ML20206J483
| ML20206J483 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 06/24/1986 |
| From: | Wilson R GENERAL PUBLIC UTILITIES CORP. |
| To: | Zwolinski J Office of Nuclear Reactor Regulation |
| References | |
| 5000-86-0926, 5000-86-926, NUDOCS 8606270178 | |
| Download: ML20206J483 (36) | |
Text
t GPU Nuclear
~
glW
{
100 lnlerpace Parkway parSippany. New Jersey 07054 201 263-6500 TELEX 136-482 Writer's Direct Dial Number June 24, 1986 5000-86-0926 Mr. John A. Zwolinski, Director BWR Project Directorate No.1 Division of BWR Licensing U.S. Nuclear Regulatory Commission Washington, D.C.
20555
Dear Mr. Zwolinski:
Subject:
Oyster Creek Nuclear Generating Station Docket No. 50-219 Responses to Draft Technical Evaluation Report (TER) for SEP Integrated Plant Safety Assessment, Section 4.11, Seismic Design Consideration Your letter dated January 9,1986 transmitted a draft TER prepared by an NRC contractor. The draft TER is based on the contractor's preliminary review of GPUN's response to sub-topics of SEP Topic No. III-6, " Seismic Design Considerations". The sub-topics reviewed by the TER are seismic assessments of piping systems and electrical equipment.
The draft TER identified several areas where additional information was needed for the staff to complete their final assessment of the sub-topics.
In our recent meeting with your staff on April 24, 1986, we discussed our draft responses to the concerns raised in the draft TER. The staff found most of the draft responses presented in the meeting were sufficient and required no further evaluation. The staff requested that the following additional information be provided in our formal response to the staff:
(1) provide the specific spectral acceleration referred to in item 5 of the draft response, (2) provide the parts of reference "i" referred to in items 4, 5 and 6 of the draft response, and (3) add a discussion about the general location of the liquid poison piping in item 8 of the draft response.
8606270178 860624
[0 (
PDR ADOCK 05000219
)
g P
PDR l'
\\
GPU Nuclear is a part of the General Public Utilities System
6 The attached are our revised responses to the concerns raised in the draft TER. Some of the draft responses have been revised in accordance with agreements reached at the April 24, 1986 meeting and include the additional infonnation requested by the staff. We believe that the responses are sufficient for the staff to complete their final assessment of the sub-topics covered by the draft TER.
ry uly yours, R.
Vice President Technical Functions RFW/pa(3530f) cc: Dr. Thomas E. Murley, Administrator Region I U.S. Nuclear Regulatory Commission 631 Park Avenue King of Prussia, PA.
19406 NRC Resident Inspector Oyster Creek Nuclear Generating Station Forked River, N.J.
08731 J. Donohew U.S. Nuclear Regulatory Commission 7920 Norfolk Avenue Bethesda, Maryland 20014 l
r i
0YSTER CREEK SYSTEMATIC EVALUATION PROGRAM Responses to NRC/EG8G comments on seismic analyses submitted by GPUN, contained in NRC letter dated January 9,1986 i
P
)
)
i I
+9 i
I l
{
s CRD PIPING REANALYSIS 1.
Comment:
MPR ratioed piping stress values calculated by EG&G to obtain results for different seismic accelera-tions.
The stresses ratioed should not have included pressure and deadweight stresses.
Response
We agree.
Howe ve r, because the non-seismic stresses are small relative to the seismic stresses, the stress results are within allowables by either method (EG&G agrees with this conclusion as stated on page 3 of of Reference a).
2.
Comment:
Credit should not be taken for friction in U-bolt pipe supports.
Response
The EG&G analysis of the CRD return piping was based on the use of a Reg. Guide 1.60 spectra with a peak ground acceleration of 0.22g and no axial restraint at U-bolt supports.
In the re-analysis of the CRD return piping by MPR, two cases were considered, a.
Reg. Guide 1.60 spectra with a peak ground acceleration of 0.22g and axial restraint at U-Bolt support at Node 1105.
Stresses exceed ASME code allowables.
b.
Site specific spectra with a peak ground acceleration of 0.165g and axial restraint at U-Bolt support at Node 1105.
Stresses are within ASME code allowables. -
The effect of the site specific spectra on the stress analysis results for the CRD piping can be obtained by a
taking the dif ference between the above two cases.
The reduction in stress due to the use of site specific spectra is approximately 2.5 to 4.4 times greater than the reduction in stress due to axial restraint at U-Bolt support at Node 1105, depen' ding on the location.
EG&G has estimated the stresses in the CRD return piping based on site specifi'c spectra, but taking no credit for axial support at U-Bolt support at Node 1105 due to friction. These stresses are within ASME Code allowables, as indicated in Table 2 of Attachment 2 of Reference a.
Thus, stresses are acceptable without taking credit for axial restraint at U-Bolt support at Node 1105.
3.
Comment:
Stress results were based on a model of a section of EG&G's full piping model.
The stress results developed using the shortened model appear acceptable provided the response of the shortened model is comparable to EG&G's full piping model.
Response
As described in the re-analysis of the CRD return piping (Reference b), a comparison of response was made between the shortened model and EG&G's full piping model.
As indicated in the re-analysis, this comparison showed the two models to have a similar response.
Results of this comparison are presented in..
~
ANALYSIS OF 4160 V SWITCHGEAR AND 460 V UNIT SUBSTATION CABINETS 4.
Comment:
The analysis of the cabinets utilizes a simple single degree of freedom model.
Justification should be provided to show the method of analysis usirig a simple model is conservative.
Response
For base mounted cabinets such as these, seismic response is dominated by the fundamental frequency.
This applies to both the horizontal and vertical directions.
As a result, modeling them as a single degree of freedom system results in an accurate representation of the cabinets.
This approach is consistent with that used by the NRC's consultant, Structural Mechanics Associates, who evaluated similar Oyster Creek cabinet anchorages in Section 4 of Reference h.
It is also consistent with more recent anchorage evaluation guidelines being developed by EPRI and the seismic Qualification Utility Group for similar equipment, as reported in Section 3 of Reference i.
These guidelines are based on evaluations of electrical equipment cabinets using the response spectra method.
The evaluations show that a single degree of freedom model results in an accurate represen-tation of the cabinets.
5.
Comment:
The analyses utilized equivalent static accelerations at the zero period acceleration in the vertical direction and the peak of the response spectra curve in the horizontal direction.
These values should be multiplied by 1.5 to account for the fact that a J
simple model was used in the analysis.
3-1 i
1
U
Response
Dynamic characteristics of base mounted cabinets are such that additional load factors to account for participation of higher frequency modes are unnecessary.
This has been demonstrated in evaluations reported in Section 3 of Reference i (included herein as ) in which dynamic analyses, using the response spectra method, were conducted on generic base mounted cabinet structures.
Results of these evalua-tions indicated that the maximum base shear was always less than the weight of the item times the spectral acceleration (either the peak spectral acceleration or the amplified acceleration at the fundamental frequency of the unit), and that the maximum overturning moment was always less than the weight of the item times the spectral acceleration times the distance to the center of gravity.
On this basis, a ststic load factor of 1.0 is appropriate.
These results weis accepted by the Senior Seismic Review and Advisory Pa-'el (SSRAP) and the NRC in recent meetings to discuss the esults of EPRI sponsored tasks in developing anchorage-guidelines, Reference k.
It should be noted that t.1e anchorage evaluations of the Oyster Creek cabinets presented in Reference d were based on the zero period acceleration in the vertical direction and the peak of the response spectra in the horizontal direction, as indicated in the EG&G comments, abo ve.
6.
Comment:
The electric equipment seismic analyses should have used 5% damping in lieu of 7%.
Response
The bases for using 7% damping in the seismic evaluation of the Oyster Creek electrical equipment are as follows:
1
For bolted assemblies similar to the electrical cabinets, a value of 7% damping for SSE loading is in accordance with Table 1 of NRC Regulatory Guide 1.61 and Table 1 of IEEE-344.
A value of 7% damping was recommended and used by the NRC's Senior Seismic Review Team for analysis of both electrical and mechanical equipment as documented in Paragraph 6.2.2 of Reference j, the NRC's report of the seismic evaluation of Oyster Creek.
For stress levels no more'than half of yield (working stress), a damping value of 7% is recom-mended as the "value that should be used in design when moderately conservative estimates are made of the other parameters entering into the design criteria",
as stated in Paragraph 5.2 of Reference 1.
For stresses at or just below yield, the recommended damping increases to 10 to 15%.
Guidelines for anchorage evaluations of electrical equipment cabinets currently being developed by EPRI and the seismic Qualification Utility Group, Reference i, remove excess conservatisms that were used in the original SEP evaluations discussed above.
Although review and acceptance of these guidelines by the NRC and the SSRAP is on-going, the general methodology and significant criteria changes have been accepted.
Additional analyses were performed for the 4160V and 460V cabinets using these guidelines to compare the results with those performed for SEP.
Results of these analyses are presented in Attachment 2 and indicate that cabinet anchors are adequate for seismic loads at both 5 and 7 percent damping.
Strength values used to evaluate anchorage adequacy were taken at one quarter of the mean strengths reported in Reference i.
~
LIQUID POISON PIPING 7.
Comment:
The analyses of the liquid poison piping utilized two models which were interconnected but not overlapped.
Response
The liquid poison piping analyses utilized several models in order to conform to computer code size limitations.
In the case of the two models referred to by EG&G, one model included overlap and the remainder of the attached piping system not included in the model was represented by an equivalent' spring.
In the second model, overlapping was not included because, after examining the locations of rigid supports, neglecting the attached piping was considered conservative.
A detailed discussion of the liquid poison piping models i
is given in Attachment 3.
I i
8.
Comment:
Sei?mic anchor motions (SAMs) were not con-sidered and should be included in the liquid poison piping analysos.
Response
SAMs were considered and judged to be not significant for the liquid poison piping analyzed by MPR in Reference e.
The main reason for this is that the portion of the liquid poison piping system analyzed in Reference e is anchored to either the reactor building structure or drywell structure at essentially the same elevation.
Both of these structures are supported from the same foundation mat.
Thus, significant anchor dis-placements between the reactor building and drywell structures are not expected.
1.
9
f This approach is consistent with the approach used by EG&G in their analyses of Oyster Creek piping in neference g (i.e., EG&G did not include SAMs in their analyses of Oyster Creek piping either).
i
.i l
1 i
(
I L
i i
't l
I
, u
I 4
i
~
CORE SPRAY PIPING f
9.
Comment:
Seismic anchor motions (SAMs) were not considered and should be included in the core spray l
piping analyses.
i
Response
SAMs were considered and judged to be not j
significant for the core spray piping analyzed by MPR in Reference c. The main reason for this is that the l
portion of the core spray piping system analyzed in Reference c is attached to either the reactor vessel or drywell structure.
Both of these structures are supported from the same foundation mat.
- Thus, l
significant relative anchor displacements between the j
reactor vessel and drywell structures are not predicted for piping in the drywell.
This approach is consistent with the approach used by i
EG&G in their analyses of Oyster Creek piping in Reference g ( i.e., EG&G did not include SAMs in their l
analyses of Oyster Creek piping either).
l 4
I 1
l i i e
e
]
i t
__.. -.. - = -
~
i r
4 MAIN STE AM AND FEE DWATER PIPING SUPPORTS I
f 10.
comment:
Current support drawings were not included'in l
the MPR Report (Reference f).
Thus, it was not possible to independently verify or spot check the correctness of 1
i j
component data.
Response
All main steam and feedwater piping supports inside the drywell have been recently inspected by GPUN.
The inspections indicated the location and direction of g
]
support-(i.e., X-direction, Y-direction, or 2-direction) s j
were consistant with the boundary conditions specified j
in the main steam and feedwater piping analyses.
l Therefore, the main steam and feedwater piping analyses i
are considered valid.
Material Nonconformance Reports
]
(MNCRs) were prepared where the inspections indicated the actual piping support details dif fered f rom those I
shown on the piping support drawings.
The effect of f
these deviations on the main steam and feedwater piping j
support analyses contained in Reference f are being j
evaluated as part of a separate GPUN project.
I I
i I
i j
i l
i i
l i
j -
1 i
i l
L.
i REFERENCES l
4 l
~
a.
NRC (J. A.
Zwolinski) letter to GPUN (P.
B.
Fieldler) dated January 9, 1986, containing a Draft Technical Evaluation Report on SEP seismic analyses of Oyster Creek j
piping, piping supports, and equipment.
b.
MPR (Wm. R. Schmidt) letter to GPUN (J.
R. Thorpe) dated i
September 23, 1982, forwarding " Reanalysis of the Control Rod Drive Return System Piping Considering Axial U-Bolt Restraint and Site Specific Spectra."
c.
MPR-777, " Oyster Creek Nuclear Generating Station -
i Seismic Reanalysis of Core Spray System Piping,"
September 1983.
l j
d.
MPR-794 " Oyster Creek Nuclear Generating Station -
Seismic Analysis of 4160 Volt Switchgear and 460 Volt Unit Substation Cabinets," November 1983.
]
e.
MPR-780, " Oyster Creek Nuclear Generating Station -
]
Seismic Reanalysis of Liquid Poison System Piping,"
December 1983.
f.
MPR-802, Rev.
1, " Oyster Creek Nuclear Generating Station -
Seismic Reanalysis of Main Steam and Feedwater Piping Supports Inside Containment," September 1984.
g.
EGG-EA-5211, " Summary of the Oyster Creek Unit 1 Piping Calculations Performed for the Systematic Evaluation Program," EG&G Idaho, Inc., July 1980.
h.
Structural Mechanics Associates (F. A. Thomas) letter tio Lawrence Livermore Laboratory (T. Nelson) dated August 11, 1981, containing a report on electrical 7
equipment anchorages at Oyster Creek Nuclear Station.
i l
i.
URS/J. A.
Blume Report (Draft), " Development of Anchorage l
Guidelines for Equipment in Nuclear Power Plants,"
November 1985.
j.
NUREG/CR-1981, UCRL-53018, " Seismic Review of the Oyster Creek Nuclear Power Plant as Part of the Systematic E valuation Program."
i 1
f 1
1 10 -
1 i
l k.
Meeting on February 19 and 20, 1986, in Bethesda, l
Maryland, with the Senior Seismic Review and Advisory l
Panel, the NRC, and the Seismic Qualification Utility Group to review EPRI sponsored tasks.
1.
NUREG/CR-0098, " Development of Criteria for Seismic Review of Selected Nuclear Power Plants," May 1978.
l 9
l l
\\
i MPR AssaclATEs. lNc.
REANALYSIS OF THE CONTROL ROD DRIVE RETURN (CRDR) SYSTEM PIPING - COMPARISON OF EGEG FULL PIPING MODEL VERSUS A SHORTENED MODEL 7
COMPARISON OF MODELS Comparison of the EG&G full piping model of the CRD return system piping and the shortened model of this system were made in three areas to verify similarity of response in the two models:
- 1) comparison of support reactions during safe shutdown earthquake (SSE) loading, 2) comparison of natural frequency responses, and 3) comparison of SSE seismic stresses.
j The EG&G full piping model is shown in Figure 1; the i
shortened piping model is shown in Figure 2.
Comparison of Support Reactions Support reactions are compared in Table 1.
As shown in this table, the full model and shortened model compare extremely well.
One difference l'n the support reactions of the two models was at node 170, the node where the full model was stopped to create the shortened model.
At this node, axial support was provided in the shortened model to simulate j^
support provided by the attached pipe in the full piping model.
1
)
Comparison of Natural Frequencies Natural frequencies for the two models are compared in Table 2.
Only the first 10 modes of the shortened piping model are compared to corresponding modes for the full model.
As shown in Table 2, the frequencies of responses of
the shortened model are comparable to the full model.
The direction and locations of responses for each model were the same.
Comparison of Seismic Stresses Seismic stresses for the two models are compared in Table 3.
The seismic stresses shown are based on Equation 9 of the ASME Code,Section III, Subsection NC.
For each model, the i
areas of peak stress were picked out and listed in Table 3.
Comparison of the results indicates that the shortened model accurately predicted the areas of peak stresses.
The magni-l tude of the stresses were on the average 18% higher for the shortened model.
This difference in magnitude is not a concern since final stress analysis results were not based on the magnitude of shortened model stresses, but on the ratio of the change in stress due to changes in loading and boundary conditions.
CONCLUSIONS i
The comparisons presented in this report show that the I
response of the shortened section of the EGGG piping model of the CRD return piping is comparable to the full model.
As a result, seismic stress ratios obtained based on the shortened section of the EG&G piping model are considered adequate.
t I
i
l TABLE 1
{
CRD RETURN PIPING COMPARISON OF SEISMIC SUPPORT RE ACTIONS FOR EG&G FULL PIPING MODEL VERSUS SHORTENED MODEL l
EG&G FULL SHORTENED NODE REACTION MODE L MODE L NO.
TYPE DIRECTION REACTIONS REACTIONS (Note 1)
(Note 2) 2 Force X
143 137 2
Force Y
158 175 2
Force Z
88 77 l
2 Moment X
1016 1020 j
2 Moment Y
3732 4247 I
2 Moment Z
12720 13145
)
35 Force X
292 322 1
35 Force Y
2 3
35 Force Z
292 213 110 Force Y
2 2
i 170 Force X
224 224 i
170 Force Y
130 64 j
170 Force Z
(Note 3) 174 1105 Force Y
91 91 l
1105 Force 2
287 324 1
1135 Force Y
l 1
1231 Force X
91 92 i
1231 Force Y
30 84 1231 Force Z
40 42 l
1231 Moment X
533 554 1
1231 Moment Y
3735 4109 1231 Moment Z
11163 12187 1235 Force X
239 267 1
1235 Force Y
6 6
]
1235 Force 2
168 202 1430 Force Y
9 9
J 1455 Force X
177 187 1455 Force Y
115 119 I
1455 Force Z
68 60 1455 Moment X
5617 5795 1455 Moment Y
9970 10510 1455 Moment Z
989 1055 1535 Force Y
2 1
1575 Force Y
10 10 l
Notes i
1.
Force type reactions are in lbs.
Moment type reactions are in in-lbs.
2.
Refer to Nu-pipe model figures (Figures 1 and 2) for-3 orientation of reactions.
t 3.
The shortened model was given support at node 170 in i
the Z direction to simulate the effect of the continued piping for the full EG&G model.
j
TABLE 2 CRD RETURN PIPING 1
COMPARISON OF NATURAL FREQUENCIES FOR EG&G FULL PIPING MODEL 1
VE RSUS S HORTENE D MODE L i
i, i
EG&G FULL PIPING MODEL SHORTENED MODEL (Note 1) i MO DE NO.
FREQUENCY MO DE NO.
FREQUENCY i
2 2.631 1
2.613 i
3 3.234 2
3.200 j
5 3.610 3
3.625 i
7 3.989 4
4.043 9
4.898 5
4.908 14 5.674 6
5.596 16 5.983 7
5.921 l
18 6.809 8
6.360 21 8.492 9
8.489 j
25 10.505 10 10.526 i
i I,
Notes:
1 1.
The first 10 modes of the shortened piping model j
are compared to corresponding modes for the full i
model.
Mode numbers for the full model that are not listed above are for response in areas of the l
full model that were not part of the shortened i
model.
l
)
i I
J
}
l l-
O 4
i TABLE 1 CRD RETURN PIPING l
COMPARISON OF SEISMIC STRESSES FOR 9
EG&G FULL PIPING MODEL VERSUS SHORTENED MODE L 1
l i
SEISMIC STRESS (psi) j E O. '9 - AS ME SECTION III, SUBSECTION NC AREAS OF PE AK EG&G FULL STRESS PIPING MODEL SHORTENED MODEL (Node No.)
(Note 1) j
}
120-125 24,270 29,595 I
1310-1315 46,174 62,104 1375-1380 52,312 58,457 1450-1455 54,236 64,864 I
1470-1475 43,933 53,793 l
i
)
Notest i
1.
The shortened model stresses were on the average 18% higher than the full model stresses.
The i
shortened model did correctly indicate areas of peak stresses.
i l
m 6
0 kJ s$
$~5 l i
.7 s
g 8
g a
g E
i i
v
=
/, e a
n 1i Ih3sklg,11 il lh, ll g
n=h l3 l
i gj g
3 Q
,r 9d$
nimumn_
p., d...., \\,,,,, y
=
=
AY jYa
'n i
- y...}nm i
"I oxuurui s
~
h e
l
.E!
8
\\
Attl3R_tti
_i
~ ~ " '
l,
~ ~ ~ ~
I(~ ~ ~
,,~,~
l
= = = = -
l
5.:
d ti
~
D 5?
e!!
E se g
,ng E
=
g 3
,El B!l!!l28
==
e I
lii i
id in 35E gg j
gj g,
c.
hw y
L$
b 4
4 I
I I
9
.l s
o 3
5
=!
l a
i s
=
=
=
5
_! l lk sr
...=;-
~
I I
~
I F
I
@j
.0
_E k!
s l
(N i
> N.
k I _H J i
.sg g;g
~
M PR ASSOCIATES. INC.
SUMMARY
OF ANCHORAGE ADEQUACY EVALUATION OF 4160V AND 460V CABINETS USING GUIDELINES BEING DEVELOPED BY EPRI AND THE SEISMIC QUALIFICATION UTILITY GROUP i
1 SHEAR / PULLOUT INTERACTION (A VALUE OF 1.0 INDICATES A FACTOR OF SAFETY OF 4 ON BOLT FAILURE)
EQUIPMENT SEP EVALUATION EPRI EPRI RESULTS (MPR-794)
GUIDELINES GUIDELINES 7% DAMPING 7% DAMPING 5% DAMPING
.l (Notes 1&3)
(Notes 2&4)
(Notes 2&4) f 4160 V - Unit lA 0.91 0.48 0.68 4160 V - Unit 1B 0.84 0.16 0.62 4160 V - Unit 1C 0.79 0.11 0.20 4160 V - Unit 1D 0.74 0.10 0.18 460 V - Unit lA2 1.16 0.03 0.05 460 V - Unit 1B2 1.12 0.03 0.05
- i Notes
1.
Acceptance criteria for SEP evaluations:
P/PA + Y/YA d.1.0; P and j
V are calculated pullout and shear loads; PA and Vg are 1/4 of mean strength values reported in MPR-794, Reference d.
2.
Acceptance criteria are as follows:
For V 1 0.4 V :
g (P/P ) 2 + (V/V ) 2 1.0 g
g For V > 0.4 V :
g 0.7 (P/P ) + (V/V ) 5, 1. 0 g
g P and V are calculated pullout and shear loads; Pg and Vg are 1/4 of the mean strength values recorded in the EPRI guidelines, i
Reference 1.
3.
Horizontal spectral accelerations are taken at the peak of the i
response spectra curve.
Vertical accelerations are taken'at the ZPA.
4.
Horizontal spectral accelerations are taken at 6 Hz based on EPRI guidelines.
Vertical accelerations are taken at the ZPA.
l t
l I
M P R ASSOCIATES, INc.
DESCRIPTION OF INTERFACE BETWEEN I
LIQUID POISON SYSTEM PIPING l
MODELS 1 AND 2 The liquid poison piping in the reactor building consists of two piping runs; 1) from the liquid poison tank to the pump i
suction, and 2) from the pump discharge to the containment penetration.
Rigid anchors were assumed at the liquid poison tank, the pump suction / discharge, and the containment pene-tration.
The portion of piping from the pump discharge to the containment penetration was broken up into two models, which were interconnected at Node 82, as shown in Figures IV-1 and IV-2 from MPR-780 (attached), in order to conform to computer code size limitations.
The model of the piping from Node 82 to the containment penetration overlapped the model of the piping from the pump discharge to Node 82.
The overlap is shown as dashed lines in Figure IV-1.
This overlap extends to the first support at Nodes 78 and 278.
Springs in the Z direction were modeled at Nodes 78 and 278 in order to model the stiffness of the remaining piping.
The combination of springs and overlap provided in this model are considered adequate for stress analysis purposes.
The model of the piping from the pump discharge to Node 82 did not overlap the model of the piping from Node 82 to the containment penetration.
This is considered conservative because the effect of the attached piping from Node 82 to the containment penetration would be to provide additional sup-port for the piping from the pump discharge to Node 82.
The additional support provided by the attached piping would reduce the stresses in piping from the pump discharge to Node 82.
Therefore, this model is also considered conservative for stress analysis purposes.
l
l i
10 U-DOLT Sut#0mf 10P-2-844 l
q 03
.._..__3
's, 7
U-DOLT SerPOh? 467-3-818 s
33 3-aOLT SeptORT Nr-3-a3 I
u-not? SuetOet me-t-es
/
3 u-SOLT Surront ter-3-23
@h giDAIN LINE DRANCM 3
3 4
l u-mott Seefeat 3
Seele84 asAa0Gsa ht-3-83A WP-3-N3 i
SPSIIAG NAacGES s
80P-3-ela CotiTAltrtf.047 PENCTRATIDel 4
V 4
3 4
5 i
i
\\,
PUMP DISCHARGE LINE-CONTAINMENT TO EXPLOSIVE VALVES FIGURE IV-1
~ ~ "
7,g*
- i. g e r cueroar
)
.-.oor cueroo i
"#",a','""'"
S,e e~,
z
, o..
, =.-
.=:-
o g[p
,/
(7 p (. z y / /
=*,e st s.% y o
z
-. c.@ m m
,/
=o &
~
/
A 3'
q
~~
e
<A e
'll N~,%
f =,
A A
~
l O'
b N
y x 4 N =,- -
l e
A s
b J;
f
,0.u,u
- ==- A j#
t k".
i PUMP DISCHARGE LINE-EXPLOSIVE VALVES T'O PUMPS -
FIGURE IV-2
Attachmsnt 4
~
i DEVELOPMENT OF ANCHORAGE GUIDELINES FOR EQUIPMENT IN NUCLEAR POWER PLANTS i
prepared for d
1 Electric Power Research Institute Palo Alto, California I
November 1985 J
prepared by URS/Jonn A. Blume & Associates, Engineers l
150 Fourth Street San Francisco, California 94103 i
e l
~~
4 EPRI EQUIPMENT ANCHORAGE PROJECT FINAL REPORT OUTLINE Included y
In This Draft In Draft
/'
Progress Executive Summary e
1.
Introduction 2.
Strength of Fasteners e
1 3.
Anchorage Analysis Methodology e
/
4.
Benchmarking N,,./
e 5.
General Description of Anchorage Guideline
'e r
6 Development and Use of Anchorage Guideline e
(Partial)
Appendix A Bibliography of Expansion Bolt Test e
Reports
.N
\\
3 Deaggregation of Bolt Strengt'h e
Statistics
'y C Histograms of Bolt Strength Data e
D Computer Code Documentation (Partial) e A
E Justification for Static Coefficent of a
Unity $4[4 x
Y 3ase Plexibility Analysis for MCC e
G Inspection Checklists (Partial) e l
ORS /Blume
i l
Section 3 ANCHORAGE ANALYSIS METHODOLOGY This section presents the analysis methodology developed to calculate the seismic capacity of equipment anchorages.
The methodology serves two purposes.
It has been used to generate the anchorage screening tables, and ip can be used for a case-specific analysis of any anchorage conditions not addressed by "the screening tables (i.e., outliers).
SUMMARY
OF METHODOLOGY
/
l The basic analysis method consists of the followin'g six steps, which are illus-trated in Figure 3.1 and explained in greater detail in the following sections:
)
1.
Determine the appropriate input seismic accelerations for the item j
of equipment for each of the three directions of motion.
1 2.
Calculate the seismic inertia loads for each of the three direc-3 tions of motion using the equivalent static analysis technique.
j N ' 'M 3.
Distribute the seismic inertia ' loads to each of the fasteners by calculating the following force components from each direction of i
motion:
/
--Fastener shear due to horizontal forces i
--rastener shear due. to torsion
/ k i'
--Fastener pullotat
'due to overturning moment (with an appropriately assumed location of the overturning axis)
-5
--Fastener pullout due to vertical forces 1
4 Combine each of the above colinear fastener forces due to each of the three directions of seismic motion using the SRSS method.
5.
Combine the forces on each fastener due to the equipment deadweight with those calculated from the combined seismic loads.
6.
Compare the calculated load on the most highly loaded fastener 1
with the appropriate acceptance criterion.
i Since the above analysis steps are quite simple and systematic, the analysis pro-cedure was automated, and a computer program was developed for the seismic analy-i 3-1 t%?/Blume
t j
4
\\
i y
Start
-[ \\N, n
i Determine input i
seismic accelerations o
Calculate seismic (
/
inertia loads u
1 Distribute inertia loads to fasteners I
Combine fastener loads due to three directions of earthquake input u \\
Combine seismic loads on fasteners with dead loads
- [g'[ loads with fastener 3
o Ware fastener
. b strength criteria t
End Figure 3.1.
Flowchart of Analysis Methods
.I s
i
j '
' sis. of anchcrage layouts using th2 procedures and techniques d2scrib3d in this chap ter.
The program is called JABSAE, and it is written in FORTRAN for an IBM.
PC.
Computer code documentation is provided in Appendix.D.
i The following paragraphs provide the technical basis for the analysis methodology.
INPUT SEISMIC ACCELERATIONS The first step in the analysis procedure is to specify the input seismic accelera-tions (as a multiple of g) for which the equipment anchorage is. to be evaluated.
Three input acceleration values must be specified for each analysis:
two horizontal and one vertical.
The accelera tion values are obtained from an j
appropriate response spectrum at the frequency and damping of the equipment.
(Equipment frecuencies may be determined by testing, by analysis, or using the guidelines given in Section 6.
Guidelines for damping are also given in Sec-tion 6.)
In general, each direction of mo tion.'.could have a di f ferent input w
seismic acceleration due to different input response spectra and equipment frequencies.
'?
One special application of the analysis procedure is the calculation of equipment anchorage capacities for the development of the screening tables used in the anchorage guidelines.
For this application h'ich is described in Section 6), the
- h/.
input acceleration values are lg in each 'of the two horizontal directions and 2/3g in the vertical direction.
These input accelerations are used for this applica-tion to facilitate the calculation of seismic capacity.
A CALCULATION OF SEISMIC INERTIAL 0 ADS Q f%
The seismic inertia load ?f,or each earthquake direction is calculated using the equivalent static load method".
In this method, the seismic analysis is performed statically by applying the inertia loads at the center of gravity of the equip-ment.
The inertia load in each direction is equal to the acceleration value described in. the. previous paragraph, times an equivalent static factor, times the weight of the equipment item.
The result of this calculation is a static load in each of the three orthogonal directions.
For a genera ~1 condition where eccentric-ities exist between the center of gravity and the center of the fasteners, the three static inertia loads will result in two base shears, a pullout force, two overturning moments, and two torsional moments on the equipment base.
Each of the two base shears and the pullout force are equal to the inertial load in each of the corresponding directions.
The overturning and torsional moments are obtained ll.RS/Blume 3-2
'by multiplying the initial forces by tha appropriata distance batween tha cent 2r of th2 fasteners and th2 center of gravity.
The decision to use an equivalent' static analysis was based on the fact that it is a widely accepted approach that is generally considered to be adequate, but not overly conservative, for equipment anchorage calculations.
A more precise evalua-tion of equipment anchorage forces could be obtained through a detailed dynamic analysis; however, this approach is not usually chosen for equipment anchorage verification.
When using the equivalent static load method, an equivalent static factor is applied to the input acceleration values used in the analysis.
This e
N factor, which is comparable to the modal participation factor iny,the dynamic response spectrum analysis method, is influenced by the mass distribution and dynamic characteristics of the equipment.
An equivalent static load factor of 1.5 is recommended by the NRC Standard Review Plan, Section 3.9.2 (Reference
).
be unnecessarily conse[vative for the evaluation of This factor is considered
- 4 equipment anchorage.
An analytical investigation./wasuperformed to develop a more appropriate equivalent static factor for the calculatibn,of anchorage forces.
The investigation consisted of a dynamic analysis calculation, using the response spectrum analysis technique, of the anchorage forces for two equipment prototypes.
N One prototype has uniformly distributed { massg nd stiffness and is fixed at the a
base, and the other consists of a relativeTy rigid superstructure mounted on flex-
\\
ible supports.
The first case is characterized by the mode shape of a cantile-vered beam, while the second is represented by a straight-line mode shape in its fundamental made of vibration.
The dynamic response of equipment under the scope of this study will be boundedsby these two prototypes.
The analysis was performeOusing a response spectrum with a constant acceleration value for all frequencies.
The results of this study demonstrated that the maximum seismic base shear never exceeds 75%.- of the weight of the item times the spectral acceleration (as a multiple of g).
Also, the maximum overturning moment was always less than or equal to the weight of the item times the spectral acceleration times the distance to the center of gravity, which is the midheight for the equipment prototypes considered.
On the basis of these resul ts, a static load factor of 1.0 is recommended for anchorage evaluations and was used throughout this study for the base shear and overturning moment calculations.
URS/Blume 3_3
The fixed-basa equipment prototype was also analyzed to determina the influsnce of higher modas.
Tha baso shears and overturning moments w3re calculated using the same spectral acceleration value for. the first two modes of vibration.
Modal responses were comoined by the squa re -roo t -o f -the -s um-o f -the -s q ua re s (SRSS) me thod.
The resulting. two-mode base shear was found to be approximately 5%
greater than that of the fundamental-mode base shear.
However, the two-mode base shear value was considerably lower than the value recommended above, which indi-cates that the recommended formulation includes an adequate margin for higher-mode i.
responses.
/\\
The overturning moment that resulted from the SRSS of the modal responses was less than 0.4% larger than that obtained from the fundamental-mode response.
This per-centage of increase is considered insignificant.
Thus, it was concluded that the static load factor of 1.0 can be used to calculate the base shear and overturning moments because it gives results that are conservative relative to the more pre-cise dynamic analysis method, and it includes an adequ' ate margin for higher-mode response.
s
~%
A further discussion of the analytical investigation described above can be found in Appendix E.
DISTRIBUTION OF INERTIA FORCES TO THE FASTENERS
\\
Once the base shears, pullout forces, torsional moments, and overturning moments are determined for each direction of earthquake motion, the anchorage forces on each fastener can be calculated for each direction of seismic excitation.
The fastenerfslear and pullout loads was developed for the general formulation for case where the center ofgrith of the equipment item is eccentrically located with respect to the centroiAof the fastener group.
+
The fastener shear due to earthquake input in one direction is composed of two components.
0ne component is the direct shear, which is equal to the total base shear divided' by' the total number of fasteners.
The other component is the shear due to the.eccentrici ty be tween the center of gravity and the centroid of tne fastener group.
This eccentricity causes a torsional moment on the fastener group that is resisted by torsional shear components in each fastener of the group.
The relative magnitude of the torsional shear to which each fastener is subjected is directly proportional to its distance from the centroid of the fastener group and inversely proportional to a constant that is analogous to the moment of inertia of 1
3-4 ES/Blume
/.
/ ~,.
i s
APPENDIX E Justification for Static Coefficient of 1.0
[
9 r
o e
G MS/Blume
w e
e e
O e
9 CONCLUSION USE SEISMIC COEFFICIENT OF l.0 X WEIGHT FOR EqulVALENT STATIC LOAD APPLY EQUlVALENT STATIC LOAD AT EQUIPMENT C.G. (TYPICALLY MID-HEIGHT) e
%S p t
e
/
.n s-s
~~_NN
'N f-
\\,,'
E-1 M lu m e
- 1
D e
- )
es
/
l
,hp/,/
~
N
/
g/.:
A v
CASE 1
'N RIGIO EQUIPMENT I
l
]
st ?,: r..* L o AD : ~ W A [.t.
i>lf <lt
.f i'
~r F virked tr 4v i
i
/ hw M
Zieim dosd~%
.,x shw = vs.AGg-40 wcr
%n : nt = we.o pe airi/ /uigM v).-
luqu
'Q
'w -
conf t =
ntAg i
pak4n.f 9an/r a.-
0 Gf?chefGff.'m94o Ca e
Cvc:2Twznin c Mu:M"= C p.4 = -M.]).cs = 7,oy,j,.ga 0
l E-2 M S lu m e 1
t
s
~
l CASE 2 FLEXI8LE EQUIPMENT WITH CANTILEVER TYPE MODE SHAPE m = mass per uni t height n (sinh (anx)-sin (anx)) + cosh (a x) - cos(a X)
$n(*
- A n
n 9 > He gri
\\A)= 704/II)tt$U: 14Acf w
MODE 1 MODE 2,
/
\\
f
,i s
s 4
[*1(x)dx=.7831
.f*2(x)dx=0.4341
$*2(x)2dx = 1
[+g(x)2dx = 1 PF2 = 0.434 PFt = 0.783 D%
9-g 4 (x)dx = 0.1884 mlSa.
= m(PF )Salf*1(x)dx = 0.6131 mlS k V = m(PF )Sa2 2
2 V,
t a
82 2'
2 = m(PF )Sa2h*2(X)*xdx = 0.0394 m1 OTMg = m(PFg)SaII+1(x) xdx=0.445 m1 5a9c 2
OTM 2
^
(tyt$nis$ l-l Q T X '.a
~~
Conservatively, Assiane Sal ? Sa2 = Sa
%i
~
Combine first and second, modes by SR55
- ST V3 = (V8 82
- Y a
OTM = (OTM2 + 0TMj) b = 0.4467 m1 5 41.O Nb 2
a 2.
o s
and GTM.
No te,insigni fican t impact of higher modes on V3 i
.-9~
v E-3 l}ff$/Blume
-.