ML20214W845
| ML20214W845 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 12/05/1986 |
| From: | Youngblood B Office of Nuclear Reactor Regulation |
| To: | White S TENNESSEE VALLEY AUTHORITY |
| References | |
| NUDOCS 8612100425 | |
| Download: ML20214W845 (19) | |
Text
p c
Docket Nos.: 50-327 and-50-328 Mr. S.A. White Manager of Nuclear Power 5 DEC 1986 Tennessee Valley Authority 6N 38A Lookout Place 1101 Market Street Chattanooga, Tennessee 37402-2801
Subject:
Transmittal of Draft Safety Evaulation on the Interim Acceptance Criteria for Cable Tray Supports Enclosed is a draf t copy of the staff evaluation on the proposed interim acceptance critera for cable tray supports.
The staff intends to incorporate this evaluation into the Safety Evaluation Report (SER) on Volume 2 of the Nuclear Perfonnance Plan which addresses Sequoyah. The purpose of this letter is to transmit the report to the Tennessee Valley Authority (TVA) and to request that TVA respond to the confirmatory items identified in the report.
With the exception of the confinnatory items, the staff has found the TVA interim acceptance criteria for c'able. tray supports acceptable. However, TVA should address the confirmatory items identified on pages 16 and 17 of the enclosure prior to restart of either Sequoyah unit.
In order for the staff to support the TVA projected restart date for Sequoyah, TVA should respond to the confirmatory itens within four weeks of the date of this letter.
If TVA cannot meet this date, it should provide its proposed date to the staff in writing within one week of the date of this letter.
If you require any additional assistance, please contact the project manager for the Sequoyah SER, Mr. Joseph J. Holonich at (301) 492-7270.
Sincerely, B. J. Youngblood, Director PWR Project Directorate #4 Division of PWR Licensing-A
Enclosure:
As stated cc: See next page a
DISTRIBUTION:
See next page
- SEE PREVIOUS CONCURRENCE o
7 PWR#4/DPWR-A PWRt4/DPtlR-A Pl!PJ4fDPWR-A PWR#4/DPPR-A fild
,JP4R-A r
- JHolonich/ rad
- TJVenyon HDUMi'n
- CStahle BJW gblood 12/02/86 12/02/86 12/J /86 12/02/86 12/ /86
% g 05000327 861205 R
P PDR
I
,1 Mr. S.A. White Tennessee Valley Authority Sequoyah Nuclear Plant cc:
Tennessee Departnent of Public Regional liHristrator, Region II Health U.S. Nuclear Regulatory Ccncission, ATTN:
Director, Bureau of 101 Marietta Street, N.W., Suite 2900 Environmental Health Services Atlante, Gecrgia 30323 Cordell Hull Building Nashville, Tennessee 37219 J.A. Kirkebo ATTN:
D.L. Williams Mr. Michael H. Mobley, Director Tennessee Valley Authority Division of Raciological Health 400 West Summit Hill Drive, W12 A12 T.E.R.R.A. Building Knoxville, Tennessee 37902 150 9th Avenue North hashville, Tennessee 37203 Mr. Bob Faas Westinghouse Electric Corp.
County Judge P.O. Box 355 Hamilton County Courthouse Pittsburgh, Pennsylvania 15230 Chattancoga, Ternessee 37402 R. L. Gridley Tennessee Valley Authority SN 157P Lookout Place Chattanooga, Tertnessee 37402-2801 M. R. Harding Tennessee Valley Authority Sequoyah Huclear Plant P.O. Box 2000 Soddy Daisy, Tennessee 37379 Resident Inspector /Sequoyah NPS c/o U.S. Nuclear Regulatory Comission 2600 Igou Ferry Road Soddy Daisy, Tennessee 37379 H.L. Abercrcrrbie Tennessee Valley Authority Sequoyah huclear Plant P.O. Ecx 2000 Soddy Daisy, Tennessee 37379
l o
O O
SE000YA4 NUCLEAR POWER STATION, UNITS 1 & 7 EVALUATION OF THE PROPOSED INTERIM ACCEPTANCE CRITERIA FOR CABLE TRAY SUPPORTS 1.
INTRODUCTION The original design criteria for cable tray supports were developed in 1972-74 time frame. The design criteria included the effects of earthquake but did not consider the effects of Design Rasis Accidents (DBA).
In 1975, the original design criteria were revised to include the DBA loads, but the original designs were never reviewed to ensure compliance with the revised criteria. Though this deficiency affected only the cable tray supports attached to the steel containment vessel (SCV), other deficiencies found in 1984 and 1986 dictated a thorough review of the adequacy of all the cable tray supports. During the current review, the TVA staff discovered that the existing cable tray supports would not satisfy the basic FSAR commitments. On July 17-18, 1986 in a meeting held in Bethesda, the TVA staff requested the NRC staff approval of a set of interim acceptance criteria less stringent than those committed in the FSAR.
It was indicated that such an approval would enable TVA to restart the plant in early 1987. As a part of the request, TVA also committed that the original FSAR criteria would be restored in an orderly manner after the restart for the affected cable tray supports.
The NRC staff evaluation consisted of (11 assuring that the proposed interim acceptance criteria were justifiable from the standpoint of safe operation of the plant, and (21 confinning that the design calculations for cable tray supports were, as a minimum, in conformance with the criteria. The NRC staff and its consultants (Brookhaven National Laboratory) visited the plant twice and held meetings with the TVA staff - once in July 71-24, 1986 and a more extensive audit in September 29 through October 3,1986. Specific reouests for additional information have been developed as a result of these meetings.
TVA responded to the ouestions resulting from the July 2174 meetings in a report entitled, "Seouoyah Nuclear Power Plant, NQC Technical Information Request on Interim Acceptance Criteria." This report contains a discussion of the justification for the Interim Acceptance Criteria and a presentation of the method of implementing the criteria, l
i
E
?
-?-
Durino the audit of September ?9 to October 3,1986, the cable tray support walkdowns perfomed by TVA were evaluated by physical inspection of the plant; calculations perfomed by TVA to evaluate the adequacy of cable tray support systems with respect to the Interim Acceptance Criteria were reviewed; additional data supportino the interim Acceptance Criteria were also reviewed; and a portion of the concrete strength test data were evaluated.
P.
INTERTP ACCEPTANCE CRITERIA (a) Damping TVA proposes to use 7% of critical damping for the cable tray supports for the SSE/DBA loading as compared with the 5% allowed in the SQN FSAR. The following arguments are made in support of this criteria:
1.
There are substantial cable tray test data to demonstrate that the damping for cable tray supports is considerably larger than 7%. The cable trays at SON have the fundamental dynamic response frequencies and general characteristics of those tested.
2.
Vogtle was allowed to use 15% damping for their cable tray supports which are very similar to those at SON.
3.
Reg. Guide 1.61 allows 7% damping for bolted structures. While some of the cable tray supports are welded most of the mass is on the trays which are bolted to the supports.
There has been a considerable amount of data which indicates that damoing in cable tray systems is larger than 5% for SSE type loadings. This occurs because of the considerable damping in the cables themselves, and in the cable connection to the tray. During the walkdowns perfomed in the week of I
September 29, 1986, it was verified that the SON cable trays and cable tray suoports are generally similar to those tested and found in other nuclear power plants.
The staff and consultants believe that the tests performed on cable trays (test-results indicate damping values in the rance of 1n-20%) are applicable to SON.
In addition, TVA has performed calculations to determine the effect of this increase in damping. The following are typical stress ratios (defined as actual / allowable) for cable tray supoorts in the Auxiliary Ruilding.
Support Member Stress Patio Stress Ratio No.
7% Damping 5% Damoing Section-P Main Member 1.397 1.379 Bracket 0.532 0.554 Joint 0.516 0.521 Anchorage 1.49 1.51 IG Main Member 1.038 1.045 Bracket 0.863 0.875 Joint 1.154 1.277 Anchorage 1.403 1.55 5
Main Member 1.04 1.005 Rracket 0.555 0.55R Joint 0.55 0.584 Anchorage 1.13 1.17 These calculations indicate that change in dampino from 5% to 7% has little effect on the stress ratios. Thus, for the restart purposes, the 7% damping proposed by TVA for DBA/SSE loading is acceptable to the staff.
(b) 09A - SSE Load Combination 1
The FSAP. commitment required the absolute sum combination of SSE and DPA loading effects. TVA proposes to use the SRSS combination for the Interim
.a -
4 Acceptance Criteria. it is araued that the SSE and DRA loads are both low probability events and are unlikely to occur tocether, therefore, use of the SRSS combination of their load effects is appropriate.
The proposed approach is reasonable because of the uncoupled nature of the SSE and DBA loadings. Both loads are dynamic and the absolute sum of their effects would only occur if the SSE and D9A events are to occur at the same time and the peak responses of the tray supports to both the SSE and the 09A events are to coincide. The probability of such a coincidence to occur has been estimated to be rather low. Thus, the SRSS method is considered as a reasonable load combination approach for the plant restart and is acceptable to the staff.
(c) Elimination of 1/2 SSE Load Case The FSAR commitment requires consideration of the SSE and 1/2 SSE loads. TVA proposes to use the SSE loading only for the Interim Acceptance Criteria.
it is argued that the SSE case is usually more severe and that the safe shutdown of the plant is assured if the SSE criterion is met.
The 1/2 SSE load is by definition, less than the SSE load (ignoring the effect of the damping ratio). Generally, the comouted SSE stresses, when compared with the SSE stress allowables tend to be more critical than the correspond-i ing stress comparison for the 1/2 SSE case.
It should be noted, however, that i
several of the proposed Interim Acceptance Criteria relax the allowable stresses for the SSE loading case. This fact could sometimes make the 1/? SSE loadina l
case more critical than that of the SSE from the standpoint of design. However, l
a demonstration that the plant can be safely shutdown for the SEE automatically l
shows that it could be safely shutdown for the 1/? SSE. The criterion assures the safe shutdown of the plant under SSE loading. Additionally, the plant technical specification requires plant shutdown subsequent to seismic event eoualing or exceeding the 1/2 SSE acceleration levels. The proposed elimination of 1/? SSE case is acceptable to the staff on an interim basis.
L
V (d) Allowable Stresses The FSAR commitment requires that the cable trav stresses be less than 0.9 times yield for SSE/DRA loadino. TVA proposes to change this requirement to 1.7 times AISC allowables for SSE plus nRA loading, and 1.6 times ATSC allowables for SSE alone. The,iustification provided by TVA point to the fact that these allowables are stated in the Standard Review Plan and have been used in the review and approval of many plants.
Considering the high ductility of the steel (ASTN A36, A441, AS?7, A572) used in nuclear power plant structures, the Standard Review Dlan allows up to 1.7 times the AISC allowable stresses under low probabilitv loadings such as SSE and DBA. This has been the existing practice in the licensing of nuclear power plants. During the audit, the staff and consultants verified that the actual AISC allowable was reduced if the structural member section was not compact and the 1.6 or 1.7 factor is applied to this reduced AISC allowable stresses. The criterion of using up to 1.7 times the working stress allowables for cable tray support calculations is acceptable to the staff and consultants for the Sequoyah restart evaluation.
3.
IMPLEMENTATION OF INTERIM CRITERIA fal Cable Tray Supports Attached to Steel Containment Vessel The reevaluation of supports attached to the steel containment vessel was re-quired to resolve NCR SONCER 8414 The NCR addressed the fact that the Cable tray supports on the steel containment vessel (SCV1 were not designed for the loadings due to the Design Basis Accident (DBA1 There are a total of 560 cable tray supports attached to the steel containment vessel. All supports are attached to the outside o' the vessel by welding to the horizontal or vertical vessel stiffeners. Support members are generally i
1 l
,- 4" x 4" or 2" x 2" tube steel members. Cable travs are generally attached to the supports by clip angles which are welded to the support member and bolted to the cable tray. Most supports are simple ?" x 2" cantilever brackets welded to vertical stiffeners. The next largest cateoory of supports are ?" x ?"
cantilever brackets welded to a 4" x 4" member spannino between vertical stiff-eners. Most supports were analyzed by grouping all similar configurations and selecting the worst case envelope of the supports within each group. The ma.fority of the supports (5511 were enveloped by five typical designs. The remaining nine unique supports were individually analyzed. A walkdown of the cable tray systems was perfomed to establish actual tray loading. Measurements of the
~
c'ross section of cable trays were taken and actual tray loadings were calculated from the profile measurements.
The GTSTRUDL computer code was used to analyze the supports. The cable trav and its supports were modeled using elastic beam elements. A typical model included two supports and one cable tray span. The flexibility of the model support points was modeled using spring constants detemined by a finite element analysis of the containment vessel and stiffeners. Cable mass and tributary mass of the adiacent spans were included as lumped masses. Response s
spectrum analysis was used to analyze the SSE and DBA events. The events were analyzed separately using 10 percent peak frequency broadened, and 7 percent damped spectra. Modal response combination was perfomed by the SRSS method.
The directional response combination for the DRA event was implemented by absolute sumation of the three directional responses.
For the SSE, the directional response combination was perfomed by taking the absolute sun of the highest horizontal response and the vertical response. The DBA response was combined with the SSE response by the SRSS method.
Finally, the response due to dead weight was combined absolutely with the combined response of the SSE and D9A. Resulting stresses were evaluated with respect to 1.7 times the AISr allowables.
The effects of containment vessel expansion due to 09A thermal and pressure loading on the cable tray supports were also evaluated using the therral
7 loadino capabilities of GTSTRUDL. The containment expansion effects due to pressure were converted to an equivalent temperature gradient and then added to the actual themal gradient. The total temperature gradient effects were applied to the cable trays and supports to determine their stresses.
The largest reaction-load from the cable trav support analysis was applied to a containment vessel model to determine stresses in the vessel wall and sti f feners. Maximum stresses were evaluated against the applicable ASME code allowables.
Supports. which failed to meet the Interim Acceptance Criteria were analyzed using the actual tray loading detemined by the field walkdown.
If the criteria were not met with the reduced weight, the load rating of the tray was reduced and controls were established to prevent addition of weight beyond the reduced capacity.
TVA has completed the calculations for all the supports attached to the containment. The results indicated a need for modification of three existing supports and the addition of twelve new supports. All modified and new supports were designed to meet original design criteria requirements. Two of the modifications were required to prevent overstressing the supports and one modification was reouired to prevent nyerstressing the containment stiffeners.
Twelve additional supports were required in areas where span lengths exceeded the allowables.
The staff and consultants reviewed sample qualification calculations and per-fomed a walkdown of the affected supports. The staff audit team also reviewed selected calculations coverino the DAA response spectra generation, themal and pressure induced displacements, stiffness of the SCV stiffeners at support attachment points, and effects of support loads on the SCV wall and stiffeners.
Based on the audit results, the staff and consultants concluded that methods used in reevaluating the SCV cable tray supports were adequate and that the interim acceptance criteria were appropriately implemented to qualify the supports for the plant restart.
.. (b) Cable Tray Supports on the Reactor Ruildino Shield Wall Many cable trays located in the annulus between the steel containment vessel and the shield wall are supported from the shield wall.
In these cases the base plate of the cable tray support is bolted directly to the shield wall using wedge type expansion bolts. These supports consist of either cantilevered tube steel configurations or tube steel members mounted parallel and bolted directly, with little clearance to the shield wall. Because the total annulus clearance is only 5 feet the maximum span length of the main member in the cantilevered configurations is less than 5 feet. TVA detemined that because the surface mounted tube supports were mounted ad,iacent to the concrete their response amplifications to seismic inputs would be negligible.
Therefore, these surface mounted supports are qualified for the seismic response of the reactor shield building at their points of attachment. All cantilevered supports, on the other hand, were qualified either by individual analysis or by comparison to cable tray and supports envelopino configurations for which analyses were performed.
Although there are approximately 400 supports attached to the shield wall they are segregated into three generic and a number of special support configura-tions representing the cantilevered and the surface mounted types. For the three generic configurations, TVA selected a bounding or enveloping case to evaluate their acceptability based on consideration of suoport location, loading and member span. Supports identified as MK 9e, MK lle and MK 18b were the bounding cases since each was installed at a high elevation, carried maximum loads (4 trays 1 and exhibited maximun member spans. The special configuration supports were each evaluated as they exhibited unique configurations. The TVA selection and categorization of the supports was accepted by the staff and the consultants.
A walkdown of all shield wall mounted supports was performed by TVA.
In the walkdown for the generic and special supports, the configurations were
o-confirmed; the dimensions of the base plate including any eccentricities of the tube attachments and bolt holes and the proximity to other bolted structures were noted; the span lengths and full profiles were recorded and the presence of thermal insulation and multiple attachments were noted.
For all other supports, a visual check of all these attributes was made and any deviation was measured, if appropriate, and recorded. The as-built information obtained in the walkdown was used in the evaluations.
Furthermore, all instances of tray overfill, base plate bolt hole oversize or attachment eccentricities and bolt hole shear cone interference were evaluated.
A walkdown in the annulus area was perfomed by the staff and consultants.
Tube attachment eccentricities and ground wire attachments were observed for supports Mk9b and Mk15, respectively, but no real deficiencies were noted. The supports and trays appeared adequately constructed and fimly anchored.
An audit of the calculations for the shield wall mounted supports was conducted. The calculations were retained in a single file identified by calculation No. CSG-86-009.
In the file were copies of all the analyses perfomed for the subject supports covering a period from April 1986 to present. This included the latest GTSTRUDL and BASEPLT II computer analyses for each generic and selected special support, the numerical development of bounding load cases, the assessment of all anchor bolt shear cone interferences and the evaluations perfomed to bound the conditions of base plate eccentricity noted in the walkdowns. In general, the calculations were comolete and under-standable. However, in those instances where revisions were made to earlier calculations, the earlier calculations were not labeled superceded, making the audit activity difficult. The audited calculations have demonstrated that each cable tray support attached to the shield wall had sufficient capacity to meet the interim criteria for the SSE load condition.
I I (c) All Other Cable Tray Supports There are 7,900 cable tray supports in Categorv I structures (excluding Steel Containment Buildino and the Reactor Building Shield Wall). Most of these are in the Auxiliary Ruilding (1,700) and the Control Building (8501 The selection of the worst case supports in the auxiliary building, documented in B25 860913 825, was reviewed by the staff and consultants. The selection process started with a review of the drawings which contained support details.
After considering factors to include the number of cable trays for each support, span length and floor elevation, ten worst case support configur-ations were identified. Each configuration may represent a group of specific supports with different geometries or it may represent a unique situation.
For those configurations that represent a group of supports, the following three criteria were used to select the specific worst cases: (11 supports having largest span lengths and largest weights; (2) maximum weight with the length selected for the first mode period at peak response of the spectrum; (31 maximum length with weight selected for first mode period at peak response of the spectrum. After review by the TVA central technical group of the above cases, five additional worst case supports were added.
The same selection process was applied to cable tray supports in the other buildings. Thus, altogether, TVA has considered thirty original worst case supports and five additional worst case supports.
It is believed that the TVA has used gond engineering fudgment in its selection of the worst cases and the approach used is.iudged as acceptable for restart purposes.
Walkdowns were performed for each of the " worst case" supports by TVA to evaluate:
1.
Weight in the trays. Profiles were measured for trays which were more than 75% full and weights calculated.
.g 11 1
2.
Any additional attachment to the support. Sketches were made detailing the attachment.
3.
Cases where the trav support is not mounted concentrically on the baseplate.
4.
Whether the support is Fire Protected.
5.
Any violation of TVA's Construction Specification G-2 (e.g., close spacing of adjacent anchorages resulting in overlapping of shear cones or anchor plates placed near an edge of a concrete member).
6.
Other unusual details.
Reports on the results of the walkdown were prepared and signed by the preparer, checkers, and a Q/C staff. Results of the walkdown were reviewed during the September 29 audit by the staff and consultants. The data on the TVA walkdown sheets were verified to be accurate with one omission. An interference was noted for support Mark 31. A 6" conduit was noted to be in a location close to a bracket of this support. Seismic induced motion could be expected to cause the bracket to imDact the conduit. All accessible supports in the Reactor Building (inside containment) were also inspected. The inspection verified the TVA walkdown findings which included cases of supports not installed concentrically on baseplates and cracked concrete under baseplates.
No additional deviations were observed by the audit team. TVA is evaluating these discrepancies and will report to the staff upon completion of their work.
TVA prepared a GTSTRil0L model of each of the worst case supports based on the 1
drawings and the results of the walkdown. The supoorts were modeled as beam elements. The mass of the cable trays were lumped on the appropriata brackets with the tray masses distributed equally to the adjacent supports. A response spectrum analysis was performed with the 7% damped spectrum used as input.
The model used for the support marked Section p-P was reviewed during the audit by the staff and consultants and was found to be acceptable.
f I
i
.1
-1?-
5 l o
~
l The design calculations of the support Mk-1G, identified by TVA as E?S 8608?7 were also reviewed by the staff and consultants. The analyses were cdrried out l,
by tiTSTRllDL. The modeling of the frame and the input of loadings pere found..to be correct and adequate. The seismic responses of the cable tray supports were obtained by dynamic modal analysas. For the checking of the frame stresses, pseudo static seismic responses were utilized. Loads were combined based on the sum of the dead load and the absolute sum of the most severe of the horizontal and the vertical SSE loads. Stress evaluations were then made for main members, brackets, joints, and anchorage.
Several issues were raised by the staff during the July 21-24 audit. ThEse have been addressed by TVA and evaluated by the staff and consultant's during the September 29 audit. A few locations were identified where the spa'n of,the trays was more than 8 feet. These conditions occurred where the trays are inclined at a 45* angle. The horizontal projection of the span is less than 8 feet, but the inclined span is greater. TVA has performed load tes'ts #(TVA B46860311003) that evaluates a cable tray in this configuration. The tray was found to have a capacity of 140 pounds /ft. which indicates a factor of safety of more than 3 over the full tray design loading of 45 lbs/ft and, therefore, is acceptable.
Several groups of cables were observed to cascade vertically from a conduit or
,t i
one tray to another in the control butiding. TVA has perfomed tests at Wyle iy; i
Labs to demonstrate that the cascading cables can withstand the SSE scismic
- j induced loading. The tests have been evaluated by an independent TVA consultant l
and are currently being reviewed by TVA staff. The TVA consultant concludes that the cables are not overstressed because they are not stressed beyoad their tension capacities. TVA should provide NRC their final evaluation of Wyle test results for review and approval.
i
.I l
r
.t Pending the resolution of the above noted item, the staff and consultants concluded that the program conducted by TVA for these cable trav brackets and supports qualification was adequate and acceptable for restart purposes.
4.
ANCHORAr,E IN CONCRETE This discussion applies to supports which are anchored to concrete by means of baseplates, anchor bolts and embedded plates.
Several concerns relating to safety factors and methods of analysis were identifi.ed at the July 21-24 meetings.- These have been addressed by TVA and were discussed during the September 29 audit. The concerns are discussed below:
TVA proposed that self-drilling (SSD) and wedge (WR) type expansion bolts used for base plate anchorages be designed for a safety factor of 2.0 under Load Combination of SSE plus DBA. The TVA staff indicated that this would be an interim criterion.
In the Phase II design qualification work, the minimum safety factors for SSD and WB would be upgraded to 2.8 and 2.5, respectively.
In defense of this proposal, the TVA staff indicated that during the implementation of IE Bulletin 70-0?, the NRC staff had accepted a safety factor of 2.0 for both types of expansion bolts on an interim basis. The same logic can apply in case of the interim evaluation of the expansion anchor bolts.at SQN for restart purposes.
The audit team discussed TVA's proposal and decided that the proposed safety factors were not adequate and that TVA should use, as a minimum, the original FSAR design criteria requiring 2.5 for WR and 2.8 for SSD as safety factors for the interim period. This requirement should be ac'ded to TVA Interim Acceptance Criteria. For the long tenn effort TVA should determine the actual safety factors and evaluate them against the requirements of IE Bulletin 79-02.
i
't
_14 5
Some of the conservative assumptions used in TVA's standard design practice tend to support a view that the actual safety factors against the pull-out of expansion bolts will, in general, be higher than those calculated.
For example, TVA uses the expansion bolt capacities based on 3000 psi concrete, whereas, the concrete strength data at 90 days indicate that the actual strength of the concrete could be in excess of 6000 psi. This could increase the expansion bolt capacities by about 40%. Another example of the conservatism is that in normal installation, TVA procedures requires preload of bolts to twice the design load. A minimum of 25% of the bolts are tested for slippage at that preload level Any slippage as indicated by a drop in load of the load indicator was regarded as a failure. This requirement is more stringent than the accepted industry practice of allowing some slippage. These conservative design and installation practices fom the basis for the staff acceptance of the above noted safety factors for the restart purposes.
TVA has comitted to the interim requirement set by the staff. This is s
acceptable to the staff.
5.
RASEPLATE ANALYSIS Frame analyses were perfonned, as discussed in previous sections to evaluate the distribution of forces throughout the cable tray supports. The cable tray mass is distributed evenly between ad.iacent suoports. Actual walk-downs were performed to evaluate overloaded trays. Trays that were less than full were considered to be full. The SSE loading was used as an input and two alternate types of analysis were performed. The first type of analyses performed were response spectrum analyses.
If there were no modes less than 33 cps, a seismic load equivalent to the tray and support mass times the zero period acceleration (ZPA) was applied to the support. The second type of analysis was a static analysis with a load eoual to the tray and support nass multiplied 1.5 times the peak spectral acceleration. The deviation between the center of cable trays' mass points and brace connection.ioints had not been considered by TVA
.t
-Is-at the time of staff audit. TVA will consider this in calculations to be developed.
It is not expected that this will lead to serious changes in response forces and will be treated as a confirmatory open item. In other respects the current analyses used by TVA are acceptable.
The loads from the frame analysis are used to evaluate the adequacy of the support members and base plates. Standard engineering methods are used to evaluate stresses in members and are considered acceptable by the staff and consultants. The BASEPLATE II computer program is used to evaluate stresses in the baseplate and bolts and bearing stresses in the concrete. BASEPLATE II is a preprocessor code that generates input data for an ANSYS solution.
Plate finite elements are used to model the baseplate and elastic springs model the anchor bolts. The concrete is modeled with an elastic spring in series with a gap element so that the concrete acts in compression but not in tension. TVA has perfomed sensitivity studies to develop criteria for the finite element modeling of the baseplate. The modeling and analysis of the baseplate are acceptable.
6.
CONCRETE Members of the audit team reviewed available data related to the concrete strength at SQN. The staff has raised several questions on the concrete issue. TVA is still evaluating its concrete records and will respond to the staff questions in the near future. The staff will issue a separate safety evaluation of the concrete issue at SON.
7.
CONFIRMATORY ITEMS The staff and its consultants identified the following confirmatory items requiring resolutinn by the TVA:
d (1) An unused bolt hole was observed in the main tube member of support MK 11d in the annulus.
It should be verified that this support is adequate.
(2) The 1/8" fillet welds used throughout the supports to the shield wall do not satisfy the requirement of AWS DI.1-85 Section 10.5.31. The adequacy of these welds is to be investigated based on data to be obtained in a scheduled TVA test proaram.
(3) The spring constant for self drilling bolts was used for the BASEPLT II analyses. Most of the bolts are wedge bolts. The BASEPLT Il analyses must be revised to reflect the proper bolt type.
(4) An error in one of the moment components for support MK 11d in the annulus was found. The evaluation of this suppnrt should be revised.
(5) An interference between a conduit and support MK 31 in the Auxiliary Building was observed during the audit. TVA must evaluate the signif-icance of this condition.
(6) The evaluation of all " worst / case" supports in the Auxiliary Building must be completed and documented.
(7) The Interim Acceptance Criteria for anchor bolts should be based on safety factors of 2.5 and 2.8 for the wedge bolts and self-drillina bolts respectively. TVA should fully document its implementation of these safety factor related criteria.
(8) TVA is to develop and submit for staff acceptance calculations which demonstrate that the eccentricity of the cable tray mass will not adversely affect the qualification of supports.
,6 (9) TVA is to provide their final evaluation report addressing the desion adequacy of cascading cables tested at the Wyle Labs for staff review and acceptance.
(101 TVA will complete all required cable tray supports modifications, as determined by the TVA evaluations against the staff approved Interin Acceptance Criteria, prior to restart.
8.
CONCLUSION Based upon the review of the material provided by TVA, field audit of TVA design documents and plant walkdown, the staff concludes that pending the satisfactory resolution of the confirmatorv items, the Interim Acceptance Criteria proposed by TVA for Sequoyah restart and modified by this staff safety evaluation report, are acceptable.
TVA should formally inform the staff of the expected date for submitting the documented resolution of the confirmatory items.
The staff review of Seauoyah concrete strength issue is continuing and a separate safety evaluation report will be published when the issue is resolved.
.