ML19296D933
| ML19296D933 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 11/21/1979 |
| From: | Broehl D PORTLAND GENERAL ELECTRIC CO. |
| To: | Engelken R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| References | |
| IEB-79-02, IEB-79-2, TAC-12369, NUDOCS 8003130579 | |
| Download: ML19296D933 (26) | |
Text
{{#Wiki_filter:TI G Y., Portland General ElectricCompany V Ccr.a d J BrceN Ass:smt Vce Pwa mt /TW ms- - e,3 November 21, 1979 -~ f ff ' s, [\\ [c Trojan Nuclear Plant s ,[ Docket 50-344 License NPF-1 L-Mr. R. II. Engelken, Director U. S. Nuclear Regulatory Commission Region V Suite 202, Walnut Creek Plaza 1990 N. California Blvd. Walnut Creek, CA 94596
Dear Sir:
The attached report completes PGE's response to IE Bulletin 79-02, Revision 1 dated June 21, 1979, regarding Seismic Category I pipe support base plates which use expansion bolts. This report con-sists of the original response dated July 13, 1979 with new infor-mation identified by a revision number in the margin. Repairs have been completed on five of the eleven pipe supports with anchor bolt factors of safety less than two. Repairs are in progress on the other six and will be completed before power operations resume. Eight of the 69 supports with anchor bolt factors of safety less than five have been repaired, and work on the remaining 61 will be completed during the next refueling outage. The disposition of the above supports is in accordance with the requirements of IE Bulletin 79-02, Revision 1, Supple-ment 1. A preliminary count of safety-related pipe supports shows that 596 of 1375 or 43 percent of the supports for three-inch diameter and larger piping are mounted on masonry walls. Based on approximately (79 8003130 m .e nmn euwana OTacn 924 ]f }[f /
Portland General ElectricCompany !!r. R. II. Engelke n November 21, 1979 Page two five anchor bolts average per base plate, we have about 3300 expansion anchors in masonry walls. More detailed information will be given in our response to Revision 2 of the Bulletin. PGE's response to Revision 2 of the Bulletin dated November 8, 1979 will be provided in a later response. Sincerely, / DJB/TEB/4mg6A21 Attachment c: Lynn Frank, Director State of Oregon Department of Energy Director, Office of I&E Division of Nuclear Operations Inspections A. Schwenrer, Chief Operating Reactors Branch #1 Division of Operating Reactors U. S. Nuclear Regulatory Commission
s Revision 1 i 11/21/79 TROJAN NUCLEAR PLANT PIPE SUPPORT BASE PLATES USING CONCRETE EXPANSION ANCHOR BOLTS I. INTRODUCTION NRC IE Bulletin No. 79-02 (Revision No.1) dated June 21, 1979, requires all licensees of nuclear power plants to review the design and instal-lation of concrete expansion anchor bolts used in pipe support base plates in systems defined as Seismic Category I by the NRC Regulatory Guide 1.29, " Seismic Design Classification", Revision 1, dated August, 1973. Pursuant to the Bulletin, extensive analytical reevaluations have been performed, and field inspection and testing has been conducted for Seismic Category I pipe support base plates using concrete expansion anchor bolts in order to confirm Seismic Category I piping system operability. This report provides results of the analyses and field testing, and our response to the Bulletin. II. RESPONSE TO ACTION ITEMS Item 1 Verify that pipe support bare plate flexibility was accounted for in the calculation of anchor bolt loads. In lieu of supporting analysis justifying the assumption of rigidity, the base plates should be considered flexible if the unstiffened distance between the member welded to the plate and the edge of the base plate is greater than twice the thickness of the plate. It is recognized that this criterion is conservative. Less conservative acceptance criteria must be justified and the justification submitted as part of the response to the Bulletin. If the base plate is. determined to be flexible, then recalculate the bolt loads using an' appropriate analysis. If possible, this is to be done prior to testing of anchor bolts. These calculated bolt loads are referred to hereafter as the bolt design loads. A description of the analytical model used to verify that pipe support base plate flexibility is accounted for in the calculation of anchor bolt loads is to be submitted with your response to the Bulletin. Response to Item 1 Original Design The original design and installation of Seismic Category I pipe supports for the Trojan plant were governed by the following documents: ANSI B31.1 huerican National Standard Code for Pressure Piping - Power Piping ANSI B31.7 Anerican National Standard Code for Pressure
s a Trojan Nuclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 2 Piping - Nuclear Piping ANSI B31.7b 1970 Addenda to Nuclear Piping B31.7-1969 ASME Section IX Boiler and Pressure Vessel Code Welder Qualification ASME Standards Parts 1, 3, 4 and 31 Bergen-Patterson Pipe Support Corp. Catalog No. 66 Pipe support design loads were determined from piping stress analy-ses performed in accordance with the requirements outlined in Trojan FSAR Section 3.7.3.3. Design of seismic supports and restraints meets the requirements outlined in Trojan FSAR Section 3.7.3.3.11. The original design of Seismic Category I pipe support base plates, where concrete expansion anchors were used, was based on a simplified method of analysis using rational assumptions and conservative allowable load criteria. The distribution of loading on anchor bolts was calculated on the basis of a rigid base plate with pure tension and shear loads distributed equally to the bolts. The effects of bending moment loads were treated as follows: 1. Moments jn planes perpendicular to the plate were resolved into tension forces on the bolts by considering the bolt rows to provide the resisting couple. 2. Moments in the plane of the plate were resolved into shear forces on the bolts by considering the bolt pattern to provide the resisting couple. Shear-tension interaction relationships, as presently developed, were not considered in the original design. Concrete expansion anchors used in the original design were Phillips Redhead self-drilling shell type anchors. Allowable loads used were based on a factor of safety of ten against the manufacturer's ultimate load values determined from tests. The standard used for minimum spacing between the expansion anchors Was ten times the nominal bolt diameter. When not otherwise specified, 6" was used as the minimum distance from anchor bolt centerline to edge of concrete.
s irojan Nuclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 3 Reevaluation Analyses For each safety-related piping system involving pipe three inch /_h and larger, all pipe anchor and support base plates using concrete expansion bolts were reanalyzed to account for plate flexibility, bolt stiffness, shear-tension interaction, minimum edge distance and proper bolt spacing. Depending on the com-plexity of the individual base plate configuration, one of the following methods of analysis was used to determine the bolt forces: 1. A quasi-analytical method, developed by Bechtel, was used for base plates with eight bolts or less. A review of typical base plates used in supporting the piping systems indicate that the majority of them ware anchored either by 4, 6, or 8 bolts. The base plates typically vary from 3/8" to 1" in thickness and are generally not stiffened. For these types of base plates, an analytical formulation has been developed which treats the plate as a beam on multiple spring supports subjected to moments and forces in three orthogonal directions. Based on analytical considerations as well as the results of a number of representative finite element analyses of base plates (using the "ANSYS" code), certain empirical factors were introduced in the simplified beam model to account for the effect of the concrete foundation and the two-way action of load transfer in a plate. These factors essentially provided a manner for introducing the interaction effect of parametric variables (such as plate dimensions, attachment sizes, bolt spacings and stiffnesses) on the distribution of external loads to the bolts. The results of a number of other case studies indicate excellent correlation between the results of the quasi-analytical method and those of the finite element method (using the "ANSYS" Code). The effect of plate flexibility has been explicitly consid-ered in the quasi-analytical formulation described above. The effect of prying action on the anchor bolts was deter-mined not to be critical for the following reasons: a. Where the anchorage system capacity is governed by the concrete shear cone, the prying action would result in an application of an external compressive load in the cone and would not affect the anchorage capacity.
~ s Trojan Nuclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 4 b. Where the bolt tension determines the anchorage capacity, the additional load carried by the bolt, due to its stiffness, decreases with increasing load. At higher loads the bolt elongation will be such that the corners of the base plate will tend to lift off and the prying action will be relieved. This phenomena has been found to occur when the bolt stiffness in the finite element analysis was varied from a high to a low value, to represent the initial stiffness and that beyond the allowable design load, respectively. A computer program for the analytical technique described above has been implemented for determining the bolt leads for routine applications. The program requires plate dimensions, number of bolts, bolt size, bolt spacing, bolt stiffness, the applied forces and the allowable bolt shear and tension loads as inputs. Allowable loads for a given bolt are determined based on the concrete edge distance, bolt spacing, ultimate capacity for the bolt based on tests, considering material strengths, and a design safety factor. The program computes the bolt forces and calculates a shear-tension interaction value based on the following interaction formula: T 2 S 2 + < 1.0 TA SA ~ Whefe T and S are the calculated tensile and shear forces and TA nd SA are the respective allowable values based on a factor of safety of five for shell-type anchors. Typically, no credit is taken for shear resistance by base plate friction due to anchor bolt prel oad. The above relationship is an appropriate representation of shear-tension effects demonstrated by test data. 2. For special cases where the design of the support did not lend itself to the foregoing method, the finite element method using the "ANSYS" code and/or other standard engineering analytical techniques with conservative assumptions were employed in the analysis.
e a Trojan fluclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 5 Table 1-1 summarizes results of the reevaluation analyses, as de-scribed above, of pipe support base plates for all safety-related piping systems (3" and larger). In Table 1-1, the numbers of pipe supports with expansion anchor bolts in the various ranges of shear-tension interaction ratios are shown. Initial evaluations of expansion bolt loads, pursuant to the Bulletin, were made using allowable load values listed in Table 1-2. These loads are based on the following: 1. Concrete: The allowable values are the manufacturer's ultimate values given in Phillips Red Head Catalogue F-1000 divided by a factor of safety of 5. 2. Masonry: The allowable values are based on the limited data available from tests performed at the Fast Flux Test Facility, Hanford, Washingtan (FFTF). These lower bound values were used irrespective of considerations of test data applicability (i.e., masonry strength, representative bolt sizes, etc.) for the purpose of eliminating from further analysis those supports which are only lightly loaded. Allowable anchor bolt loads which are considered to be appropriate d for Trojan are discussed in the following section. Detailed Review of Anchor Bolts With Interaction Ratios Greater Than One The results of the base plate reevaluation analysis for interaction ratios greater than one were reviewed pursuant to IE Bulletin 79-02, Revision 1, Supplement 1. The results of this detailed review are shown in Table 1-3. Allowable bolt loads for this review were based on the following: 1. Concrete: The allowable values are the manufacturer's ultimate values given in Phillips Red Head Catalegue F-1000 divided by a factor of safety of 5. 2. Masonry: Values for heavy masonry (l) are based on a mat rial correction factor of 0.6 applied to the concrete values and a further reduction factor of 0.71 for standard weight masonry The 0.6 factor is based on International Conference of Cuilding Of ficials (ICB0) Standard flo.1372 (February 1975) which lists (1) Heavy masonry is high density concrete block with f' } 4000 psi. Standard masonry is regular weight concrete block with Q } 2000 psi.
i Trojan Nuclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 6 allowable values for Phillips self-drilling anchor bolts in both concrete and masonry. Allowable tension values for 3500 psi concrete were compared to values listed for masonry. Ratios obtained were greater than 0.6 for all sizes of bolts. Since the majority of masonry at Trojan to which safety-related pipe supports are attached is high density block, this ratio is conservative. To ensure that allowable values for standard masonry were also conservative, a further strength reduction factor was added. This further reduction for standard masonry is based on the square root of the design strength ratios or V2000/4000 = 0.71. The allowable values used for the detailed review are also given in Table 1-2. The results of the FFTF anchor bolt testing in masonry were reviewed and found to be not applicable to Trojan for the following reasons: 1. The materials used in the FFTF test differed from those at Trojan. The block used in the FFTF test had a strength of fs = <2000 psi (minimum) whereas the high density block at Trojan had f6 = 4100 psi and the standard weight fs = 2700 psi. The mortar fcm was 3500 psi at Trojan compared to 2500 psi at FFTF, and the core fill ff was >6000 psi compared to 2000 psi. These values are based on actual testing during Trojan construction. 2. Only one size (3/8") Phillips self-drilling anchor bolt was tested. d 3. The bolts were installed with a twist bit rather than using the anchor itself to drill the hole. 4. The location for 25% of the tests (i.e., within 1" of the block edge) is not representative of any of the anchor bolt installations at Trojan. The FFTF test data is not applicable to Trojan, and no examples of thorough tests for self-drilling anchor bolts in masonry were found. It is judged that the values for allowable anchor bolt loads in masonry are suf ficiently conservative. In order to confinn these values, a test program at Trojan in the near future will be performed. Results of this demonstration testing will be made available as a supplement to this report.
Page 7 TABLE 1-1
SUMMARY
OF REEVALU/TI0tt ANALYSES OF PIPE SUPPORT BASE PLATES USING EXPANSION ANCHOR BOLTS Seismic Category 1 litmber of Supports with Anchor Bolts in Interaction Retio Range Piping System 0<0.25l 0.25<0.50! 0.50<0./5 f 0.75<1.0 1.0 <1.2 5 f 1.2 c < 1. b t 1.5 <' l 27 l 3 3 0 1 0 0 L 1. Reactor Coolant 2. Chemical Volume Control 195 22 16 10 1 5 15 3. Residual Heat Removal 99 18 8 7 1 2 11 4. Safety Injection 98 14 13 11 3 2 9 l 5. Containment Spray 38 6 4 3 1 0 5 6. Main Steam 44 4 3 0 1 2 5 Steam Generator A 7. Blowdown (2) 0 0 0 0 0 0 11 8. Condensate & Feedwater 79 8 2 3 0 0 2 9. Component Cooling Water 176 19 10 11 1 1 8
- 10. Service Water 173 33 9
3 1 2 19
- 11. Diesel Fuel (2) 30 6
3 3 0 1 2
- 12. Miscellaneous (1) 97 23 8
3 0 0 7 TOTAL 1056 156 79 54 10 15 84 Total flo. Supports Analyzed = 1454 (1) Miscellaneous systems include: Radioactive Gaseous Waste Dirty Radioactive Waste Containment Vent Monitoring Miscellaneous Gas Supply Spent Fuel Pool Chilled Water Primary Makeup Water Solid Radioactive Waste Clean Radioactive Waste Containment Condensate Process Steam (2) Typically small diameter piping (less than 3" diameter)
Page 8 TABLE 1-2 ALLO'.lABLE PHILLIPS RED HEAD SELF-DRILLING ANCHOR BOLT LOADS (ALL LOADS ARE IN KIPS) I f I. Concrete Size Tension Shear 3/8 1.13 0.67 1/2 1.70 1.34 5/8 2.34 2.38 3/4 3.24 3.24 7/8 3.57 3.69 t II. Masonry - Used in initial Bechtel Reevaluation Analyses zik Size Tension Shear 3/8 0.260 0.260 1/2 0.385 0.385 5/8 0.529 0.529 3/4 0.784 0.784 7/8 0.808 0 808 III. Masonry - Used in Detailed Review Analyses Heavy Masonry (1) Standard Masonry (2) Size Tension Shear Tension Shear 3/8 0.68 0.40 0.48 0.29 1/2 1.02 0.80 0.72 0.57 5/8 1.40 1.43 1.00 1.01 3/4 1.94 1.94 1.38 1.38 7/8 2.14 2.21 1.52 1.57 (1) Heavy Masonry fs 2 4000 psi. (2) Standard Masonry f5 2 2000 psi.
Page 9 TABLE 1-3 DETAILED ANALYSIS RESULTS SUPPORTS WITH ANCHOR BOLT INTERACTION RATIOS GREATER THAN 1 No. of Supports with Mounting Factor of Safety (3),
System Location (1)
<2
>2,
_..t_. n -. e _ n..- u=s=$=v a-
- =1=s=t=t=L.
t _t-
.$ _ t.. e n.._
_- gg - 30 4n_.30__40 re rlp ---9 0---- 110-- 12 0
~-
-~
_ 9 Z -==-!NDMBER fCE-DEFECTSi
=t-E- =-i r s E- ~_N. : _m- __i FIGU RE~- 4 A __
..r
- .=. _-
...-,_ j. _.. _... _-1_ _,_-j _,_3 ') ~~ m-
- s
Trojan Nuclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 21 6. Evaluation of thread engagement and base plate hole enlargement data. A. Thread Engagement Minimum thread engagement lengths necessary to develop the full manufacturer's ultimate strength in concrete for the expansion anchors were calculated using ANSI Bl.1, Appendix A, as a basis. For bolts identified in the field test program as having less than minimum thread engagement, values for ultimate strength of the bolt threads were calculated for various lengths of thread engagement. These values are shown in Table 4-3. The field data was reviewed for anchor bolts with less than the minimum bolt thread engagement lengths necessary to develop full anchor strength. There were 380 bolts on 190 supports with less than minimum engagement. Of these 380 bolts, 208 were checked against calculated bolt loads. These 208 were either on large pipe supports or on small pipe supports (< 2" diameter) which were computer analyzed. /_11 The remainder were on small pipe supports which are typically very conservatively designed since detailed computer analysis was not performed. The results of this review showed that there was a factor of safety less than 5 on only two bolts. The first hol a factor of safety less than 5 based on the analysis results and will be repaired in accordance with Supplement No. I to IE Bulletin 79-02. The second was a computer analyzed small pipe support. The bolt factor of safety was reduced to 4.56. The support will be repaired by repl acement of the existing bolt. The results of this review show that only 2 of 208 bolts, or less than 1%, were af fected by reduced thread engagement lengths. B. Bolt Hole Enlargement The anchor bolt inspection identified 181 bolts on 68 supports which had base plate holes enlarged beyond the 1/8-inch oversize specified on the design detail sheets. The reason for enlarging the holes during construction was to compensate for slightly offset placement of the anchor shells. The enlargement was generally done with a cutting torch.
<J Trojan Nuclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 22 The enlargement of individual bolt holes does not degrade the support unless the bolts are lef t loose or the hole is enlarged enough to allow the bolt heaa to pass through the plate. No case was found where the hole was enlarged enough to allow the bolt head and washer to pass through. In reviewing the data for those bolts with enlarged holes, there was no condition where the bolt was not tightened af ter installation except for failed anchors. The individual bolt preload serves /$i to keep the plate tight against the the anchor shell or counting surface and p> ' vents excessive shear displacement. Therefore, the effects of reduced thread engagement and bolt hole enlargecent as identified in the test program are not significant with regard to the structured adequacy of the supports.
.. b., Page 23 i TABLE 4-3 PULLOUT FORCE VS BOLT SIZE AND EMBEDMENT LENGTH (FORCES ARE IN KIPS) Ultimate Tensile Lf(1) Strength of Ex-Bolt 1/8 3/16 1/4 6/16 3/8 7/16 1/2 pansion Anchor (2) Dia 3/8 2.16 3.24 4.32 4.65 4.65 4.65 4.65 4.65 1/2 2.92 4.38 5.84 7.30 8.50 8.50 8.50 8.50 5/8 3.74 5.61 7.49 9.36 11.2 11.7 11.7 11.7 3/4 4.55 6.83 9.10 11.4 13.7 15.9 16.2 16.2 2h 7/8 ' 5.37 8.06 10.7 13.4 16.1 17.9 17.9 17.9 (1)Lf = Thread engagement length (2)ln concrete f6 = 3500 psi.
..J., Trojan fluclear Plant Pipe Support Base Plates Using Concrete Expansion Anchor Bolts Page 24 7. Evaluation of Defective Anchors identified. In accordance with plant procedures, all Seismic Category 1 pipe support base plate expansion anchor deficiencies identified during the field inspection and testing program were documented in nonconformance reports. All nonconforming base plate anchor 1 installations have been dispositioned as acceptable. All but three nonconfonaing anchor installations have been repaired. Analyses of the three base plate anchor deficiencies that have not been repaired show that the pipe supports still have adequate factors of safety. However, the conditions associated with these three deficiencies are scheduled for repair as soon as practicable. In summary, no defective anchor condition identified by field inspection and testing of pipe support base plates affected the operability of any Seismic Category 1 piping system. TEB/ rcn/3j f 55.35B1 .}}