ML17059B664

Lead Expansion Anchor Load Capacity in Reactor Buildings at Savannah River Site, Rev 0

ML17059B664
Person / Time
Site:	Nine Mile Point
Issue date:	08/15/1989
From:	BRYANT E L WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML17059B657	List: ML17059B655 ML17059B658 ML17059B659 ML17059B660 ML17059B661 ML17059B662 ML17059B663 ML17059B664 ML17059B665 ML17059B666 ML17059B667 ML17059B668 ML17059B669 ML18040A314 ML18041A055
References
C.0SQUGANCHOR-R, C0.0SQUGANCHOR-R00, CO.0SQUGANCHOR, CO.SQUGANCHOR, RTR-2661, NUDOCS 9708070135
Download: ML17059B664 (50)

Text

Westinghouse Savannah River Company ATTACHMENT I A@A mt CALO NO REVISION PAGE NO P.O.Ai.28802 RTR-2661 August 15, 1989 TO: S.D.Curry FROM: E.L.Bryant,$W LEAD EXPAN I AN R L AD IT~IN REA T B ILDING AT THE VANN H VER I INTR D Tl N Walkdowns to determine the seismic adequacy of selected safety systems have been performed for L and K reactors utilizing the Unresolved Safety Issue (USI)A-46 methodology sponsored by the Seismic Qualification Utility Group (SQUG).Actual experience data from earthquakes, test data compiled by the Electric Power Research Institute (EPRI), and other data for certain classes of nuclear plant equipment make up the SQUG data base.Anchorage is also addressed in the USI A-46 methodology.

However, lead expansion anchors are not included.Some safety related equipment in the Savannah River Site (SRS)reactor buildings is supported by lead expansion anchors.Design load capacity was based on the allowable loads listed in SRS Standard B8M which was intended to provided a factor of safety of 4.However, walkdowns have revealed that some lead anchors are not installed per B8M, and therefore the actual load capacity of these anchors may be less than indicated in the B8M.A testing program(SP-2449) was developed and 107 abandoned anchors in K, L, P and R reactor buildings were tested to determine shear and tension load capacities.

This RTR documents the results of this testing program and provides a method for confirming adequacy of the lead shell type anchor.Test results show that the tension load capacity of lead" 9708070i35 97073i PDR ADOCK 05000220 P PDR OSR25 82'Wta 89>~~q oh')

~a RTR-2661 Page 2 August 15, 1989 lTTACHMENT C CALC NO o REVISION~<IiF NOtype anchors at failure is below the design failure loads (i.e., four times the B8M).Therefore, the new lower allowable loads listed in Table 1 should be used for lead shell type anchors.If the calculated load on the anchor is greater than the Table 1 load, the anchor still may be adequate providing it is load tested to prove that it.is capable of carrying the seismic demand load.A regression analysis correlation between torque and load has been developed by SRL based on test measurements and can be used for proof load testing those lead type shell expansion anchors that exceeds the new lower allowable (Appendix A).The anchor bolt should be torqued to induce a tension load on the anchor which confirms its capacity to carry its calculated seismic demand load or the BSM design allowable load.This approach may result in a lower factor of safety.The factor of safety of 4 was intended to provide a high level of confidence that the anchor will support the design load.Because the anchors will be load tested to the seismic demand or design allowable load, a lower factor of safety is considered acceptable.

Tests also show that the BSM values for shear loads are acceptable

.Table 1 Anchor Size, inches'urrent Tension Allowable Load (BSM)700 1200 2000 3000 5500 New Recommended Allowable.Load Based On Tests 600 870 970 1280 3160 Proof Torque For BSM Design Allowable, ft-Ib 24 50 37 69*Applicable only for anchors that have passed the tightness torque per RTR-2582

RTR-2661 Page 3 August 15, 1989 IgfEFtiF<REWSipg PA()E NO ANCE o~1.The mean measured tension failure load of lead type expansion anchors is less than 4 times BSM.2.The new lower tension load capacities of lead type expansion anchors that are listed in Table 1 should be used for walkdowns.

3.The tension load on anchors to be seismically qualified should be calculated and shown to be below the Table 1 values.4.All lead type anchors should be proof loaded if the calculated seismic demand load exceeds the load capacity isted in Table 1 or if B8M allowables are assumed.5.The torque necessary to induce a tension load on installed anchors should be determined from the correlation given in Figures 8-11 at the 95%confidence level.6.BSM shear load capacities are adequate.7.Further use of lead type expansion anchors at SRS should be discontinued.

Lead type shell expansion anchors have been used in the reactor buildings to support some safety-related equipment.

As part of the seismic upgrade work, the adequacy of these anchors was evaluated using the USI A-46 methodology.

This methodology is based on application of actual earthquake experience data, test data compiled by EPRI, and other data and is used to evaluate the seismic adequacy of certain classes of nuclear plant equipment.

Anchorage is addressed by SQVG but specific data on lead type expansion anchorage is not available.

Allowable loads specified in BSM were originally used during the walkdowns for the purpose of anchorage evaluation.

The walkdowns revealed that many of the lead anchors were not installed per BSM and the actual load capacity of lead type expansion anchors was questionable.

RTR-2661 Page 4 August 15, 1989 CALC NO REVISION PARE NO AdcHoR A total of 107 Jead type anchors covering 5 anchor sizes were load tested to failure in tension and 36 anchors, 3 sizes, were tested to failure in shear per SP 2449.Based on the test results, new lower allowable loads for tension were determined and-are listed in Table 2.The shear failure loads exceeded those predicted by B8M and therefore, the B8M values were retained.The lead type expansion anchors exhibited poor load performance in tension as evidenced by a wide variation in failure loads.Because of this poor tension performance, it is concluded that further use of lead type expansion anchors should be discontinued at SRS.Tests were performed under Special Procedure 2449 on abandoned lead type expansion anchors in SRS Reactor Areas.The following conditions were prerequisites to testing:-Anchor did not pull out or become loose when subjected to a tightness torque of approximately 20%of the installation torque.-Perpendicularity of the anchor with the wall was less than 5 degrees offset.-Anchor location was greater than 10 diameters from the next loaded anchor or a concrete edge.Because of the limited number of available anchors, some anchors closer than 10 diameters to the next abandoned anchor were used, The test was considered acceptable if concrete spalling did not protrude into the area of the adjacent anchor.Also, some expansion anchors were tested if part of the threaded shell was above the concrete surface but all of the lead ring that secures the anchor to the concrete was below the face of the concrete.In this case, a metal spacer was used to provide a gap between the base plate and anchor during the tightness torque.TET P The test setup is shown in Figure 1.A hydraulic ram was used to apply a load to the anchor for tension tests.A stud rod with a

RTR-2661.Page 5 August 15, 1989 N CH os double nut on one end was inserted through the bore of the ram and screwed into tAe anchor about one diameter.Then the ram was used to apply a load approximately 325 lbs to the anchor.All subsequent loads were applied to the anchor by torquing the aut and measuring the resisting load on the hydraulic ram that is transmitted.to the anchor.Both the torque and the load were recorded in step increments until failure.For shear tests, a special shear plate assembly was made (Figure 1).A bolt was inserted through the shear plate assembly into the anchor.The hydraulic ram was then placed under the shear plate assembly to push the shear plate upward.The resisting pressure at failure was recorded.Both tension and shear load capacities (Table 2)for lead type expansion anchors were determined.

The mean measured tension loads at failure were all below the expected failure loads (i.e., B8M x 4).Therefore, new lower design allowable loads are recommended which includes a factor of safety of 4.A tightness torque was applied to screen out poorly installed anchors during walkdowns and was also applied during these tests.Anchors that failed the tightness torque were not load tested.Thus, the new lower recommended values are applicable to anchors that have been screened (i.e., passed the tightness torque).Additionally, as part of the walkdown effort, calculations should be performed for all lead shell type anchors to show that the seismic demand load is equal or below the new lower recommended loads.If the seismic demand load exceeds the new recommended value, the anchor may still be adequate providing the anchor bolt is torqued to a proof load.The induced proof load will confirm whether the anchor can carry its calculated seismic demand load.Most pipes at SRS was designed to ANSI Standard B31.1 using B8M anchorage allowables.

Pipes design to ANSI B31.1 has proven to be seismically rugged based on earthquake experience data.Therefore, if seismic qualification of SRS piping is accomplished by evaluating

'he as-built condition against ANSI B31.1 (i.e., no load calculations), then B8M allowables must be assumed, and the anchor bolts must be

RTR-2661.Page 6 August 15, 1989 ATTACHMENT CALC NO proof torqued to induced the BSM design allowable load.ln this case, only one torque test (no tightness torque)would be required and calculation of the anchor load would not be necessary.

SQUG methodology.

permits a factor of safety of 2 for non-shell type anchors but requires a factor of safety of 4 for the shell type anchor.However, this proposed program requires 100/o verification.

of strength of those lead shell type anchors that exceed the new lower recommended loads or if B8M allowables are assumed by performing a torque test.The torque to be applied should be selected based on the lower bound 95%confidence curve (Figures 8-11)and will induce an equivalent mean load on the anchor of approximately1.5 times the seismic demand load.Thus, the SQUG requirement of achieving a 95%confidence level that there are no more than 5%nonconforming anchors is met.The BSM shear load allowables will be rdtained because the measured shear load capacities (Table 2)exceeded the B8M failure values.Shear tests of the 3/4" and 1" anchors was not performed because no.suitable anchors could be found.Table 2 Lead Shell Expansion Anchors Anchor Size, inches Tension Tests Number of Tests B8M Design Allowabe Load Failure Load Per B8M'ean Measured Failure Load New Recommended Allowable Load Torque Based On B8M Design Allowable Load, ft-Ib Shear Tests Number of Tests B8M Allowable Failure Load Per B8M Mean Measured Failure Load 35 34 17 10 9 700 1200 2000 3000 5500 2800 4800 8000 12000 22000 2623 3746 4456 5870 14058 600 880 970 1280 3160 24 37 50 69 16 12 8 400 800 1400 2000 3500 1600 3200 5600 8000 14000 2796 5736 7934

RTR-2661 Page 7 August 15, 1989 ATTACHMENT CALC NO.O PAGE NO DA NL The new lower recommended allowable loads are based on the mean load measured at failure for the sample population tested.The mean (i.e., sample mean)was then statistically adjusted to obtain a true mean load at the 95%confidence interval.The failure load distribution was assumed to be Gaussian.The allowable loads listed in Table 2 provides for a factor of safety of 4.For loads exceeding the new lower recommended loads, torque tests may be performed to determine anchor adequacy.A least-squares regression analysis correlations between torque and load at the 95%confidence level were tabulated by SRL (Appendix A)and have been plotted on Figures 8-11.The torque to be applied should be selected based on the lower 95%confidence bound curve and will induce an equivalent mean toad on the anchor of approximately 1.5 times the seismic demand or the B8M load.Also, the torque to be'applied should not be extrapolated beyond the range of the 95%confidence lines.TEN I NTE Tension tests were performed in P,L,K and R reactor buildings for five different sizes of lead shell expansion anchors.These tests were performed in many different locations of the reactor building: 0 level,-20 and-40 clean areas and radiation zones.A total of 107 tension tests were performed (Appendix B).Fewer tests were performed for the 5/8, 3/4 and 1 inch anchors because of the low availability of anchors.Some anchors were load tested without torquing because the anchors were cross threaded.All 1-inch anchors were found at the floor level and contained excessive amounts of rust, therefore the torque to load relationship was not developed.

Tests showed a wide variation in tension load capacities (Figures 2-7)for each anchor size, reactor area, and mode of failure.This wide variation may be expected because the holding capacity of the anchor is largely determined by the correct hole size and compacting of the lead ring between the concrete and the threaded shell.The amount

RTR-2661 Page 8 August 15, 1989 ATTACHMENT PAGE NO C.5 of force for e~uring adequate compacting of the lead was probably not specified during the original installation.

Approximately 7%of the anchors failed due to cone failure af the concrete.Some of the failures (22%)were due to a combination of a concrete failure (i.e., concrete cone did not penetrate to bottom of anchor)and slippage of the anchor out of the hole.Another 22%'of the anchors failed because the non-lead portion of the anchor broke.The largest portion (47%)of the anchors failed because of slippage (i.e., the anchor was pulled out of its hole without spalling).

About 40%of the slip failures were due to poor installation (i.e., no concrete spalling or significant galling of the lead sleeve was apparent indicating an oversized hole or insufficient packing of the lead sleeve).Nearly all of the 3/4 inch anchors were not installed per B8M.Inspection of the anchors after pullout showed that in most cases only one of the multiple stacked units provided support.A list of the anchors that failed the tightness tests (all turned in hole except1 anchor which was torqued against base plate)or the perpendicularity requirements was not kept but it is estimated that about 10'-15%of the anchors tested were rejected based on those requirements.

E TE Data was also collected to determine the correlation between torque and load for the 3/8, 1/2, 5/8, and the 3/4 inch anchors.Regression analysis curve fits (Figures 8-12)were made for each size anchor and show a good correlation between mean torque and load.Also, a good linear correlation between the mean failure load and bolt diameter was determined.

The lower bound 95%confidence lines.for-torque and load are also potted on Figures 8-11 based on SRL analysis.Torque tests should be used to demonstrate anchor installation adequacy for the B8M design allowables or if the seismic demand load is calculated and exceeds the new lower allowables (applicable to K and L areas only).The torques selected from the Figures for load testing the anchor should be based on the 95%confidence lines and will met the SQUG requirement that there are no more than 5%nonconforming anchors.Wide variation in torque at any specific load was also experienced.

Torque values..differ because of perpendicularity of the anchor with the wall.This type of installation would cause the ram to pull the

RTR-2661'age 9 August 15, 1989 cl3K C9..~eHoR anchor out at yn angle and would yield an increased torque value.Also, some bolts do not fit tight in the anchor until torqued against the base plate (i.e., thread contact between bolt and anchor is not tight).In some cases this could allow the ram to-be mounted with some offset from perpendicularity.

Additionally, the concrete wall was visibly not flat in some locations which would allow the ram to pull the.anchor at an angle yielding higher torques.All of these variables could result in different torque values for the same load.Shear testing was performed only in R Reactor Area.However, tension tests showed no significant variation in failure load capacities as a function of reactor area (Figure 12).Thus the shear tests are expected to be representative of all reactor areas.The mean measured shear load was determined for three different size anchors (Figure 13,14)based on the results of 36 different tests.Tests were not performed on the 3/4 and 1 inch anchors because none could be found.The mean measured shear loads at'ailure were 50 to 74%greater than failure loads indicated in BSM.Only one of the 32 anchors tested failed at a load slightly less than the BSM value times 4 (Figure 14).

FIGURE I ATTACHMQf CALC NO REVISION PAGE NO T 0 G U U RESSURE GAUGE ANCHOR TORQUE NUT II II II HYDRAULIC RAN ANCHOR HEAR ASSENBLY PRESSURE GUAGE HYDRAULIC RAN

ATTACHMENT

."""~NO.CJ1 FAILURE LOAD Sc MODE TENSION TEST 3/8" ANCHOR 6000 5000 A-ANCHOR FAILURE C-COteaETE VAILURE S-SUP FAILURE B-S8 CFAI URE 4000 Cl C 0 3000 MEAN w2623 AASSSSSSSSSACBSSSSSACSSAASBBSSASAAS FA1LURE MODE F1QURE 2

FAtLURE LOAD 8c MODE TENSlON TESTS 0/2" ANCHOR CAI.C NO CJZ 8000 6000 A-ANCHOR FAILURE C-CONCRETE FAILURE S-SLIP FAILURE 8-S 8 C FAILURE 4000 MEAN~3746 2000 ASSSASSSSSSSSSSSASBBBASBBCSCCAAAAS FAILURE MOOE FIGURE 3

CAl.C NO FAlLURE LOAD 5, MODE TENStON TEST 6/8" ANCHOR 12000 10000 8000 A-ANCHOR FA1LURE C-CONCRETE FAlLURE S-SLlP FA1LURE CS-S 8 C FAlLURE 4000 MEANa4456 CS S A S CS C S CS S S A S C S CS CS A FAILURE QQM~ROURE 4

ATTACHMENT CALG ND llDriSiON O FAILURE LOAD 5 MODE TENSION TEST 3/4" ANCHOR A-ANCHOR FAlLURE S-SLlP FAlLURE C-CONCRETE FAlLURE MEAN a5870 S C S S S A A S S A FAILURE MODE FlOUAE I

ATTACHMENT CALC NO O~REVISION n o u ZM eHoP-MEAN FAILURE LOAD TENSION TEST 3/8, 1/2, 6/8,&3/4 ANCHORS 7000 6000 Cl t CO Itl 4000 O 3000.0.375 0.'5 0.625 0.75 ANCKOR OIAMETER (IN.)FIGURE 6

CALC NO>AGF NO MEAN FAILURE LOAD VS.ANCHOR SIZE TENSION TEST y~-529.20+8360.8x R"2~0.985 Ol 4000 N CI 0 3000 z 2000 0.000 0.$25 0.250 0.375 0.500 0.625 0.750 ANCHOR SEE FtQURE 7

TORQUE VS LOAD TENSION TEST 3/8" ANCHOR c PAGE NO 6000 5000 4000 3000 Q 0 2000 1000 SPYe LOWER CONFSEN BOuWO)ya-798+64.86x 0 20 40 60 80 TORQU E (FT-LB)FIGURE 8

ATTACHMENT CAI.C NO REVISION PAGE NO C.l-H op TORQUE VS LOAD TENSION TESTS 1/2" ANCHOR 7000 6000 5000 4000 8 8~8 8 3000 2000 m ma BmG Q 1000 y~-1002+60.31x 0 0 20 40 60 80 100 TORO UF.(FT-l.B)FlGURE 9

TORQUE VS LOAD TENSlON TEST 5/8" ANCHOR ATTACHMENT CAI.C NO o.REVISION d PAGE NO 10000 8000 6000 4000 2000 9P/e LOWER CONFID BOUND I y~-1602+73.08x 20 40 60 80 100 120 140 TORQUE (FT-LB)FIGURE 10

TORQUE VS LOAD TENSION TEST Sia-ANCHOR ATTACHMENT veHo&REVISION OO optic pe t 2d 10000 8000 4000 2000 95Ye LOWER CONFIDENCE BOUND Y~-2147+?5.2X THIS CORRELATION EXCEEDS THE 95'lo LEYEL BY 10 FT-LB TO PERMIT LOAD TEST TO-1.5 x 20 40 60 80, 100 120 140 TORQUE FT-LB FIGURE 11

CGDZDE ATTACHMENI C CALC NO REVISION PAGE NO MEAN FAILURE LOAD TENSION TEST ANCHOR SIZE&REACTOR AREA 3/4-K 5871 5/8-ALL 4456 5/8.R 4704 5/8'100 1/2-ALL 1/2-FI x ILI N CO lL O 1/2-P 1/2-L 1/2-K 3722 3237 2623 3/S-R 2321 3/S-L 3108 2842 0 1000 2000 3000 4000 5000 6000 7000 MEAN FAILURE LOAO (LB)FIGURE 12

REVISION PAOF NO C~MEAN FAILURE LOAD SHEAR TEST 3/8, 0/2,&5/8 ANCHORS 10000 8000 Oi lh IA CJ IA 4000 2000'0 0.375 0.5 0.625 ANCHOR DIAM.(IN.)FIGURE 13

ATTACHMENT CALC NO REVISION o o PAGE NQ SHEAR FAILURE LOAD 12000 MEAN MEASURED MEAN PER BSM s/s 1/2'NCHOR SIZE (INCH)FIGURE 14.3/8'

Niagara Mohawk Power Corporation Nine Mile Point Nuciear Station, Unit No.1 In-Structure Response Spectra Excitation SSE (NUREG/CR-0098, PGA of 0.13G)BUILDING: Reactor ELEVATION:

259'OW/COL:

LA8 DAMPING: 5%MODEL NODE: 58 2.0 1.5 6 C 0 1.0 I I+I I I I I I I I I I I I I I I I I I ee/I J ge I I I I~I I~/I er I I~/I.,j/jl tl I I I I I~l'll'ell'el'll I el"4 I , fe I I I I I I I L I I I I~I r I I I I I I I I I I I t'I I I I I r r I I I I I I~e e~I//I L I I I I 1 I I I I I I I'I I I I I I~~---e.~t-I~I I I~I e I I I~I I I I I I I I't I 1.5 x Bounding Spectrum------North/South

---East/West I I I I I I I t---v I I L I I I I r--"~I L J I I I I I I---r I L I I I I.I I-r-I L I I 1 I I I I L L I I I I I L I I I I I,I I I I i I I I I I~I~I I I'I I I I I I I I I I I I I r I I I I I I 4 I I I I I I I I I I I I I I--.D(RECT(ON" I I I I I I I I I I I I 0.5 I~/----r'-I e/Ie/e/I'.I I r I I I I I I t I I I J I L I I I I I I I I I I I I I I J I I I I I I I I I L L I I I I L I I I I I I I.I I I I'I I I~e I I'I I I I t I I I I I't J't I J I I'l I I I I I I I I I I L I""r"--~lae--I I L I I I I I-t---r---I I 0 0 10 20 Frequency(Hz) 30 aW 40 m~I-e Z o m

Nine Mile Point Unit 1 AQ6 Bounding Spectrum Vs.SSE Plot 0)C 0 0.1 Q tD O 0.01 I I I I I J T T I I I I+I I t T I I I I I I I I I I I I I J I I I I I I I I I I I I-ra r I 4 I--L J I I I I 8 I I I I I I I I I I L4 I I I I I I I I-4 I I I I I I I I I I L J-I-I I I I I I I I I I I I I I I I I I I J I I I I I I I I I I I L J L I i C'I i I I t I I I L J L I I I I I I r1 i I I I I I I I I I I I I I I I I I ro r T I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I Q I i'I 1 I T T I I I I gi I I I I I I I I I I I I T I 1 I I I I I I I I I I I I I I I I I I+9 I H I I I I I I I ra r I I I I I I I I I I I 4 I I I I I I I 4 I I I I I I I I L4 I-I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I L J I I I I I I I I I I I I I I L J I I I I I I I I I I I J.I I I I I I T 1 I I I I I I I I I I I I I I I I I I I I I I T I I I I I I I I I I I I I I I L J I I I I I I-r-i--I I I I I I I I I I I I I I I I I I r I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 4 4 I 4 I I I I 1 I I'I I I I I I I I I I I I I I I I I I I I I I I L.I I I I I I I I I I I I I I I I I 1 I I I I L I 4 I I I I I a I I I 4 I I I I I I I I I I I I I L 2 I--J I I I I I I I I M~I I I s r I I I I I I I I I I I.I I I I I I I I I i I I I I I I P1 SSEi" J L J J L I-L I I I I I I I I I I I I I I I--w-a-t w-r I I I I I I I I I I I I I I J L J J L I L I I I I I I I I I I I I I 1 1 T i I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I---a---i--1-1

-r-i-r I I I I I I I I I I I I I I I I I I I I I I I/I I I I I I I I I I I I I I I I I I Bhund@~Spectrum"" 0.1 Frequency (Hz)10 100 U W Ttl W c O ril