ML18136A009
| ML18136A009 | |
| Person / Time | |
|---|---|
| Site: | Surry  |
| Issue date: | 09/05/1979 |
| From: | Proffitt W VIRGINIA POWER (VIRGINIA ELECTRIC & POWER CO.) |
| To: | James O'Reilly NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| Shared Package | |
| ML18136A010 | List: |
| References | |
| 725, NUDOCS 7910010780 | |
| Download: ML18136A009 (20) | |
Text
,,
\\
VIRGINIA ELECTRIC AND POWER COMPANY RICHMOND,VIBGINXA 28261 i.. : *;
September 5, 1979
~.-.. ~*::.
~-- --.
"'*--...,.'.,.j
~1C)
_1::.. :JJ Mr. James P. O'Reilly, Director Office of Inspection & Enforcement U. S. Nuclear Regulatory Commission Region II 101 Marietta Street, Suite 3100 Atlanta, Georgia 30303
Dear Mr. O'Reilly:
Serial No. 725~
PO/RMB:svm Docket No.:
5~80 License No.:
DPR-32 This letter is in response to your letter of August 31, 1979.
The concerns of your staff as stated in the referenced letter and in the Inspection Report No.
50-280/79-44 are answered -in the attachment to this letter.
The concrete anchor bolt sampling program as addressed in our letter of July 24, 197 9 _has been further reviewed.
The intent of the review was to clarify any questionable data and to modify the acceptance criteria to include your criteria for interim acceptability of plant operation as per the August 20, 1979 Supple-ment to Bulletin 79-02.
This review has been completed and the results indicate that the anchor bolts do meet your.requirements for interim operation.
The accessible supports will be upgraded expeditiously as deficiencies are noted during plant operation and the non-accessible supports will be upgraded during 1;he steam generator replacement outage as required by I.E.Bulletin 79-02.
The field "as-built" program per Bulletin 79-14 as addressed in our letter of August 31, 1979, has been completed for the non-accessible areas with no not-able discrepancies found on the safe shutdown systems. All other field infor-mation is being evaluated and will be handled per the intent of the Bulletin.
We plan to fully comply with the intent and schedule of Bulletins 79-02 and 79-14 and to operate Unit No. 1 in compliance with the restrictions of the lifting of the Order to Show Cause dated August 22, 1979.
The results prior to startup demonstrate that the Unit No *. 1 will operate safely and will be
~apable of being safely shutdown following a seismic event.
As you are aware, the Virginia Electric and Power Company has a strong, con-tinuing commitme~t to assure safe operation of all of our nuclear units.
T-he activities associated with the Surry units this** year has caused us to con-tinuous~y assess the safe operation of Surry in relation to the Show Cause Order of March 13, 1979, the changing regulatory climate, and the 'disposition
~f several Inspection and Enforcement Bulletins, including IE Bulletin 79-02.
Our deliberations and investigations have caused us to draw certain conclusions about the operation of Surry and its affect on the health and safety of the public and our operating personnel.
Our conclusions are indicated below:
- 1.
Power plant piping including Surry's, has an inherent margin of safety associated with it even considering major seismic activity.
This margin
e VIHOINIA ELECTRIC AND POWER COMPANY TO Mr. James P. O'Reilly, Director of safety, admittedly incalculable, does not negate its presence.
Ex.ist-ing power plants have shown* 0an almost remarkable resilience to seismic activity.
(Appendix F, "Seismic Capability of Nuclear Piping," Vepco submittal Serial No. 453, June 5, 1979).
- 2.
The Show Cause Order of March 13, 1979 has caused Vepco to reanalyze all piping originally analyzed dynamically.
The effort constitutes all major piping inside the containment of greater than 6" diameter.
Modifications have resulted from this reanalysis, but the modifica-
'tions were the result mainly of as-built conditions.
The as-built verification effort in IE Bulletin 79-14 currently underway at Surry will resolve all as-built differences.
All modifications to piping systems in inaccessible areas will be completed prior to start~up and this provides us with a high degree of assurance of the operability of safe shutdown systems.
- 3.
Base plate flexibility considerations, in accordance with IE Bulletin 79-02, have been and will be incorporated for those supports for which new loads, due to the Show Cause reanalysis, exceed the original de-sign allowable loads and for any new supports required by the reanal-ysis effort (Vepco Serial No. 5ll, June 25, 1979, W. C. Spencer to Harold R. Denton).
Therefore, base plate flexibility considerations have been a part of all present and future modifications associated with the Show C~use effort.
- 4.
Sur>y Unit 1 cannot operate and must be kept in a shutdown conditioµ for any seismic event which exceeds a ground acceleration level of 0.01 g (p. 3, Order, Docket No. 50-280, signed by Harold R. Denton, August 22, 1979).
This value is a small fraction of the Category I piping system design ground acceleration value of 0.15 g.
We believe this is a very conservative approach of interim operation.
- 5.
The staff of the NRC Office of Nuclear Regulation has thoroughly evaluated the procedure for design of the piping at the Surry plant not subjected to computer seismic analysis, i.e., those systems now falling in the IE Bulletin 79-14 scope of work.
They have compared the design procedure to the NRC' s Standard* Review Plan 3. 7. 2 and found our methodology to be acceptable (p. 9, Safety Evaluation, Surry Power Station, Unit No. 1, Docket No. 50-280, dated August 22, 1979).
We believe the NRR staff shares our confidence in the design of these systems.
- 6.
During any one month, there is a hazard of 4.5 x 10-4 of equaling ~r exceeding an'earthquake with a peak acceleration of 0.04 g.
Thus the chances are very slight that the plant will experience any signifi-cant shaking due to an earthquake during the period of interim opera-
~
tion prior to steam generator replacement.
The chances of experienc-ing the DBE are extremely small (Section 7.8 of report tr~osmitted to NRR, June 5, 1979, Vepco Serial No. 453) *
- 7.
The Surry Power* Station has some anchor bolts that ar~ not per the manufacturer's instructions.
We do not agree that dition unilaterally voids the anchor bolt design capacity.
installed this con-In fact,
VIRGINIA ELECTRIC AND POWER COMPANY TO Mr. James P. O'Reilly, Director 8.
- 9.
- 10.
our sampling program has demonstrated our stated confidence that the anchor bolts as instal_led can indeed carry the loads required in order for the supports to perform their intended function.
Based on this fact, coupled with other arguments presented herein, it has been shown that the Surry piping systems can perform their intended func-tion.
IE Bulletin 79-02 states case histories of poor anchor bolt installa-tions at other units.
Surry Power Station does not have the anchor ool t installation problems to an extent even remotely close to the units cited in the Bulletin.
Documentation problems exist, but this is not to be confused with lack of design.
As this submittal demonstrates, acceptable and con-servative design procedures were used or-iginally.
Long term programs for IE Bulletins 79-14 and 79-02 have been initi-ated and are continuing.
The results of these ongoing activities are available for your inspection at aRy time, and the alternative courses uf action are not obviated by interim operation.
In the process of evaluating these factors, we can only draw the conclusion that interim operation of Surry Unit 1 will have no adverse affect on the health and safety of the public or our employees.
Your expeditious review is requested as the Unit No. 1 will be ready for startup on -September 6, 1979.
Very truly yours, W. L. Proffitt Senior Vice President - Power RMB/svm:2K3 cc:
D. Burke
I' ATTACHMENT TO LETTER rc __ _
ANSWERS TO NRC REGION II CONCERNS NRC CONCERN VEPCO could not demonstrate that stress analysis had been performed forcer-tain systems for Surry l..
These systems could be categorized as piping systems less than 6 inches in diameter except those included in the NRC Show Cause Order.
Systems for which documentation was not available included certain ones (both inside and outside containment) that would be necessary for the safe shutdown of the plant following the seismic eve!1t.
VEPCO believes that this field run piping had been designed using hand calculations to determine stresses and that the supports were located using "typical" guidelines.
Data presented or made available to NRC inspecto:cs did not appear to provide necessary infor-mation to support a finding of technical adequacy.
Documencation was not available that would assure that the pipe support systems including expansion arichors for the piping systems described above had been desj_gned to support tbe loads that would be imposed.
Based on the above, we believe that additional assurances by VEPCO are neces--
sary and that VEPCO should provide justification for a. samplj_ng test progra.m that would provide assurance that the unit can be saf_ely operated and can be safely shut dom1 in the event of a seismic event.
VEPCO RESPONSE Information available at the me.etin;:~ of *August 28~ 1979, was not complete P-Dd r;bould
. be inspected with our Architect Engineer; howevc~r, the above concerns ba.d 1neviously been answeren to. the NRR staff during their overall inspection for the Orc1.er to Show
. Cause.
To address the NRC concern expressed above, the piping systems referred to
- were-designed using the criteria and design requiJ:ern2nts in effect dur:f*,,g the design and construction of Surry.. The design of piping systems using 11decal 11 loads is an acceptable method of design which hns b2en reviewed and approved by the NRC.
When Surry was constructed, provisions were not made to incorporate all field engineered changes j_nto the construction drawings and, therefore, 11as-builtn discrepc:.;:,cies can be expected.
The documentation and record-keeping requ.irements in effect today are more stringent than those in existance when Surry was built and therefore, supporting calculations and drawings for these discrepancies are not as accessible as for a plant recently constructed.
Since the design requirements for this piping were the basis for field design modifications, VEPCO believes that the "as-built 11 discrepanci.e.s were engineered changes for whj_ch the documentation is not readily available and not changes which were made,v-ithout any design basis.
The use of 11decal 11 loads as an acceptable method of design was reviewed by the NRR staff during their overall inspec-tion for the Order to Show Cause.
After the inspections by the NRR Staff at our Architect Engineer's offices in Boston, they,-,ere satisfied and addressed this. question
- j_n their SER of August 22, 1979 as quoted below:
Verification of Analysis Methods We have reviewed the acceptability of the analytical methods which are currently a basis for the facility piping design.
The licensee has identified the following computer codes/analysis methods as applicable:
PSTRESS/SHOCK O (Initial 3 Versions of SHOCK 1)
Static Analysis Methods NUPIPE
- ~.. PSTRESS/SHOCK 0 This code was used for 12 safety r~_lated system problems and although it did not algebraically sum responses, the code was not equivalent to current prectice.
The licensee, therefore, reanalyzed these systems with the NUPIPE code.
Statis Analysis Methods used for design of the piping at the Surry Plant not subjected to computer seismic analysis were based on simple beam formulations which, in essence, controlled seismic stress levels through use of pre-established seismic spans.
Th~se simple beam formulations were utilized to calculate maximum allowable spans based upon an assumed acceleration factor of 1. 5 times the pe3k acceleration obtained from the response spectra.
In calcula-ting the maximum span lengths, it was conservatively assumed that a longi-tudinal pressure stress of 4,000 psi and a maximum deadweir,ht stress of 1,500 psi were present in the pipe.
This combined value of 5,500 psi was subtracted frm-;i the allowable stress (1. 8 Sh for pressure and dead~\\'eight and seismic) to obtain a seismic allowable s*tress.
Calculating mB.ximum spans by this procedure results in maxj_mum allo:,rabJc spans greater than the deadweight spans recommended in ANSI B31.J.
Thus deacJweight goven1S and pi.O'lides a greater number of supports resulting in closely spaced. restraints.
To minimize effects of concent1.-ated weights) restraints were placed as required at valves and o~1:..cr ce--::.:::2::tr:itcd masses.
Tor Surry Unit 1, piping 6 inches in diameter and smaller was generally
- ~nalyzed 1wing the:. simplified static method~ with the option of utilizing more rigorous rn.ethods available to the analyst.
Piping 2 inches and belo,., was sho,-m on the piping drawings diagram;-:,a ti-cally (i.e., without detailed dimensions).
The stress engineers located supports during the inst2.llation process working at the site with erection isometric sketches.
As described above, the stress analysis was performed by assuming many simple supported straight beams, the spans of which are governed by dead load spacing requirements of ANSI B31.1.
The piping fundamental frequen-cies associated with these maximum allowable spans (9.7 to 13.6 cycles per
- second) are not in rc;sonance with the building in which they are located (2 to 8 cycles per second).
The method of equivalent static analysis outlined in this procedure has been compared with the NRC's Standard Review Plan 3.7.2 and is found to be acceptable.
End of NRC SER quote
2 -
VEPCO RESPOMSE (CONT'D.)
Static Analysis Methods used for design of the piping at the Surry Plant not subjected to computer seismic analysis were *oased on simple beam formulations which, in essence, controlled seismic stress levels through use of pre-established seismic spans.
These simple beam fonnulations were utilized to calculate ma~imum allowable spans based upon an assumed acceleration factor of 1.5 tines the p'eak acceleration obtained from the response spectra.
In calcula*-
ting the maximum span lengths, it was conservatively assumed that a longi---
tudinal pressure stress of 4,000 psi and a maximum deadweight stress of 1,500 psi were present in the pipe.
This combined value of 5,500 psi was subtracted from the allowable stress (1.8 Sh for pressure and deadweight and*seismic) to obtain a seismic allowable stress.
Calculating maximum spans by this procedure results in maximum allow-able spa,ns greater than the deadweight spans recommended in ANSI B3L 1.
Thus de.adweight governs and provides a greater number of supports resulting in closely spaced restraints.
To minimize effects of concentrated weigh~s.
restra:i.nts were placed as required at valves: and other concentrated m<'::c:~-:;cs, For Surry Unit 1, piping 6 inches in diameter and smaller was generally analyzed using the sirr:.plif:i.ed static method) with the option of utilizing more rigorous methods available to the analy.:::t.
Piping 2 inches and below was shown on the piping draw:Lngs dj_ar;rcrcJ.u.ati-cally (i.e., without d2taiie_d dimension~-).
The stress engineei~s loc2.ted supµori::s during the installaU.on process working at the site with erect:\\.OJ!.
isometric' sketch2.s.
As described above, the stress analysis was performed by assuming many simple supported straight beams, the spans of which are governed by de.ad load spacing requirements of AKSI B31.l.
The piping fundaraent2J.. frequen-*
cies associ;ited with. these maximum allowable spans (9. 7 to 13.6 cycles per:
second2 are not in resonance with. the building in which the.y are loc2.. ted (2 to 8 cycles per second).
The method of equivalent st.s.t:i.c analysis outlined in this procedure has been compared with the NRC 1 s Standard Review Plan 3.7.2 and is found to be acceptable.
- VEPCO RESPONSE (CONT'D.)
The followin8 is a listing of the static analysis methods used for design of small piping.
SIMPLIFIED METHODS FOR PIPING A1'lALYSIS SURRY *-POllER STATION LICENSHTG REQUIREHENTS:
Piping systems 6 in. in dizmeter and smaller were stress analyzed, using accel-eration loads from ground response.° spectra.
Stops, guides, and snubbers, were located to preclude piping resonance with the supporting structure..
Displace.-*
ments of the piping were cbecke.d to ensure that there were no interferPnf'_es with anv other equipment or piping. The documents listed below specify the initial design requirements.
Date Ref.
Method
!ppli~ability 6-23-69 7-25-69 10*-28--69 4-23-70 Date 7-17-67 PS-4 PS*-1 PS Rev. 1 Aides
- From/To Preliminary Issue Simpli-fied analysis standard support loads.
Kellogg method to span supports +/- 50% of the peak structural resonant fr2c;,ue.n--
cy.
. Sinplifj_ed analysis standa.rd support loads (total).
Computer tabu1a t ed seismic stress for giveU" pipe) spa.n~
a:;id acceler:ation plus nomo**
graphs developed from equations for seismic support span> pipe frequency, the::rmal stress and flexi.b ili ty.
Nominal pipe size 2 11 through 24';
LoF te.mpcrature pipiP..g under 6 11 <li.a.
Nom:t.r,,al pipe through 2Lf.
Nominal pipe and unclej:"
s:i.2e 2" size (, 11 Content Stone & Webster Procedure Description of:
Pipe stress analysis for earthquake and tornado resistant systems specifying:
Systems designed for earthquake Type of analysis Technique Seismic forces Thermal and seismic stress Earthquake reaction Combination of stresses Allowable stresses Reaction forces and moments Tornado protection Structural response (seismic)
Date 3-12-68 From/To R. l-1. Rome to M. H. Pedell 4 -
Content "If all seismic pipe stress problems were to be analyzed dynamically, a simplified method of computer input is mandatory."
Needs:._.
- 1.
Accept input from stress program 2~
Call dead loads
- 3.
Ecompass limitation of 75 nodes 4 *.
Auto. thermal, seismic and weight combination for cofilp3rison with the allowables,.
. 12-22-70 R~ P. Klause to "Seismic analysis of 'field run piping" 1971 1971 9-15-71 O. Surgecoff
- 11. II. Pedell to R. H. L'.Ar.loureu..-:
s & w PSAR s & w 2" and smaller shown diagrammatically.
Boston stress engineer to analyze the lines at the job.
Engineer would v:alk lines ancl make freehand isometrics.
Hanging of diagrammatic lines i.s field responsibility.
Clip 3/ 8 11 SQmple lines every 4-5 t_ to walls or cab le trays.
"VEPCO Surry -
Units 1 2.nd. 2 Stress analysis re.vised and se:i.mn:Lc criter:Lio."
as per Surry FSA.R seismic lo2ds on all CJ.asr,: I piping must l.i0. incre2.sed~
Discussion of impc.cr. of stress~ support loading and schedule:'..
Seisnic acl.e.quac.y review -*
Covers response spectra, reactor coolant systera and Class I piping, ec;1.1iprn.ent, and cor:iponents, Section 4.1.l "Hethocl of Dynamic Analysis 11 Piping 6 in. and smaller were seismically supported to "preclude piping resonance with the. supporting structure."
Seismic design review Equipment and piping Surry Power Station Same as above (of 6-1-71) revised 9-15-71 Ref. Sections remain the same Includes response spectra Therefore, documentation is available to ensure that the p1;.pe suppqrt systems) including expansion anchors, for the smaller than 6 inch lines of NRC concern had been designed to support the loads that would be j_mposed.
Additional assurances are available in that, per Bulletin 79-14, ali 2 1/2 inch and larger piping will be or has been re-verified as to 'as-bu,ilt' conditions.
The Unit No. 1 piping has all been redrawn.
The smal~ bore inaccessible safe shutdown piping has been initially evaluated with no nota-ble discrepancies found.
All other safe shutdown inaccess:i.ble piping has been re-analyzed per the Show Cause Order and modifications made.
- VEPCO RESPOHSE (CONT'D.)
The small bore safe shutdown accessible piping has been initially "as-built" and initial walkdown Quality Control comments and discrepancies are being verified.
and will be evaluated on a priority basis and handled per the require.me:1t:_s of 79-14 as stated in our letter of August 31 1 1979 Based on the above, we believe that sufficient assurances have been provided and that in fact the unit cnn be safely operated and can be safely shutdmm in the event of a seismic event.
6 -
NRC CONCERN
- 2.
VEPCO should demonstrate that anchors with sleeve top to expansion plug dimensions that indicate less than full sleeve embedment or less than full sleeve expansion will withstand a loading of at least two times the design load (1/4 manufacturer's ultimat_e). It is suggested that anchors with the worst di,mensional anomalies be load tested but equivalent methods would be acceptable.
VEPCO RESPONSE The Anchor Bolt tnspection and rest Program conducted required that anchors in the sample be proof loaded to demonstrate that the anchor will withstand at least its design *capacity (J_/4 of manufacturer's ultimate) in tension in accordance with the test requirements outlined in I.E.Bulletin 79-02.
Tensile load was applied to the.bolts by applying a torque to the bolt.
The torque values used to apply the proof load were furnished by the manufacturer and were based on tests performed by others.
These values are listed oelow:
Anchor S12 S58 S34 S78 Table l BolLUin.)
1/2 5/8 3/4 7/8 Proof Torque (Ft-*Lb )_
30 45 70
':J:J
. The above values were compared to those obtained from an analyt:i.cal method to correlate installation torque to bolt tension.
Since. the proof load torque.
listed above was significantly higher than the *c;.alculated torque required to produce the required tension using this analytical method, the proof load
--torques were judged to be acceptable.
The analytical method used is presented in the. text "l1echanical Engineering Design" by J. E~ Shigley, HcGraw-Ilill Book Company, 1963, and employed a coefficient of friction of O.15 as a value for average bolts and nuts.
Original design crite.ria for the anchor bolts installed at Surry Pm,er Station was based on allowable bolt loads factored for 3000 psi concrete with a factor of safety of four (4) based on the manufacturer's pub-lished data.
These allowable bolt loads were the tensile loads for which the corresponding torques *were calculated.
Table 2 shows these calculated torque values and the actual proof torque values provided by the manufacturer as indi-cated in Table 1 above.
T.able 2 also shows the factor of safety proven by applying these ~orques assuming the torque-tension relationship T = K Fd given by $higley as referenced above
- Size 3/8 1/2 5/8 3/4 7/8
- VEPCO RESPOHSE (CONT'D.)
Design Load(l)
Allowable (lb) 1218 1826 2517 3487 3835 Calculated Proof (2)
Torque. (Ft-Lb) 8 15 26 42 54 Table 2 Actual Proof Torque 25 30 45 70 95 (Ft-Lb)
K (3)
Value 0.204 0.2-01 0.199 0.194 0&193 (1) based on factor of safety of four (4) and adjusted for 3000 psi concrete.
(2).
based on analytic.al.method referenced above
(_3)
Torque coefficient IC given by Shigley; 1963 edition.
(_4)
Assumes.s.pproxirnate relationship T = K Fd Factor Of(4)
Safe t*,
3.13 2.00
- 1. 73 1.66
- 1. 75 The proof load -torque as discussE:cl above demonstrates that the. shell is suffic:ien tly expanded and embedded to resist some margin above design allowable tension loads, and -therefore, justifies concluding that it is also adequate for shear and combined shear/tension design lo.s.d,:;.llowables.
The criteria used to evaluate the acceptability of shell movement during proof loading of the anchors was that shell outward movement was insufficient to permit the shell to come up into contact with. the baseplate or that the shell did not rotate when the proof load torque was applied.
Minor shell movement was accepted if the shell lod~ed up and did not contact the baseplate when load tested.
On August 20, 1979, the NRC issued Supplement No. 1 to Revision 1 of I.E.Bulletin 79-02 to establish criteria for evaluation of interim acceptability of plant oper-ation with less than the design factors of safety for piping supports due to as-built problems, under design, baseplate flexibility, or anchor bolt deficiencies.
The supplement requires that for an anchor bolt, th.e factor of safety be equal. to or greater than two (2).
In order to demonstrate by test that a safety factor o*f two
(_2) exists in anchors with sleeve top to exp_ansion plug dimensions that indicate less than full sleeve embedment or less than full sleeve e~-pansion, several anchors with the worst dimensional anomalies and: pasi:;ed the proof load test were *re-tested as suggested by the NRG.
The anchors were re-tested to at least two times the design load (J.t4 manufacturer's ultimate in 3000 psi concrete) by applying an equivalent torque.
The equivalent torque was again calculated using the approximate relationship given_ by Shigley as referenced above.
Since this is an approximate relationship, a factor of 1.25 was applied to the calculated required torque to account for unknowns and to ensure that an equivalent tensile load was applied.
The following relationship was thus
8 -
VEP CO RF.:SPONS E ( CONT r D. )
used to calculate the required test torque to demonstrate a safety factor of two (2) o TS.F. = 2 *(2) (1.25)(K Fd)
Where Ts.F. = 2 = Torque Required to Test Bolt to S.F. = 2 K
torque coefficient F = design load allowable based on FS of 4 and 3000 psi concrete d = diameter of bolt The test torque values used are as listed belrn~:
.Bolt Size (i~:;_2_
3/3 1/2 5/8 3/lf 7/8 Test To;_q~~
25 l10 70 lJ.O 135 Safety Factor Assuming
- !:EE_roximate Relations hi£ T=-=L ~5 K.!.i 2
2 2
2 2
The bolts re*-tested to demo11stra.te a safety f2c.tor of two (2) are listed in Table 3o The fourteen (11+) bolts selected for re--test: were representaU.. ve of the anchors wij:h the worst dimensional anomalies.
The test results indicate that all but one (.1) of the belts re--tested p8ssed the proof load re-test
. proving that a safety facto-.c of greater than two (2) exists in these. anchors..
One anchor, when removed to check the dimensions, could not be re-inserted fully into the anchor when the anchor and sleeve began turn:i.ng together.
.. The.refore, it_ could not be shown that a safety factor of two (2).existed in this and10r prior to the re-* test.
This anchor will be repaired..
Based on the results of the re-test, we believe that reasonable assurance exists that a safety factor of at least two (2) exists in the bolts included in the representative sample which satisfactorily passed the initial t1roof load test
- TORQUE TEST RESULTS v ING SAFETY FACTOR OF 2
~--*-
fiANCT!OJ( r*AcKOFl'T rnlPTI~ 'f mR!Wl-Pi1SnELLl
--=
=
IPING SYSTEH
! l SJiQUEl!CE DEFICIENCIES NOTED ID NO.
TORQUE OF SHELL ENGAGE LOAD MOVE-AND HAUGER I AND SET I l'Ro.r 'N rnmT TOHQUE MENT 1RAWING SERIES NUMBER nurnmrr SIZE I I PLUG I
, z:;~-~ f~-;1_~1~1~:;L F:~~11, c~~:J-
~SRS 1-9 H-92
,_1448-FP-13A I~
I i
l{odified Safety 3-1 11-7 12-3/4 I 70 I 2-11s ! -11s 3/4 110 No njection I
1448-MSK-2A I
t
.;afety Injection 3-2 H-10 1-3/4 70 2
t 0
I*
1 110 No 1448-ASK-2A i
I I
I 2-1/81-~/8 I
i I
RHR 11448-FP-12A 3-17 H-30 5-3/ li 70 I L
1 110 No j
I I
l 1-3/41-1/4 I
RHR 11448-FP-12A 3-27 I
H-45 15-3/4 70 I 1-3/8 110 No
\\.C I
12-3/4 !
I i
l *1-1/41
~vcs 11448-FP-lOC I 5-53 H-206
<70 2
I -11a 110 Ho I
- i.
I i
I JSSRS 11448-FP-9A 6-12 U-111 I 3-3/ 4 !
70 2-112 \\ -118 I 1 110 No I
I i
i l
i I
l 12-3/41 i
i I
)
I
'SSRS 1+48-FP-9B 6-21 11-24 70 l 2-1/s I -11s 1-3/ s1 Could not perform torque test I
i
- anchor & shell started l.
j I
I
~
I I
turning.
I i
I t
I I
- 2-1/ 41 j
I 1-3/41 r
1-1/8
- II 11448-FP-lOR 6-39*-
II-170 70 f
0 110 No I
I I
l I
i I
I i
j -
I I
I I
i 2-1/4 ! -1/8
~H 11448-FP-lOR 6-41 H-157 1-3/4 J 70 i
l 1
110 No r
l
' i i
I I
i l
I I
i l
j J " '4 !
I I
... HSI 1148-FP-llA 7-29 ll-22I3 70 I
/
j 0
1-1/8 110 No j _-:;/
I 2-1, 41 I
I I
I I
i 111-7/81 I
- omponent Cooling 8-30 CC-Al4 50 2-3/lf 1 -1/8 1
135 No 11448-FP-16G
' I j
IPING SYSTEM AND MWING SERIES I
-* tiNCHOR Bi\\CK.()liF ri51frnfl rTilREi\\D [PROOF. s1fau:===-==-
. ID NQ6 TORQUE l O'F I SHELL I ENGL\\.GE-, LOAD MOVE-DEFICIENCIES
=
AND
, s:rr 1:PROJ 1N1 H.EtE' TORQUE MENT
. SEQUENCE II.t\\NGER NUMDER nrnrnrm 3IZE !
I PLllG
=======-=-f*=====~=====**=~-ill_z) ! (Ft.-lh;l __ (:ln.J~LJ-ihl/ (it\\~.,)~-TulJin._L..
omponent Cooling J-1/2 f-:1..,l-1/41 1
_ -1/8
- 1i, 3/8 40 I No
=
9-5 R-134 1448-FP-16M o:....;m;....p;...;;o_n..:;.e.::..n_t;::.C~oc.c..o-l-i-n-g--+-------+-----~!--1--1-/-2-1!
30 1-3/81 0 1
1
- 3/lf 40 ! No 9-15 CC-llil-231 1448-FP-16C j
1 1
I I
I I
_.I i
~
I i I I I
I I
i I !
l t
I I l i
I '
I ! ' ! I
)
I I I I i l l j l
I I I I
' ! i I j
I I I
i l I
' 1 I
I I
I I
i l
I I
l I
I I
I I
I I
I l
I f
~
l'
~
1 i
j l
i I
t
~
l I
I I ! I i
I
! I NOTED I
I-C I
11 -
NRC CONCERN I
- 3.
VEPCO should address, quantify, and analyze each case where significant de-ficiencies such as broken anchors, adjacent nuts missing and smaller than drawing required number and/or size of anchors installed were observed but excluded from the sample and thus from the swnmarized failure results in your letter of July 24, 1979.
VEPCO RESPONSE The calculated failure rates presented in VEPCO letter Serial No. 146 were based on the deficiencies found during inspection of the 200 anchor bolts included in the sample program for which the installation attributes could be measured.
The calculated failure rates can therefore be taken as represen.tative of the total population since they were chosen by a random selection.
Deficiencies or discrepancies noted during the inspection but not included in the sample such as broken anchors, adjacent nuts missing, or smaller th2.:-L required nun.her and/or size anchors have been listed on the corrected data summary sheets contained in the Anchor Eolt Inspection and Test Progrem.
These deficiencies/
discrepancies are being analyzed or corrected in conjunction with deficiencirss noted on bolts included in the representative sample.
All de'ficiencies/ discrepaa**
cies noted during conduct of the test program are addressed and analyzed in the following paragraphs.
Discrepancies found to exist bet*ween the original desj_gn drawings and the.
"as-built" condition are being evaluated to determine the~ adequacy of the. support:.
These discrepancieB consist of (_1} different anchor si~e, (.2) different munber of bolts and different configuration, (_3) weldecl baseplate in::tead of bolted,
(.4) missing sup?orts, and (5) other 11a..s-built" discre_panc.i.e::; a::; de::;c.:.:dbed belm*J, The evaluation consists of making "as-built" drawings and re-analyzing the support to dete.rmine the anchor bolt loads using the original anaJ.ytiec:1 methods.
The loacl :i.ng combinotions being used are consistent with those which have been approved under the Surry Task Fo:rce interim start-up pipe. support design criteria.
Any non-conformance identified as a result of the re-analysis will be classed in accordance with the criteria given in Supplement No. 1 to I.E.Bulletin 79-02 to determine operability.
The discrepe.ncies being analyzed are quantified and analyzed below:
(1)
Different Anchor Size The proper anchor bolt size was not always installed as required on the sup-port detail drawings.
Those supports detennined to have improper size anchors are being 11as-built 11 to the existing field conditions and re-evaluated as described previously.
There are 47 support_baseplates containing different size. anchors.
The worst. case found to exist involved the situation wher.e anchor bolt of 1/2 inch diameter was installed in place of the required anchor bolt of 3/4 inch diameter.
In an effort to qualify the difference in anchor size, the following relationship was used:
T(l/2) = [(K)(F)(d)] (2)(1.25)
~ [(0.201)(3487)(1/2)) (2)(1.25)(1/12)
= 73 Ft-Lb' s VEPCO RESPONSE (CONT'D.)
(2) where: Tis the Test Torque used to qualify the 1/2 inch anchor K is the Torque Coefficient for the existing anchor Fis the Design Load Allowable for the required anchor assuming 3000 psi concrete and a safety factor of~-
d is the existing bolt diameter.
In the worst case referenced above, three 1/2 inch anchor bolts were torqued to 75 ft~lb's.
This torque value justifies the ability of the smaller sized anchors to withstand a tensile load of equivalent to at least twice the design allowable load of the required anchors.
The results of the testing are shown in Table 4 titled "Test Results for Anchors Smaller Than Design".
The test was performed in accordance with procedure P-4 of Special.Test 39 except for the increased test torque value.
The test results indic2ted no deficiencies.
The test results indicate that although lesser sized anchor bolts have been installed in place of the required anchors, they are capable of handling loads equal to twice the design alowable load for the required anchor size.
In addition~ preliminary resultG of "as-built" analysis of the 24 hangers analyzed, shows that a safety factor of at least two (2) is present for 4500 psi concrete.
Easepla.tes were alco ::c:.:r.::1 2~:.c. ~!er:U::Led to have different numbers of bolts and different bolting configurations, Forty-six base.plates were found to exist in th:i:'.s condition.
Each of the supports associated *with these base--
pletc::s are be:tng 11as-built 11 to their existing field conditions and will alt;o be re-evaluated as described previously.
Prel:iJnina.ry rr::sults of the seventeen (17) hangers re-analyzed to date indicate that a safety factor of at least two (2) exists with 4500 psi concrete.
(3)
Heli_ed Basepla t-~~
There we.re nine (9) baseplates found to be weJded to structural members rather than: bolted using the required concrete expansion anchors.
The supports containing these baseplates are being "as-built" for re-evaluation, However, the initial detennination is that.this type of attachment meets the design requirements.
(4)
Missing Supports Several supports were listed as missing on the data sheets and as such were excluded from the sample program.
All but one of these supports has been accounted for and identified during subsequent inspections.
Therefore>
these discrepancies have been resolved
- NO. l~
TEST RESULTS FOR S:HALLER TUAN DESIGN (SAFE
'"'~oR OF 2)
-* --=
'fXNci10R1DicicoffriT-mi"~---
PIPING SYSTEM
- . 1ID NO. ! TORQUE I OF l SliELL AND SEQUENcg U./\\NGER l ANJJ I
l SET
!*PRC,JvN DRAWING SERIES NUHDER ummER* I SIZE ~
I PLUG !
t~-i (Ft.-~,L:h~~-:.Li_ (in.)_
eves 11448-FP-lOT 5-24 ll-46 12-112 I 30
, 1 1-1/4 1 -11s I
I I
I
=rTII REJ\\Dr~()()f
~1'.!=
=
=
SHELL ENGi\\GE LOAD MOVE-DEFICIENCIES NOTED I HENT lTORQUE HENT
,i.!.'!J.... r",;::l!, __(in0
- r=.....
=
l 3/ lf I 1s No 1/2 inch, 3/4 inch required I
I I
1-1/2 I 1-7/lci -3/16 eves 11448-FP-lOT 5-55 II-220 30 I I
I I
I eves 11448-FP-lOD 5""'.'70 H-212A
- 4-1/2 I 30 I 1-3/S; -1/4 l
I f
l I 5/8 75 No 1/2 ~.nch, 3/4 inch required i 3/4 75 No I
I 1/2 inch, 3/4 inch re4uired 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
l I
f l
I l
1 I
i I
l
/
I l
i I
I I
I I
i l I I
I l
I
~
i t
I
}
I l
I I
t i I
I I
I l
l I
i l
I 1
~
I r
l
[
\\
i i
J i
i l
I r
' ' I I
I I I
i
~
I I
- 1 I
I I i I I
I I
I I
l.'..> -
NRC CONCEHN
- 4.
VEPCO should address and justify the fact that the random sample did not include anchors from all safety related systems.
~SPONSE As described in the Anchor Bolt Inspection and Test Program, the baseplates inspected were selected randomly from those baseplates in safety related systems for which design data was available at the tine.
As a result, the random selection did not include anchors installed in all safety related piping systems.
Although the inspection did not include bolts from each safety related systemj tne test program did include bolr:s installed in all piping systems 2 1/2" and larger that*are required to operate for safe shutdown.
Bolts were tested in each of those systems except the Auxiliary Feedwater System.
These bolts were excluded from the sample since the baseplates were grouted and the studs could not be removed for inspection.
Those baseplates with anchor bolts tested are considered to be representative of the total population and the workmanship at the time of construction since there j_s no unique character:i.s tic of these supports that would affect their randO"mness and no dif~erent construction practices -were used for bolt instal-lation from one system to anothe:r.
Since t:he bolts were. selected randomly and includEid all systems requirP.c1 for safe shutdown with accessible a.'1.chor bolts~ the progrc.m conducted is represen***
tativ,~ of all safety related piping. 3ystems and serves as justification for interim operation.
NRC CONCERN
- 5.
VEPCO should summarize the results of the test program based on the reviewed and corrected datao The format should be similar to VEPCO's letter of July 24, 1979, and include the calculated failure rate(s).
~VEPCO RESPONSE All initial data sheets completed have been reviewed. and those data sheets which had "questionable" data, such as large or small depths to the set plug or thread engagements that exceeded the space available or that exceeded the actual thread length measured in new anchors of. that size have had the data verified and. dis-crepancies have been resolved.
The corrected data has been entered in the summary tables.
Results of the test program based on the reviewed and corrected data is smmnarized below.
The results given are based on the 200 bolts included in the sample for which the installation attributes could be measured.
Deficiencies noted durj_ng the inspection but not included in the sample such as broken anchors J adjacent nuts nissing, or smaller than drawing required number and/or size of anchors installed are addressed in Item No. 3 above.
111e calculated failure rates given below can therefore be taken as representative of the total population of anchors installed in safety related systems since they are based on the 200 bolts selected randomly.
The results were evaluated in accordance with the acceptance criteria contained in the Anchor Bolt Inspection and Test Program to determine if the anchor bolt installation was acceptableo The acceptance of the anchor w2s reduced to only four (4) attributes to detcrrniue if the bolt would perform its de.sip,n funct:ion:
anchor size, initial tightness, thre2d engagement) ancl proof load capability, The inspection results indicated that 26 anchor bolts of the 200 :Lnclude.d j_n the test sample or..13% had an installed anchor smallP.r than the anchor diameter specified on the ori8inal design drawingc Documentation could not be located to ~erify the adequacy of the 11as-built 11 installation.
For those baseplates where the same nur.:i1er of anchor bolts were used, but smaller anchor bolts were provided, the design factor of safety based on 3000 psi concrete is reduced from
!f to a minimum of 2 for tension and bet,veen 1. 7 and 3. 5 for sheo.r depcndi.ng on bolt size.
The reduced factors of safety are higher when the actual concrete strengths are considered.
The supports. in which smaller anchor bolts exists are being "?is-built" and re-analyzed by Stone and Webster Engineering Corporation as described in our xesponse to Item 3 above. Any nonconform.ances identified will be classed in accordance with the acceptance criteria given in Supplement No. 1 to I.E.Bulletin 79-02 to determine operability.
Of the bolts inspected, 9 anchor bolts of the 200 included in the test sample or 4.5% had no initial tightness.
So~e initial tightness is required to restrain
- the baseplate to prevent displacement when the load. is applied.
The inspection results indicated that 4 anchor bolts of the 200 included in the test sample or 2% did not have a thread engagement of at least ~/2 of the bolt diameter.
Therefore, it can be concluded that a thread engagement problen does not exist at Surry.
Bolts in the sanple were proof torqued to a value equal to a minimum-of 1.66 times the. allowable tension design load based on 1/4 of the ultimate pull out value for 3000 psi concrete.
Of the 200 inspected, 13 or 6.5% did not hold when loaded to the proof load. Five (5) of the 13 bolts which did not meet this criteria rotated in the hole and thus would not hold the required torque.
Even though the shell turned when the bolt was torqued, the bolt*s were recog-nized to have some capacity~ Additional proof load testing conducted on
VEPCO RESPOtISE (CONT'D.)
anchors with the worst dimensional anomalies as described in Item 2 previously provided reasonable assurance that a safety factor of at least DYO exists.
In summary, the testing program conducted at Surry showed that 92.8% of the attributes measured to determine acceptability for interim operation were considered to be acceptable.
The def~ciencies observed during the sampling program indicate primarily that the safety factor above design load was less than required by I.E.Bulletin 79-02 for final design margins, but satisfie~
the interim operability requirements of Supplement No. 1 to I.E.Bulletin 79-02.
- ~~ *..