ML20239A706
| ML20239A706 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 12/22/1987 |
| From: | Gridley R TENNESSEE VALLEY AUTHORITY |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| NUDOCS 8712300014 | |
| Download: ML20239A706 (25) | |
Text
-
- _ _ _ _ _ _ _ _. =
j y., /
TENNESSEE VALLEY AUTHORITY CH ATTANOOGA. TENNESSEE 37401 SN 157B Lookout Place DEC 22198T U.S. Nuclear Regulatory Commission ATTN: Document Control Desk Hashington, D.C.
20555 Gentlemen:
In the Matter of
)
Docket Nos. 50-327 Tennessee Valley Authority
)
50-328 SEQUOYAH NUCLEAR PLANT (SQN) - REVISED RESPONSE TO NRC INSPECTION REPORT NOS. 50-327/87-06 and 50-328/87-06, ITEM ~NO,_CEB-12, AND SUBMITTAL OF SUPPORTIVE DOCUMENTATION FOR HIGHER CONDUIT DAMPING VALUES
References:
1.
TVA letter to NRC dated July 2, 1987, " Tennessee Valley Authority (TVA) - Division of Nuclear Engineering (DNE)
Design Calculation Effort for Sequoyah Nuclear Plant (SQN)"
2.
" Summary Test Report on Damping in Electrical Conduit" dated June.23, 1987 is our revised response to NRC Inspection Report'Nos. 50-327/87-06 and 50-328/87-06, Observation Civil Engineering Branch (CEB)-12.
In the previous response ( eference 1), TVA stated that we planned to develop and issue a summary report demonstrating the conservatism of the damping values used for the SQN electrical conduit.
TVA would then await NRC approval of the proposed damping values before revising the Final Safety Analysis Report (FSAR) and implementing the requested values in engineering analyses.
A copy of TVA's summary test report, reference 2, was given to NRC personnel in a meeting in our Knoxville offices on November 5, 1987.
Based on a thorough review of the relevant test data and the licensing dockets of recent or near-term operating license projects, a technical justification and precedence exist for the use of damping values for electrical conduit that exceeded those currently specified for TVA. is a brief report entitled " Justification of Damping Values Requested," which further summarizes the TVA basis for this higher conduit damping request.
Additionally, because SQN is an Operating License (OL) plant, any change.in damping values would require a revision to the SQN FSAR and a subsequent Unreviewed Safety Question Determination (USQD) for the effects of the revised criteria.
This USQD has been completed and is included as an attachment to enclosure 1.
8712300014 871222 DR ADOCK 05 g7
'\\'
An Equal Opportunity Employer
O l
% I U.S. Nuclear Regulatory Commission DEC 221987 In summary, supportive test data has been prepared by TVA and others demonstrating that the reasonable values of damping for electrical conduit should be 7 percent for rigid steel conduit.
Testing performed by TVA supports the conservative use of 15-percent damping for rigid aluminum conduit.
The proposed damping values are for application to the operational basis earthquake, safe shutdown earthquake, and design basis accident events. As an attachment to the enclosed USQD, you will find a draft of the FSAR revision necessary to support the damping values proposed herein.
The TVA technical staff is available to meet with the NRC staff for additional discussion on this request at any time convenient for you.
We would suggest an informal meeting forum for these technical discussions.
Very truly yours, TENNESS E VA AUTHORITY T
R. Gridley, Dir ctor Nuclear Licens ng and Regulatory Affairs Enclosures cc (Enclosures):
Mr. K. P. Barr, Acting Assistant Director for Inspection Programs TVA Projects Division Office of Special Projects U.S. Nuclear Regulatory Commission Region II 101 Marietta Street, NW, Suite 2900 Atlanta, Georgia 30323 1
i Mr. G. G. Zech, Assistant Director for Projects TVA Projects Division Office of Special Projects U.S. Nuclear Regulatory Commission 4350 East-West Highway EWW 322 Bethesda, Maryland 20814 Sequoyah Resident Inspector Sequoyah Nuclear Plant 2600 Igou Ferry Road Soddy Daisy, Tennessee 37379
a ENCLOSURE 1 Observation No. CEB Use of Variable Damping for Conduits TVA design criteria SQN-DC-V-13.10, seismically qualifying conduit supports, was revised on November 20, 1985, to include the span lengths and the support loads as developed in TVA calculation B41 851105 028. A review of this calculation showed that a variable damping ratio was used in determining the seismic loads on the conduit supports.
TVA used a damping value of 2 percent for frequencies greater than 10 Hz and 5 percent for frequencies less than 10 Hz.
TVA's commitment, as shown on table 3.7.2.4 of the Sequoyah FSAR, is a constant 2-percent damping value for the safe shutdown earthquake (SSE).
The use of a higher damping value would lower the conduit support loads and might yield an unconservative design.
TVA Response The fourth sentence in the observation is not correct:
1.
The FSAR reference is for piping, not conduit.
2.
The FSAR commitments for conduit are actually in section 3.10 and are given in the context of underground electrical conduit banks.
3.
The FSAR damping value referenced in section 3.10 is 1 percent, without reference to earthquake level.
4.
The 1-percent damping value contained in section 3.10 of the FSAR is out of place and was never used in design.
There is no explicit regulatory guidance for damping in electrical conduit.
Regulatory Guide 1.61, paragraph C.2, allows test-based damping to be used to justify values different from those given.
SQN FSAR section 3.7 (Seismic Design) lists damping values to be used for design of Category I structures, systems, and components (including large and small diameter piping), but does not explicitly define damping ratios to be used for seismic support of rigid metal conduit.
This resulted in the practice of treating rigid metal conduit in a fashion similar to small piping, including the use of Variable Damped Spectra, which the FSAR allows for small piping.
The analogy to small piping was originally reflected in design criteria SQN-DC-V-13.10 as 2-percent fixed damping for the SSE condition.
The design criterion was later revised to reduce the peak on the 2-percent curve to correspond to the peak level for a 5-percent curve.
Also, the spectrum used for input to the design criteria was broadened approximately 20 percent on the high frequency side of the peak and broadened to 0 Hz on the low frequency side. Problem Identification Report (PIR) SQNCEB8756 has been issued against this observation.
4 Based upon existing TVA test data and test data prepared by ANCO Engineers, Inc., in 1978 that has been accepted by the NRC staff on other nuclear projects, a technical justification has been developed for the use of damping values higher than suggested for piping by Regulatory Guide 1.61.
Since SQN already has its OL, the mechanism for evaluating the change to the design criteria is by an Unreviewed Safety Question Determination (USQD).
The determination that the revised damping criteria would not result in an unresolved safety question has been reviewed and approved within TVA (see attachment 1).
The FSAR change to reflect the revised electrical conduit damping values, along with other clarifications to the analysis criteria, has been prepared (see attachment 2).
1 i
i i
[
EP I,* -
~
QA Record
"~
UNREVIEWEQ,$AF ETY OUESTION DETERMINATION f
Damping Values for j
usoy
- identais, Electrical Conduit - SAR Balnr.'FI O
y,,
' To Ranii5vnN %efone WM mius Accession No.
g g '
'Rou ' '
b DATE N
mav Tor PatrAnto ntviewto ArPaovso APPO ALst I
\\
NO.
PP
,3 3
B ' * "'W W?W.ny/r; JBlk j
l 2
R 3
3 R
4 R
5 5
Prosect ord Affected Unit (d SQN Units 1 and 2 Dets of hment FCR/SCR/MCR/DCR Nureter L Dete ot Document Other Document identifier SAR Sections 3.7 6 3.10 See USOO Sheet No.
F
- See USOD Sheet No.
Potential Tech Spec Chge Species Aewirernents?
O ve.
Le; no O ve.
T u.
TVA Test Report No. CEB-BN-1028, dated 23 June, 1987.
us,-
ANCO Engineers Test Report No. 1053-21.1-4, dated 15 December,1978.
A change in the FSAR design basis for seismic damp g values for electrical conduit-Dewet on of chenee
'g Ely, to the following s
from 2% and 1% of critical for SSE and OBE, re dern indu t, accepted g
tabulation [ This change in the design criteri benecessaryM standards $1s required to minimize eine field modifications tha% ou as a result of revised in-structure response spectra coupled with the existing acceptance criteria for electrical conduit.
7%
~
Aluminum Conduit 15%
15%
I Continued on Sheet 3 cc (Anschmenuh RIMS, SL26 C K I
/
i.
3; 33,,g g TV A 1065110E 445)
\\
)
l s
l' UNREVIEWED SAFETY OVESTION DETERMINATION poet 2 hasauf.se Damping Values for
- t. is e. proe.unty of occurreace w she conwquence of an accideat or N'wacoon of eqwerneat importsat is wMy.-eM eve 8ved ia the sewy Aney ie Repc<t tausend?
Ov O so
~
Astification:
j Damping values in electrical conduit used in seismic analyses are important to the prediction of conduit stresses and support reaction loads. The damping i
values currently used are unrealistically low, thus resulting in predictions that are excessively large when compared to those realized during the Safe Shutdown Earthquake (SSE).
The proposed change in design criteria is based on the referenced test j
data, and in fact is still conservative, but far less than the current criteria.
j Consequently, the adoption of the more realistic damping analysis criteria for the qualification of electrical conduit will not increase the probability h
of occurence of an accident, or the consequences of an accident or malfunction of equipment important to safety.
- 2. le tre pomrutity for en occideat or ensifvacuoa of a different type thea eay evelveted prirnously in the Safety Analys&s Report created?
O ve. B no l
kstification:
The adoption of the more realistic d'ampin'g analysis criteria will minimize the need for additional supports, or modifications to existihg supports, to satisfy existing design criteria commitments for forces and stresses. Therefore, the adoption of this damping analysis criteria will not effect, or add to, the existing congestion within the plant and thereby does not introduce the g
j possibility of an accident or malfunction of a different type than any evaluated J
previously in the SAR.
1 is the trargia of safety es defined > the beeis for say technical asocificellori reduced?
C ves O No Astification:
The design of electrical conduit systems is not covered by a technical specification, and therefore does not effect the margin of safety as defined in the basis thereof. The adoption of the more realistic damping values will increase the margin of design allowable responses to those computed.
1ik f
4 9
sJ UNREVIEWED SAFETY OUESTION DETERMINATION sh 3
sepaufw Damping Values for SQN Units 1 and 2 Electrical Conduit - SAR
~CM below (a) the subject for this page Special Requirernent(s) or Precaution (s)
Initials Safety Evaluation Preparer BFH h
g Additionellnformation Reviewer I
Description of Change ~- Continued from Sheet 1 The adoption of the damping values for electrical conduit in excess of the USNRC Regulatory Guide 1.61 requirements has been approved for all SEP plants, and for plants with CP's that have requested it.
The test data that provides the basis for the rest of the industry was performed c * " w.-- w c: = " + -- - N by ANCO Engineers Inc./Bechtel Engineers.
The NRC Staff approved the use of the ANCO data on the plants with CP's with the caveat that they justify that their conduit materials and configurations conform to those tested by ANCO.
The testing performed by TVA was for TVA specific conduit and installation configur-ations, and it has been determined that the test specimens were indeed similar, i
The testing performed by TVA was done je without the knowledge of the ANCO i;
test program and results. *A comparison of the test results obtained by ANCO and TVA show a very close agreement of recommended damping values.
The ANCO tests on electrical conduit did not extend to aluminum conduit as did the TVA testing.
The damping values requested for electrical conduit is enveloped by both the TVA and ANCO test data results.
Other nuclear projects that have used the ANCO test data results, with the I
approval of the NRC Staff, to support similar increases in steel conduit. ares l
Limerick (docket 50-353) for 7%; Grand Gulf (dockets 50-416 & 50-4171 for 7%;
J and, Callaway (dockets 50-483 & 50-486) for,7%, (values shown are applicable for both OBE and SSE).
\\
3 V
e ATTACINENT 2 SNP CEB - 12 1:
..l\\
moments along the length of the CRDM are esiculated.
These values are then compared to the allowable seismic bending moments for the equipment, to' assure adequacy of the design.
79ggg7 A
3.7.4 Seismic Instrumentation Protran In order to assess the effects on the plant of earthquake s which may occur that exceed the ground acceleration for the 1/2 SSE (1/2 SSE = 0.09 ground acceleration) seismic instrumentation s:
provided.
The instrumentation program is described in the f oll owin g sections.
3.7.4.1 Concarison with NRC Rennistorv Guide 1.12 The instrumentation is described in Sec tion 3.7.4.2 b el ow and meets the requirements of NRC Regulatory Guide 1.12.
3.7.4.2 Location,and Description of Instrumentation The seismic instrumentation locations are s h ow n in Figures 3.7.4-1 through 3.7.4-8.
The instrumentation consists of the following:
1.
A strong motion triaxial accelerometer in the Unit 1 Reactor Building on the base slab at elevation 679.78 in the annulus tetween the Shield Building wall and the containment vessel j
as shown in Figure 3.7.4-1.
The full scale range of the r
transducer is from 03 t o 1.0g with a bandwidth of 0.1 B to 25 Hs and a temperature effect of less than 2 percent per 100*F change.
The accelerometer is connected to a battery-operated tape recorder which will record the accelerations (z.
y, and a) and a time reference trace on magnetic tape.
The recording system is located in the Auxiliary Building, see item 6.
A starter, see item 5, w il l initiate operation of the recording system.
2.
A strong motion triaxial accelerometer installed inside the Unit 1 Reactor Building on the floor slab at elevation 733.63 as s h ow n in Figure 3.7.4-2.
The accelerometer, which is
~
identical to the one described in item 1, is oriented with
]
its recording axes sligned with the accelerometer discussed i
in item 1.
3.
A strong motion triarial accelerograph inside the Diesel Genera t or Building on the base slab at elevation 722.0 as s h ow n in Figure 3.7.4-3.
The accelerograph consists of three aecclerometers (z.
y, and :) identical to those in items 1 i
and 2, a cassetic-type magnetic tape recorder vith a nazient recordina time of 30 minutes and a t rigg e r which actuates the recording sy st em vhen an acceleration of 0.01 g or greater (L. l 3.7-46 i
1 w_______.
.J
. e.
-w' SNP
- 4. *f.
17.
Broms, Bengt B., D e s i g n o f L a t e r a l l y L e a d e d F i l e s,
{R 'l
_ J ou rnal of the Soil Mechanics and Foundation Division, ASCE, No. SM3, Ma y 1965, pp 79-99.
ItJSERT C
t
.f#
4
-k.
3.7-55
e w
TABLE 3.7.1-3
' f".
. {,.
DAMP 1NG RATIOS USED IN ANALYSIS OF CATEGORY I STRUCTURES SYSTEMS, COMPONENTS, AND SOIL FOR STRUCTURES LISTED IN TABLE 3.7.1-1 Damping Ratio, Percent of Critical Viscous Dampint 1/2 Safe Shutdown Safe shutdown Item Earthquake Earthquake Steel Containment Vessel 1
1 1*
Concrete Shield Building 2
5 7
and Internal Concrete l
St ruc ture Other Welded Steel 1
1 2
St ruc tures Bolted Steel Structures 2
2 5
other Reinforced Concrete 5
5 7
St ruc tu res Bolted or Nailed Wooden 5
5 5
St ruc tures Damping for Determining 10 10 10 Amplification through Soils for Soll Supported Structures Vital Piping System **
0.5 0.5 1
l3 E LE(.TPdC 4L Co.G mT -
STEEL 7
j
.L l
AtuM Wuh 15
<G
'd.
{
- Damping values used when stress levels are at or near yield. All other damping values are for lower stress levels.
- Response spectra were also computed f or f requency variable damping of 5 percent to 10 hertz, decreasing linearly to 2 percent at 20 hertz, and remaining at 2 percent to 33 hertz as described in ASME Code Case N-411.
Variably damped spectra used in piping analyses were not modified to include higher damping permitted by U.S. NRC Regulatory cuide 1.61.
Variable dampin; was used in new analyses as well as reanalyses (reconelliation work) using the response spectrum method but not'in time 3
history analyses.
To account for possible increased flexibility when Code. Case N-All damping was used, piping system displacements were checked for adequate clearance with adjacent structures, components, and equipment. Also, i
involved equipment was checked to withstand increased motion.
I r
g, k~
Revised by Amendment 3
_ __=___
e f \\ l]
t' G
f SNP-3
/
s TABLE 3.7.2-4 j
DAMPING RATIOS USED IN ANALYSIS OF CATECORY I J'
4
- y STRUCTURES, SYSTEMS, AND COMPONENTS FOR STRUCTUItES LISTED IN TABLE 3.7.2-3
'.5yy f
+
f (percent of Critical Damping)
)
(
, l[/
\\.
C.J s
"f
~
I
/
p 1/2 Safe Safe Shutdown Structure or Component Shutdown Earthquake 2 Earthquake gi
\\p '
(
I 1
, l'
- Equipment and large-l3 diameter piping systemss,
,2 pipe diameter greater p.
than 12 inches 2
3 e
t1
- Small-diameter piping g3 systems, diameter equal to or less than 12
(. _
I 2
L inches....................
Welded steel sttvetures 2
4 3
Bolted steel structures.....
4 7
Prestressed concrete
,1 f
structures..................
2 5
'r
'f tJS E5?T l
Heinforced concrete p
{
r structures.
4 7
t l
- As an option, for some cases of piping response spectrum seismic analysic, variabic damping of 5% to 10 hertz decreasing linearly to 2% at 20 hertz and remaining at 2% tu 33 hertz was used for both 1/2 SSE and SSE as described in ASME Code Case N-411.
Variably damped spectra used in piping analyses were not modified' to include higher damping permitted by U.S. NRC Regulatory Guide 1.61.
Variable damping was used in now analyses as well as reanalyses (reconciliation work) using the response specuum method but not in time history analyses.
3 To account for possible increased flexibility when Code Case N-411 l
damping was used, piping system displacements were checked for D,
l adequate clearance with adjacent structures, components, and I
equipment. Also, involved equipment was checked to withstanf
)
'?
increased motion.
l v
I Revised by Amendment 3 5
s
1 SEP '4 f 7 i
i 3.
Interim Acceptance Criteria f
[
An interim acceptance criteria is being used for temporary resolution of
!? '
cable tray support issues for the restart of Sequoyah Nuclear Plant (Phase a
I).
The criteria involve changes to the original FSAR Licensing Commitments. Application of this interim acceptance criteria is limited to cable tray supports.
TVA'sutilizationofthisinterimacceptancecriteriaisdiscussedthi greater detail in the following docwsents.
1.
R. L. Celdley's letter to B. J. Youngblood dated August 18, le8?s 4
(L44 860818 803).
F 2.
Volume 2 of the Nuclear Perfomance Plan which addresses Sequoyah 3.
B. J. Youngbloods letter to S. A. White dated December 5,1986 (A02 861210 008).
4.
R. L. Cridley's letter to B. J. Youngboood dated January 14, 1987 (L44 p61105 801).
AuAlv515 (bucT) ottat)
)
A conduit Banks fitn$
s y
4 estraint Measures
~
/
t The Category ctrical exposed conduJb.-su60rts, and electrical conduit bM et 6 ten designoO to provide vertical and horizon 1%cQe spacing occommended in TVA Count p l Specification.
y DELDE
'5_
j
'f The Category I underground electrical conduit tanks which run from the auxiliary building to the diesel-generator building, to the auxiliary cooling towers, and to the pumping station were analyzed by two methods, The conduit banks were first analyzed as a beam with unconstrained ends, on an elastic foundation and were found to have the same motion as the soll deposit in which they were buried.
Secondly,thesolldepositwasthenassumehtobean infinitely long uniform soil deposit resting on a rigid
)
foundation which responds to earthquake motion by' moving a.
continuous sinusoidal plane wave. The displacement' of the soil is 4
f; 3.10-7 G.
J
$qr b
SNP
/
2
=
a (2. n )
(DLF)
D where a=Maximena roch acceleration T-Fundamental period of'so!! deposit DLF= Dynamic amplification factory which is the maximum entface acceleration of the soil deposit divided by the maximum rock accel-erstion The fundamental period and the DLF of the soit deposit were found by modeling the soil deposit as an elastic medium and making a dynamic anstydis cf a s11ce of unit thickness using only the horizonal shooring resistance of the soil.
The wave 1ergth (Y) is W = T V,7,
3 where VST = Avere,te shear wave velocity of the soit daposit Using the results from the above equations, the bending moment due to the earthquake is 2
n M = E1D (
)
where E*Eoung modelas of conduit bank I=Nonent of inertia of conduit bank L=0ne-half of the wave length e period, maximua sof. sale displacement, maximum strees, and rea ons were de ce rni ned for a range of lateral suppori spectags qasidered sigt.ificant.
The m aximum suppor pacing is 10 feet.*-IAe 1 e r, d s, u s e d to design the suppor were d e r iv e d by: vs l a g emga x :1 m um response spectru his spectr,um was developc6'hy t a hlYs t h e maximum sei e acceleration say period from the two hor 3s al res se acceleration g rg
((
spectra.
This)was combir.ed w the maximum vertical ULL reuponse a c c e l'e r a t i o n sysS&(a at ercent damping.
The-first four natural po ds of the con were detstmineo and 44,4'1 th or periods the corresponding sponse for e ack of computed.
T effects of higher numbered natur eriods
,y
.s s t e c o n,Sef v a t i v e l y considered by sett(ng the fourt erlod rospetse equal'to the resonance response after the four
{
,3 ss'Ariod esceedeo resonance.
The modal participation factors v
u seer g
b 3.10-3
.m_. _. _ _.
__.__._-_____.__._.m
i s'
SNP-4 for eac e calculated and the important design d N enEF stress intensificat on f 2.3 was u ed-fMreaded joints and
{.';
- they were assumed to be loca d-a f maxisua moment. Where conduit tennin conduit box, the conduit attachment rested as a support point.
DELE TE Welds Welding for structural supports was in accordance with the American Welding Society " Structural Welding Code." AWP D1.1-72 as implemented by TVA General Construction Specification C-29C. Nuclear Construction Issues Croup documents NCIG-01. Revision 2. may be used after June 26, 1985, to 4
evaluate weldsents that were designed and fabricated to the requirements of AISC/AWS. When invoked. NCIC provisions will be implemented as indicated in section 3.6.8.
i l
4 a
e I
3.10-9 i'
~
e g
INSERT A 3.7.3.16' ELECTRICAL CONDUIT DAMPING VALUES For electrical conduit with cable'and their_ supports, the damping values'for both OBE and SSE shall be 7 percent for steel conduit and 15 percent for aluminum conduit.
The damping value for steel-.
conduit is based upon tests performed by Bechtel (20), reviewed by the NRC (21), and published in technical:11terature'(22, 23).
The I
configurations. tested by Bechtel'are similar.to those used at I
Sequoyah.
In addition, TVA-performed testing (24) of typical installation configurations supports the Bechtel test conclusions on:
steel conduit, and developed the basis for the: aluminum conduit by using identical test methods as.for steel.
i j
l l
1 i
i i
d_.
_..______________________._________i.__u______._J
INSERT B CONDUIT SUPPORTS Design criteria have been established for support design and spacing for all conduit mounted in Category I structures other than directly on to the SCV.
s The period, maximum seismic displacement, maximum stress, and reactions were determined for a range of support spacings.
The loads used to design the supports were derived by using a maximum response spectrum.
This spectrum was developed by enveloping the two horizontal in-structure response spectra, for the appropriate building elevations, and then added to the vertical in-structure response spectra. A stress intensification factor was used for the threaded joints, and they were assumed to be located in the region of maximum conduit bending stress. Where the conduit tarminates in a conduit.
box, the conduit box and its attachment have been trsated as a support point.
For situations where the use of the design criteria loads is not practical,- a rigorous analysis may be performed in accordance with the following criteria.
1.
The damping values for dynamic seismic analyses are defined in subsection 3.7.3.16.
2.
For each direction of motion, the individual modal responses are combined by the square-root-of-the-sum-of-the-squares method.
3.
The seismic design loads are based on the absolute addition of.the worst horizontal direction response with the vertical direction's response.
For electrical conduit runs mounted to the steel containment vessel (SCV) and subjected to the dynamic effects of a postulated loss of coolant accident (LOCA) event, a rigorous analysis will be required using the following additional analysis criteria.
1.
The damping values for use in the evaluation of dynamic loads will conservatively be taken as the same values as for the SSE level defined in subsection 3.7.3.16.
2.
The individual modal responses from three components of input, along with the cross components of response, are combined together by the square-root-of-the-sum-of-the-squares method.
3.
The resultant loads, stresses, and displacements are absolutely added to seismic induced effects and then combined with the self-weight effects.
4.
For electrical conduit runs that are subjected to both the dynamic LOCA effects and the differential support motions induced by accident thermal conditions,.two separate loading conditions are used. Accident thermal effects are not combined with dynamic accident effects.
e INSERT C
- 18. ANCO Engineers, Inc., Cable Tray and Conduit Raceway Seismic Test Program - Release 4 (Final), Test Report 1053-21.1-4, December 15, 1978.
19.. Letter from R.-Kosiba to Dr. Franz Schauer (NRC - Structural Branch),
-January 8, 1980.
(Test Reports were provided by Bechtel:to the NRC
~ following a meeting held on January 8,1980, in Bethesda, Maryland, to discuss the test report-results on a generic basis.)
20.
P. Y. Hatago and G. S. Reimer, Dynamic Testing of Electrical Raceway Support Systems for Economical Nuclear Power Plant Installations, Presented at the IEEE PES Winter Meeting, New York, New York,.
February 4-9, 1979.
21.
R. B. Linderman and A. H. Hadjian, " Development'of Bechtel Electrical Raceway System Test Program," Proceedings of the American Power Conference, 1981.
22.
B. B. Neely, et al., " Summary Test Report on Damping in Electrical.
Conduit," TVA Report CEB-BN-1028, June 23, 1987, l
l
o INSERT D 1/2 SSE SSE Electrical Conduit - Steel Aluminum 7
7 15 15
\\
i
e 1
ENCLOSURE 2 This report will provide a brief explanation of the basis for which higher damping values are technically justified for electrical conduit.
The report examines the available tests data and compares it with TVA-derived data on electrical conduit damping.
The Sequoyah criteria and FSAR commitments imply that a deterministic design basis analysis be performed for the electrical conduit.
The acceptance criteria for conduit stresses are based on American Institute of Steel Construction (AISC) allowables with an intensification factor added for the threaded connections.
Fragility tests for structural and functionality performed on three span conduit runs of different sizes showed that their inherent capacity is in excess of a Regulatory Guide 1.60 ground spectra anchored at 0.75g (reference A).
Thus, the excessive conservatism introduced by the use of Regulatory Guide 1.61 damping for piping is not justified for-l electrical conduit since sufficient margin exists over the peak ground accelerations expected at the TVA sites.
The NRC staff has recognized this fact by accepting the ANC0/Bechtel testing data (reference A) as the licensing basis for projects that have requested it.
1 The damping values and seismic fragilities observed by ANCO (reference A) exceed those developed by J. A. Blume and Associates (reference I) for the I
systematic evaluation plants (SEPs).
This difference is expected when the test setups and test techniques are compared.
The ANCO test data was collected on specimens in Phase VII of their testing program that were attached directly to the rigid test frame; and in Phase VIII, stiff frame conduit supports typical of " modern" construction were tested. Conduits were introduced into the Blume testing in Phase IV (section 3.2 and figure 3.10) as a threaded rod hung system that is typical of SEP vintage plants.
For seismic motions, the threaded rod hanger will act as a base isolator and thereby reduce the base input accelerations to the test specimens. As a result of this mounting technique, the resulting mechanism is a very low damped system.
Examination of the Blume data from Phase IV in table 3.4 confirms these expectations. The tests on rigid steel (RS) conduit are numbered 150 through 168. Tests 150, 151, and 152 are for a 2-inch conduit supported by threaded i
rod with bracing at both ends and thereby beginning to approach a rigid l
support condition tested by TVA and ANCO.
The damping values from tests 150, 151, and 152 were 9, 8, and 11 percent, respectively.
It should be noted here that both ANCO and TVA recommended 7-percent damping for steel conduit.
Tests 153 through 158 have bracing at one end or none at all, and their associated damping values are about half of those from tests 150, 151, and 152.
Tests 159 through 168 were for empty conduit, and therefore are not relevant.
The measured frequencies for the Blume tests 150 through 168 range from 3.7 Hz to 4.8 Hz for 2-inch conduit.
The ANCO frequency data on 2-inch RS conduit ranged between 11.5 Hz and 12 Hz. Therefore, the Blume data is for the most part only applicable to threaded rod hung systems and not the seismically designed supports typical of modern construction and the TVA design.
b____________._______________________________.___..____.___
e
,, The caveats imposed by the NRC staff reviewers on projects using the ANCO data (references E, F, and G) for electrical conduit consist of comparing the material and the installation configurations used with what was tested by ANCO.
TVA has exceeded the NRC. requirements by performing independent testing of standard materials and configurations in use at TVA sites that also conform to what ANCO tested.
The testing by TVA was performed without knowledge or benefit of the ANCO work.
This had the positive aspect of total independence I
and not biasing the results.
An examination of the testing techniques involved will aid in the interpretation and comparison between the TVA and ANC0 test data.
The TVA testing was based primarily on single span specimens versus three span specimens by ANCO.
TVA based the damping value data on snapback testing versus variable' amplitude slow sine sweeps from base excitations by ANCO.
These differences in test specimens and techniques produce a significant difference in the resulting damping values.
The single span conduit (beam) tested by TVA will have modes of vibrational response that differ from a multispan, base excited, continuous conduit run tested by ANCO.
If the spans were equal between a single span beam and a multispan beam, the fundamental modes computed analytically would be identical.
The fundamental mode shape would resemble a half sine wave between each support.
In each successive span of the multispan beam, the Eigen vectors would change sign to produce a regular serpentine shape.
For even numbers of spans, the area underneath the fundamental mode relative to the undeformed shape will add to zero.
For odd numbers of spans, the area under the fundamental mode relative to the undeformed shape will not be zero, but will be very small.
For the case of uniformly distributed mass as it is for conduit, these comparisons directly reflect on the modal participation factors. A mode with a zero, or near zero, participation factor will not influence the dynamic response of the structural system when compared with the contributions from the other modes that do respond. ANCO noted in their report in section 8.2 that the snapback testing technique will excite modes of response that base excitation methods cannot.
From snapback tests of their own, ANCO noted that these first few nonparticipatory modes will have damping values in the range of 1 to 5 percent. The testing performed on RS conduit with cable by TVA was solely for first mode damping and exceeded the expected range of damping stated by ANCO.
l The TVA materials and test configurations on RS conduit comprise a subset of l
the ANCO test program on conduit.
Because the test specimens used were representative of TVA site installations and the first mode damping values recorded by TVA envelop the ANCO recommended damping values, it is concluded that for similar test specimens the testing techniques must be comparable as used.
Neither ANCO nor Blume tested aluminum conduit to determine their damping values as did TVA.
The rigid aluminum conduit is schedule 40 pipe as is the RS conduit. However, the ratios of cable weight to total weight are l
.__._______.____m_._.____m.-_____
_ _ _ _ _ _ _ _ _____m.
t
.. substantially different. As an example, for a 4-inch conduit made of aluminum
)
and steel, the respective 100-percent cable fill weights to total weights are 0.71 and 0.45.
Similarly, the ratios of the dynamic properties (EI/m) for I
aluminum and steel 4-inch conduit are 2.66E+07 and 4.93E+07, respectively.
The testing by ANCO did include electromagnetic tubing (EMT) that is -
manufactured from 14-gauge steel sheet metal and has weight ratios and dynamic properties of 0.70 and 2.33E+07, respectively.
By comparison of the weight ratios and dynamic properties, it can be seen that the EMT very closely resembles the rigid aluminum conduit. Damping ratios for EMT determined by i
ANC0 were "about twice that of the RS," but with about 1 30-percent scatter about the mean.
This randomness is explained by the unpredictable nature of the coupling devices used on EMT that make use of set screws versus the j
threaded end connections of rigid aluminum or steel condult.
Even though this j
is not a perfect comparison between like conduit, it does show the increased 4
influence of the cable weight ratio on the effective damping ratio.
Because the test methods used by TVA to determine the damping r:tios of rigid aluminum conduit were identical to those used on the RS conduit and a parallel i
comparison can be drawn to existing test data by ANCO, the use of a damping ratio of 15 percent is justifiable and yet still conservative.
The reference nuclear projects (references B through G) have been licensed to use a single value of the ANCO-derived damping on conduit for both the operational basis earthquake (0BE) and SSE seismic events.
Damping values are therefore independent of ground totion acceleration levels.
The basis for this position l
was established through low level sine sweeps of between 0.05 g and 0.20 g and I
i the observation that, Leyond a very minimal input acceleration level, the damping values tended to approach asymptotically some higher value.
The j
analytical benefits of increased modal damping are only obtained for systems that have modes of response in the frequency band of amplified structural responses. The accelerations in the horizontal directions in the region of amplified responses are on the order of several g's; this is typical of most projects.
For conduit systems whose frequencies are above the region of structural amplification, the selection of a damping value is not important because there will be little or no difference between the low and highly l
damped response spectra curves.
The transition from the amplified response region is usually fairly abrupt.
A copy of the TVA summary test report (reference H) was given to the NRC in a meeting in TVA's Knoxville offices on November 5, 1987.
The report contains a number of plots to display the damping data relative to frequency and initial strain levels in the test specimens. These plots presented all of the data collected.
If the plotted data from the conduit specimens are removed from those that were without cable in them, the very low stressed conduit, and those that are dynamically rigid (above 33 Hz), the evaluation of the collected data becomes readily apparent.
Figure 10 from attachment H presented the test data on RS conduit whose initial snapback stress level was 3 kilopound per square inch (ksi) or above.
Figure 10A.(attached) presents the same plot, but has had the damping data additionally screened for the i
i
__ - _ _- _ _o
e i
empty and the dynamically rigid test specimens. Similarly, figure 14 from attachment H presents the test data on the rigid aluminum conduit data screened for 2 ksi initial stress or above.. Figure 14A (attached) presents the same plot, but has had the damping data screened for the empty and the dynamically rigid conduit test specimens. On both figures 10A and 14A, the plotted numerals indicate the conduit specimen diameter. A review of the test data corresponding to the 2-inch RS conduit on figure 10A that had damping values less than the requested 7-percent damping shows that signal cable was used in lieu of power cable.
This produces a significant difference in that (1) for the same cable fill ratio, the signal cable weighs half as much as does the power cable; and (2) the plastic jacket covering each bundle of wires is very slippery as compared to the energy-dissipation insulation on power cables.
Therefore, the data for the 2-inch RS conduit should be screened out.
The five 3-inch rigid aluminum conduit data points in figure 14A that lie below the requested 15-percent damping value are for the most part in the dynamically rigid frequency range.
By 25 Hz, there is little or no discernable separation between response spectra curves of different damping values at the TVA sites.
In summary, supportive test data Fas been prepared by TVA and others demonstrating that the reasonable values of damping for electrical conduit should be 7 percent for RS conduit and 15 percent for rigid aluminum condult.
The proposed damping values are for application to both the OBE, SSE, and DBA events.
References:
A.)
ANCO Engineers, Inc., Cable Tray and Conduit Raceway Seismic Test Program, Prepared for Bechtel Power Corporation, Report No. 1053-21.1-4, Release 4, December 15, 1978.
B.)
Final Safety Analysis Report for Callaway Units 1 and 2, Docket Nos. 50-483 and 50-486, Section 3.7.
C.)
Final Safety Analysis Report for Limerick Units 1 and 2, Docket Nos. 50-352 and 50-353, Section 3.7.
D.)
Final Safety Analysis Report for Grand Gulf Nuclehr, Station Units 1 and 2, Docket Nos. 50-416 and 50-417, Section b 7.
E.)
USNRC, Safety Evaluation Report Related to the Operation of Limerick Units 1 and 2, NUREG-0830, Supplement 1, January 1982, Section 3.7.
F.)
USNRC, Safety Evaluation Report Related to the Operation of Limerick Units 1 and 2, NUREG-0991, August 1983, Section 3.7.
1 i
7 dB Pego 5 of 5 G.)
USNRC. Safety Evaluation-Report Related'to the Operation of Grand Gulf Nuclear Station Units 1 and 2, NUREG-0831, September 1981, Section 3.7.
H.)
.B. B. Neely,'et.al; Summary Test Report on Damping in Electrical Conduit, Prepared by TVA, Report No. CEB-BN-1028,..
June 23, 1987.
I.)
URS/J. A. Blume and. Associates, Engineers, Shaking-Table Testing for Seismic Evaluation of Electrical Raceway Systems... Report No. URS/ JAB 8050, Prepared by..the SEP. Owners Group, April.1983.
Prepared by:
h
'Date:
//. 30-Q B. F. Henley
/
/dY"k J
Date:
Reviewed by:
D. J. Dombroski Approved by:
Date:
/4-t-#7'___
B. Bf Neely
^
l 1
5 3
)
f f
o
...5 t.
u.
C 0
. 3 i
s.
p.
e 0.
u l
0.
a 5
03.
v
. 2
(
g n
g.
i p
n.
m
)
a i.
d t z
r 0 t i
i.
d u i 2 r
W.
ed e
t n H
h.
s o
(
e c t.
u ql y
iW.
e e c
5 r e 1
n t
t.
ss e
u i
u, R r q
V o e
d.
TF r
n 4 Q 4
F o.
4 0
C.
1 l
5 5 5 3
3 3 e.
e t.
k22 S.
5 r.
o.
f s.
t l
0 u
s0 5
0 5
0 5
g e3 2
2 1
1 R
xv o pg& mC~ hep
~
ym@%, 5>
)
5 f
3 f
o t
- 7E1-g
- j :1: 7 E
- Z-3 u.
3 3
C 3
0 i
. 3 s.
p.
0.
3 e
1 0.
u 0
l 2.
a 5
(
v i 2 g.
g n
i n.
pt i.
mi
)
r au i.
d d z
t n
0 W.
d o
, 2 r
h.
e c e
t H
t.
s m
(
e u i
W.
u n y
qi e m c
t.
r u 5 n 5
1 5
l i
e u.
s a u
d.
Ar q
n V o e
o Tf 5
r C.
F 0
m.
55 3
3 1
g u.
3 n
i.
5 m.
u l
A.
. 5 r.
o.
f s.
t.:-
0 E-1:
- E E-.:- : E-
_E 5
0 5 5
0 5 0 5
0 5 0
5 0 5 g l
u7 7
6 5
5 4
4 3
3 2
2 1
1 s
n*N.v 03em GC aEeA e
R 39g' $>
I l
l