supports with ductile overhead anchorages, the load reaction imposed on the support anchorage during the Lateral Load Checl. does not need to exceed the buckling capacity of the brace or its connections. For example,ifit is shown that a brace buckles at 0.80g lateral load, then this load should be used for the lateral Load Check and not 2.0g. For diagonally braced supports where the anchorage is not ductile, the portion of the lateral load that is not resisted by the brace should be redistributed as bending stress to the overhead connection. The loads in the diagonal brace will cause additional vertical and horizontal lods on the anchorage, which should be accounted for.
"An upper and lower bound estimate should be used for buckling capacity of the brace, whichever is worse, for the overhead anchorage. There is considerable variation in test 7
R:spons2 to NRC R:qu;st for Addition:1 Inform; tion Fcbru:ry 11,1998 Exhibit A data capacity for light metal strut framing connections. An upper bound estimate of 2.0 times the realistic capacities discussed in Section 8.3.8 can be used for these connection types."
(GIP-2, Yart II, pg. 8 29)
Therefore, we conclude that the NRC concern that "all the cable trays could conceivably be screened out from A 46 evaluations" does not accurately reflect proper implementation of the GIP.
8a. Define ductility in engineering terms as used at Monticello for the USI A-46 evaluation.
Clarify how this definition is applied to actual system configurations at Monticello plant consistently for the purpose of analytical evaluation.
Response
Ductility as it applies to cable raceway supports is defined in the first paragraph of Section 11.8.3.3 of the GlP as follows:
"An evaluation should be conducted of the supports selected for review to characterize their response to lateral seismic motion as either ductile or potentially non-ductile. Supports suspended only from overhead may be characterized as ductile if they can respond to lateral seismic motion by swinging freely wnhout degradation of primary vertical support connections and anchorage. Ductile, inelastic performance such as clip angle yielding or vertical support member yielding is acceptable so long as deformation does not lead to brittle or premature failure of overhead vertical support."
(GIP-2, Part II, pg. 8-24)
Each of the worst case samples of raceway supports which were selected for the GIP Limited Analytical Review (LAP j were evaluated using the procedure given in GIP Section 11.8.3.3 to determine whether they are ductile or non-ductile based on the examples shown in GIP Figures 8 7 and 8 8.
The GIP methodology does not require e.n et aluation of the " maximum ductility" for any of the cable tray supports. We consider our review of the effect of ductility to be in accordance with GIP-2, as approved by SSER No. 2, which implicitly accounts for ductility in its derivation.
8b. Provide the total number of raceways that were selected for worst-case analytical calculations and were classified as ducti!c in your A-46 evaluation and for which you did C
R:sponse to NRC Requ:st for Addition:l Inforrn: tion Fcbruary 11,1998 i
Exhibit A not perform a horizontal load evaluation. Indicate the approximate percentage of such raceways as compared with the population selected for analytical review. Discuss how the ductility concept is used in your walkdown procedures.
Response
A total of twelve Limited Analytical Reviews were done. Ihed on the procedure in GIP 2, Section 8.3.3, Ductility Check, the total number of these raceway supports which were classified as ductile for the Limited Analytical Review is eight (67%). As provided GIP 2, Section 8.3.4, Lateral Load Check, an explicit lateral load check is not required for any of these supports.
The GIP method for evaluating the seismic adequacy of cable and conduit raceway systems is a two step process. First, a plant walkdown is conducted in which the Seismic Capability Engiricers (SCEs) inspect the raceway systems m the areas where cable or conduit for safe shutdown equipment may be located. These raceways are evaluated against the " Inclusion Rules" described in GIP Section 11.8.2.2; they are also evaluated for "Other Seismic Perfonnance Concerns" described in GIP Section 11.8.2.3. During this plant walkdown, the SCEs would select a sample of the raceway supports which appear to have the lowest estimated seismic margin based on l their experience and judgment (GIP Section 11.8.2.4). The size of the sample would depend upon the diversity and complexity of the design and construction of the plant's raceway support system. A Limited Analytical Review (LAR) cf these worst-i case supports is then performeA using the guidelines and criteria in Section 11.8.3 of the GIP. The nurpose of LAR is not to simulate potential scismic performance or stresses, but to analytically correlate, approximately, conditions within the plant with conditions that. perfonned well in the earthquake experience database.
The walkdown procedures used at our plant are described in Section 8.2 of GIP 2.
These procedures do not require the ductility concept to be used during the walkdown. The only place where the ductility concept is used is in the Limited Analytical Review. The procedure used for this review is contained in GlP-2, Section 8.3.
Sc. Describe the typical configurations of your ductile raceways (dimension, member size, supports, etc.).
Response
Typical ductile raceway supports are light metal framing such as vertical Unistrut P1000 or P1001 members that are supnorted by standard Unistrut angle fittings. The tray is supported Unistrut brackets attached to the vertical member with spring nuts.
9
R:sponsa to NRC Requ:st for AdditionalInform tion Fcbru:ry 11,1998 5
Exhibit A The length of the vertical member varies depending on the number of tray being supported.
8d. Justify the position that ductile raceways need not be evaluated for honzontal load. When a reference is provided, state the pag: number and paragraph. The reference should be self contained, and not refer to rmother reference.
Response
The justification for stating that ductile raceways need not be evaluated for horizontal load is provided in Section 8 of the GIP and in the Senior Seismic Review and Advisory Panel (SSRAP) report (Reference 11) upon which the GIP is based.
The GIP, Section 11.8.3, Limited Analytical Review Guidelines, states the following:
"As shown in Figure 8-6, supports characterized as ductile do not require an explicit lateral load check. Instead, seismic ruggedness for ductile raceway supports i;; assured by the Vertical Capacity Check (Section 8.3.2). The high vertical capacity of the ductile data base raceway supports is the main attribute credited for their good seismic performance."
i (GIP-2, Part II, pg. 8-19) l The basis for not evaluating horizontal loads during the Limited Analytical Review of raceway systems with ductile supports is described further on page 17, paragraphs 1 and 2 of Reference 11 where it states the folicwing:
"A limited analytical review shall be perfonned on those cable tray supports selected b/the SRT and the walkdown engineers as representative of conditions within the plant with the lowest estimated seismic margin. The intent of this limited analytical review is not to simulate potential seismic performance or stresses, but to correlate, approximately, conditions within the plant analytically with onditions that performed well in the experience data base.
"It is imp srtant for the analyst to understand the philosophy behind this limited analytical review. As previously discussed within this report, cable trays and supports have performed extremely well in past earthquakes and shaking table tests with few exceptions. The trays and their supports typically act as pendulums and wiggle and sway but do not fail. Ductile inelastic performance such as yielding of clip angle supports or steel vertical support members is completely acceptable as it allows the cable tray to deform and move without brittle or premature failure. The high damping inherent in cable tray systema reduces the dynamic motions resulting from the inelastic performance and maintains --
integrity."
(Reference 11, pg. 26) 10 i
R:sponsa to NRC Requ:st for Additional Inform: tion Fcbruary 11,1998 Exhibit A Reference 11 goes on to say on page 26, paragraph 1, tha t eartiluake experience and shake table tests show that if the ductile raceway supports have a large dead load margin (i.e., they pass the GIP three times dead load ch ck), no lateral load check is required:
"It must be kept in mind that this limited analytict.1 review is not intended to simulate potential seismic performance or stresses, but to correlate approximately conditions within the plant analytically with conditions that performed well in the experience data base. The rational for the checks is as follows: . . . The 3.0 times dead load without eccentricity check is a simple check to insure that the basic j connectors have a large dead load margin. The experience data bas: supports pass l this check and verification of a large dead load margin provides assurance that if an isolated support should fall for some unforeseen situation, that a progressive support failure mechanism is unlikely. For ductile mechanisms, no lateral load check is required consistent with the experience data base and shake table test experience."
(Reference 11, pg. 26)
The NRC stafTthoroughly reviewed the GlP methodology during the period of time from 1987 to 1992. As a result of that review, the staff took the position on pages 30 and 31 of SSER No. 2 that:
... the plant walkdown guidelines represent an acceptable approach for evaluating the seismic adequacy of existing cable and conduit raceways in USl A 46 plants.
Also, the stafragrees that the proposed analytical procedure is a reasonable approach to ensure that the cable and conduit raceways and supports in USI A-46 plants, when all the guidelines are satisfied are as rugged as those observed in the past carthquake experience data. Although the proposed guiden nes would not require detail analyses and, therefore, would not predict the structural response of the raceway support systems, they should provide the needed rationale tojudge the seismic adequacy of the raceway support systems with a reasonable factor of safety. Therefore, the staff concludes that the proposed guidelines for evalaation of seismic adequacy of cable and conduit raceways and their supports are l acceptable subject to the staff evaluations described in this supplement."
(SSER No. 2, pgs. 30,31)
We conclude from the above statement that the NRC staff position is that the GIP
! muhod is acceptable for evaluating the seismic adequacy of cable and conduit raceways and their supports. We consider our review of the effect oflateral loads on cable and conduit raceway systems to be in accordance with the requirements and intent of GIP-2 as approved for use by the NRC staffin SSER No. 2.
11
R:sponse to NRC Requ:st for AdditionalInform tion Fcbru:ry 11,1998 Exhibit A 8e. In the evaluation of the cable trays and raceways, if the ductility of the attachments is assumed in one horizontal direction, does it necessarily follow that the same system is ductile in the perpendicular direction? Ifit is not ductile in the perpendicular direction, how was the seismic adequacy of the attachments evaluated?
Response
The ductility of supports for cable and conduit raceway in the longitudinal direction is addressed using the procedure in GlP, Section 11.8.2.3, where our raceway systems were evaluated for "hard spot" supports. No other analyses for longitudinal forces are required for caole tray supports hung from above or attached to a wall. Aside from this check for "hard spots," the GlP does not require an evaluation of raceway support ductility in the longitudinal direction. The basis for this is found in SSRAP Report,
" Review Procedure to Assess Seismic Ruggedness of Cantilever Bracket Cable Tray Supports," Reference 11, page 27, paragraph 2 which states:
"SSRAP also does not intend that the cable tray supports hung from above or attached to a wall be checked for longitudinal lateral forces, i.e., lateral forces parallel to the long run of the cables. There are numerous cases of this condition in the experience data base without damage or distress to the cables (Reference 7),
and SSRAP does not believe that analytical checks are needed for this condition.
The experience data base justifies this situation, and ductile pendulum action will be sullicient for good performance."
(Reference 11, pg. 27)
Reference 7 cited in the above quotation is the same as GIP Reference 19, EPRI Report NP 7153,"Longitudir.al is,ad Resistance in Seismic Experience Data Base Raceway Systems," March 1991.
As discussed in response to Question 8d above, the GlP nu. Sod, including the loadings on the raceway supports in the longitudinal direction, was thoroughly reviewed and accepted by the NRC staffin SSER No. 2. We consider our review of the effect oflongitudinal loads on cable and conduit raceway systems to be in accordance with GIP-2 as approved for use by the NRC stafTin SSER No. 2.
M f. Discuss any raceways and cable trays including supports in your plant that are outside of the experience data. Explain what criteria are used for establishing their safety adequacy and specify your plan for resolution of outliers that did not meet the acceptance criteria.
- Provide examples of the configurations of such raceways and cable trays including supports. Also, indicate the percentage of raceway and cable trays examined during the walkdowns of the safe shut down path, llow are they going to be evaluated and disposed?
12
R:sponse to NRC Requ:st for Addition:I Inform; tion Fcbru ry 11,1998 6
Exhibit A j
Response
All the cable and conduit raceway systems in the plant are within the scope of the seismic review procedures contained in the GIP except those which were identified as outliers. GlP Section 11.8,0, page 8 2, paragraph 2 describes the scope of raceway systems which are covered by the GIP as follows:
"The scismic review guidelines contained in this section are applicable to steel and aluminum cable tray and conduit support systems at any clevation in a nuclear power plant, provided the Bound:ng Spectmm (shown in Section 4. Figure 4 2) envelopes the largest horizontal component of the 5% damped, free field, safe shutdown carthquake (SSE) ground response spectrum to which the nuclear plant is licensed."
(GIP-2, pg. 8 2)
The raceways which were classified as outliers did not meet the " Inclusion Rules,"
had "Significant Other Seismic Performance Concerns," or did not satisfy the
" Limited Analytical Review Guidelines" contained in the GlP. The resolution of outliers may include additional analysis, investigation, or a modification to the r may which would allow the raceway to meet the GIP screening criteria.
The following are the cable and conduit systems outliers and the resolution of the outliers.
The following raceways were deemed an outlier because they did not meet the inclusion Rules.
TB A7 - Some cantilever bracket supported tray were found to not be secured to the brackets, hold downs were added to preclude the potential for the tray to slide off of the brackets.
The following raceways were deemed an outlier beccuse they did not satisfy th.
Limited Analytical Review Guidelines.
CB-A2 - Braced cantilever bracket supports and cantilever bracket supports, were modified to meet the Limited Analytical Review Guidelines.
The following raceway was deemed an outlier because they had significant Other Seismic Performance Concems.
DPil 1 - A loose conduit support was found, it was tightened.
RB A19 - During the walkdown a vertical run of tray, that was supported off of embedded Unistrut, could not be verified as meeting the tray span criteria, a suppe: was added to meet the span criteria. A subsequent investigation confirmed that the tray bad previously been supported such that it did meet the span criteria and the added support would not have been required, t 13
R:sr" % NRC R:qu:st for Additionti inform: tion February 11,1998 Exhibit A The following raceway was deemed an outlier because they had spatial interaction concerns.
RB A17 - A rod hung conduit had the potential to pull on a cable going into a panel. This could potentially damage cable, a support was installed to prevent the conduit from swinging during a seismic event thus climinating the potential of damaging the cable.
The RAI question also asks for the percentage of raceways and cable trays examined during the walkdowns. The total number of rr.ceway and cable tray included in the walkdown was not collected during the walkdown since it is not part of the GIP guidelines. As provided in GIP 2, Section 8.2.1, the purpose of the plant walkdowa is to:
"(1) verify that the cable and conduit raceway systems meet the inclusion Rules given in Section 8.2.2;(2) note and evaluate any of the Other Seismic Performance Concerns given it. Section 8.2.3; (3) select a sample of representative worst-case raceway supports as described in Section 8.2.4; and (4)j; ige whether there are any seismic spatial interactions which could adversely affect the perfonnance of the raceway system as outlined in Section 8.2.5."
(GIP-2, pg. 8 8)
Although the total number of raceway and cable tray is not know based on an estimation of the total number, the supports that v re deemed outliers are estimated to be less than 2% of the whole population of raceway supports.
8g. Submit the evaluation and analysis for four of the representative sample raceway (one single non-ductile, one single ductile, one multiple non-ductile, and one multiple ductile raceway), including the configurations (dimension, member size, supports, etc.).
Response
Multip' support analysis is not part of the GIP methodology. Evaluations of a single ductile WAR RB-A13-L1) and a single non-ductile (LAR RB A8 L1) raceway support have been previously provided in NSP letter to the NRC dated January 15, 1998 in response to the NRC November 12,1997 request for additional infonnation.
14
R:sponse to NRC Requ:st for Addition:l Inform; tion Fcbrutry 10,1998
. Exhibit A Reference _s_;
A. Letter from William J. Ilill to NRC titled," Response to Request for Additional Information on the Resolution of Unresolved Safety issue A-46 (TAC No. M69460)," dated April 29,1997
- 11. Letter from NRC to Northern State Power Company titled," Request for Additional Information on the Resolutioa of the Unresolved Safety issue A-46 (TAC No. M69460)," dated January 29
. 1997 C. NRC (J. Stolz) letter to SQUG (N. Smith) dated December 2,1997, Evaluation of Seismic Qualification Utility Group's Response to Generic issues included in NRC's Request for Additional infonnation.
D. Supplement No.1 to Generic Letter (GL) 87-02 That Transmits Supplemental Safety Evaluation Report No. 2 (SSER No. 2) on SQUG Generic Implementation Procedure, Revision 2, as Corrected on February 14,1992 (GIP 2) 3 E. NSP Letter to NRC dated September 21,1992 " Response to Supplement I to Generic Letter 87-02 on SQUG Resolution of USI A-46" F. NRC letter to NSP dated December 10,1992 "Monticello - Evaluation of Licensee's 120 Day Response to Supplement No. ! to Generic Letter 87-02 (TAC No. M69460)
G. NUREG-1211 " Regulatory Analysis for Resolution of Unresolved Safety Issue A-46, Seismic Qualification of Equipment in Operating Plants"
- 11. SSRAP Report," Review Procedure to Assess Seismic Ruggedness of Cantilever Bracket Cable Tray Supports," Senior Seismic Review and Advisory Panel (SSRAP), Rev. 3.0, March 1,1991, 15