ML20206K601

SER Supporting Util 860513 Proposed Replacement of Hydraulic Snubbers W/Energy Absorbers on Main Steam Bypass Line

ML20206K601
Person / Time
Site:	Point Beach
Issue date:	04/10/1987
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20206K593	List: ML20206K588 ML20206K601
References
NUDOCS 8704170006
Download: ML20206K601 (40)

Text

'

l

s l l

SAFETY EVALUATION REPORT l POINT BEACH NUCLEAR PLANT I

ENERGY ABSORBERS AS REPLACEMENTS FOR SNUBBERS ON THE UNIT 1 MAIN STEAM BYPASS LINE 1.0 Introduction Wisconsin Electric Power Conpany has applied for NRC approval to install energy absorbers as replacements for snubbers on the Main Steam Bypass Line at Point Beach Unit 1. Energy Absorbers are flexible supports designed to limit seismic piping response by introducing significant danping into the system.

The design and application of these devices is based on proprietary development work by Bechtel Power Corporation.

The staff and its consultant, Brookhaven National Laboratory, have

. reviewed the information provided by the Licensee in support of this '

application. Licensee's responses to NRC request for additional information

on energy absorbers (attachment 1) were also evaluated. The results of the staff evaluation are included in this report.

2.0 Description of Enerqy Absorbers -

In supporting nuclear piping systems from the effects of earthquakes and other dynamic-type loads, hydraulic or mechanical snubbers have been commonly used at locations where thermal expansion considerations preclude the use of rigid supports. Because of the conplex nature of snubber mechanisms, extensive inservice inspection, functional testing, and maintenance programs have been necessary to insure reliable performance. These programs have .

resulted in increased levels of personnel radi'ation exposure, higher operating' costs and may not always enhance piping reliability due to possible existence of undetected malfunctioning snubbers. The nuclear industry has, therefore, begun to develop ways to lessen its dependence on snubbers. ,"

The proposed energy absorbing. devices along with a pseudo-linear -

analysis method were developed by Bechtel to replace snubbers. They are simple, flexible supports made of ductile steel plates which are specially shaped to attain desired energy absorption characteristics. Their operating principle is that they act as flexible supports with finite stiffness to acconodate piping thermal expansion with reduced piping dynamic response by imparting large amounts of apparent damping to the piping system as these devices undergo controlled yielding.

Experimental testing of basic energy absorber concepts was performed at the Earthquake Engineering Research Center of the University of California, Berkeley. These tests, which were conducted under various U.S. Department of Energy and Electric Power Research Institute programs, demonstrated the

, feasibility of the concept. Based on the initial success of these experiments, Bechtel had undertaken an extensive energy absorber development program which included theoretical and analytical studies, energy absorber design, fatigue testing, and other design and application evaluations. ,-

4170006 87043o p ADOCK 05000266 PDR

t ,

l 3.0 Bechtel's Energy Absorber Development Program The Bechtel development program and its resJlts are described in a proprietary report (Ref.1). Major aspects of the program are summarized below.

I 3.1 Operating Principles i

Energy absorbers of Bechtel design are made of simple, specially shaped steel plates designed to exhibit a defined force-displacement relationship.

They exhibit moderate stiffness to accomodate thermal expansion of piping.

Under seismic and dynamic loads they are designed to yield and have a defined hysteretic behavior under cyclic displacements. The hysteretic behavior causes an increase in damping of the system thus providing the desired control to its dynamic response. Since energy absorbers are designed for cyclic operation in the plastic range, they must possess sufficient fatigue endurance capability to maintain the plate's integrity, as required by Code Case N-420 of the ASME Boiler and Pressure Vessel code. -

3.2 Analytical Method of Modified ME101 Computer Program

- ~

The nonlinear behavior of energy absorberg creates a practical problem in analyzing the seismic response of piping systems. Conventional piping systems are seismically qualified using linear modal response spectrum analysis methods. Nonlinear dynamic analysis is normally performed through the application of an acceleration time history of the earthquake and direct integration of the equations of motion. Several computer programs are

available to perform this type of analysis. However, for complex piping i

I systems, theonprocess application is very basis.

a production time consuming and expensive, thus precluding ~its Bechtel,,therefore, developed an equivalent linear response spectrum analysis method to approximate the nonlinear response of piping systems with energy ~ absorbers.

Bechtel's development effort involved several stages. The first stage -

was the development of a specialized nonlinear computer program for more. .

j efficient analysis of systems with localized nonlinearities (energy absorbers). The program solutions were conpared and correlated against . )

analytica) s,olutions from other nonlinear programs and test results. The next stage was the development of a linear analysis method which approximates the 1

nonlinear response through the use of equivalent linear damping and i stiffness. Since both the equivalent damping and stiffness of the piping  !

system are dependent on the displacements of the energy absorbers, an '-

iterative procedure was required. The method was incorporated into a computer program with the capability of performing either linear modal superposition time history analysis or modal response spectrum analysis. The linear program solutions were correlated against the specialized nonlinear program solutions. The last stage was the development of a design procedure which can

,be applied on a production basis using standardized energy absorber designs.

Bechtel's existing ME101 program for piping analysis was modified to incorporate both the specialized nonlinear methodology and the linear -

methodology.

1

/ ,

3.3 Fatigue Testing and Inservice Inspection Protypical testing was performed by Bechtel to establish load-deflection characteristics of energy absorbers required for analysis. Cyclic fatigue tests were performed to establish fatigue design curves in accordance with ASME Code Case N-420. In addition, energy absorbers constructed and installed in conformance with conditions specified in Regulatory Guide 1.84 are subject to inservice inspection. A visual examination method (VT-3 method) as specified in paragraph IWA-2213 of ASME Code Section XI,,wasadopted. The energy absorber design incorporates two aids aimed at simplifying and enhancing visual ISI examination. The first is windows cut out of the sides i of the energy absorber enclosure box to facilitate viewing of the energy 1 absorbing plates and their connecting bolts for visually detecting fatigue cracks. The second is a scratch plate and markers to provide physical evidence of pipe movements and their anagnitudes. The pipe movement can be directly correlated to design conditions and fatigue endurance.

4.0 Application to Point Beach 7 'The Licensee has proposed the installation of energy &bsorbers at selected locations on the Unit 1 Main Steam Bypass Line, piping isometric P-107. ,The system currently includes seven snubbers. The proposed revision

~

• " eliminates all snubbers and replaces them with.five en'ergy absorbers. The

.- revised piping system was reanalyzed using the modified Bechtel ME101 Computer Program which performed equivalent linear analysis (see section 3.2). A description of the analysis and a summary of results was provided in Reference 2. This was later revised and resubmitted in Reference 3.

The piping system with snubbers had been qualified for seismic loads using the original plant seismic design criteria as stated in "NRC IE Bulletin

. 79-14 Final Report" dated January 28,1983. The system was analyzed for thermal expansion, weight and seismic loads..

The revised piping system replaced the seven snubbers with five energy -

absorbers identified as follows:

~

Energy Absorber Size Replaced Snubber EAB1 4x5/16x5 R-EB-2-3 EAB2 4x5/16x5 R-E B-2-1 EAB3 4x5/16x5 R-EB-2-6 EAB4 4x5/16x5 R -E B 4 EAB5 4x5/16x5 EB-2-H17 1

I O

t

4.0 Application to Point Beach (cont'd)

It was noted that Snubbers EB-2-H7 and R-EB-2-7 were removed but net replaced by energy absorbers.

The configuration modified with energy absorbers was reanalyzed using Bechtel's modified ME101 coguter program for all load cases for which it had been previously qualified. For thermal and weight loads, the energy absorbers were modeled as linear spring elements with a stiffness equal to the elastic spring rate of energy absorbers because the resulting movements are lower than the yield displacements.

The original piping seismic design for Point Beach is based on an analysis for a two-directional earthquake (maximum response of each of the two orthogonal horizontal direction responses combined separately with the vertical direction response), 0.5% damping coefficient and SRSS modal response combinations. This was the basis of the analysis used for IE Bulletin 79-14 with snubbers installed. The response spectra used were an envelope of the response spectra at elevation 48' of South Wing Auxiliary Building, at *

elevation 2.5' of Central Auxiliary Building, and at' elevation 105' of

. Containment Building.

" The ' seismic reanalysis of this system with energy absorbers was performed

~~

using the more stringent se'ismic analysis rules of Regulatory Guide 1.92 for a combination of closely spaced modes, and a concurrent three directional earthquake. The equivalent linear analysis method of the modified ME101 com uter program was used. A 0.5% overall system dag ing was used. However, since energy absorbers cause additional energy dissipation to the system, unique damping ratios were calculated for each mode. These damping values are based on the composite effect of the system daging combined with the .

~

equivalent modal damping due to energy absorber participation. The results indicate that for the SSE load case, modal daging varies between 0.5% and

_ 17%.

The results of the reanalysis done by Bechtel indicated that the code ~

allowables of ANSI-B31.1 for normal, upset, faulted and thermal expansion- -

conditions are met. For the faulted condition, which includes the combined effects of pressure, weight and SSE, the maximum piping stress was 31860 psi in the nonseismic portion of the pipe run cogared to an allowable stress of 36000 psi (2.4Sh) of A-106 Grade B carbon steel.

l

?

9

< o 5.0 Staff Evaluation The staff and its consultant, Brookhaven National Laboratory, reviewed the Licensee's submittals (Ref. 2,3) and the Bechtel Technical Report (Ref.1).

Based on the review, additional information was requested in the areas of analytical methodology, fatigue considerations, daging, design considera-  :

tions, test correlations and Regulatory Guide requirements. On October 23,  !

1986, a meeting was held with the Licensee and Bechtel to discuss these l issues. The Licensee's formal responses were documented in Reference 5 'and i are included as Attachment 1 of this report. The Licensee also provided additional information in later transmittals (Ref. 4,6), including the  ;

technical specification, fatigue calculations and the ISI inspection  :

procedure.

According to Regulatory Guide 1.84, Revision 24, the energy absorber should be constructed and. installed in accordance with the requirements of ASME Code Case N-420. In addition, the licensee is required to provide

~ additional information. The Code Case and Regulatory Guide . Requirements and '

i . the staff's evaluation of the proposed Point Beach application to these l[ , _ equirements r are discussed in the following sections.

E '-

~5.1 ASME { ode Case N-420 Requirements 4

Am wue me N-420 requires energy absorbers to be constructed for service in accordance with Section III, Division 1, Subsection NF, and defines additional requirementson material, design, fabrication, and documentation.

Based on a review of t'ne information provided by the licensee, the staff has determined that the energy absorbers should be constructed and installed in accordance with the requirements of the Code Case. The specific requirements and results of the staff evaluation of the Point Beach energy absorbers to .

these requirements arediscussed below. - '

5.1.1 Material Requirements 1 1. Energy absorber plate material shall not exceed one inch nominal thickness.'

t 2.. Weld repairs of energy absorber material shall not be permitted. - -

l

3. After machining to final dimensions, each energy absorbing material shall be visually examined for conformance to the drawings.

4. In lieu of NF-2311(a)3, impact testing shall be required in accordance with N F-2300. For SA-516 material, igact testing shall be required when the nominal section thickness exceeds 5/8 inch and the lowest tegerature for Classes 1 and 2 is less than 30*F (normalized) or 70'F (as rolled), and for i Class 3 is less than 0*F (normalized) or 50*F (as rolled).

5. All other requirements of NF-2000 shall apply.

Evaluation a

The staff reviewed the technical specification for the energy absorbers I and determined that the material requirements are met. The energy aborber l

plates are fabricated from SA516 carbon steel plate material of 5416 inch thickness. Weld repairs are not permitted. Finished plates are visually examined for conformance to the drawings. Impact testing is not required-

since the thickness does not exceed 5/8 inch. All other requirements'of

• NF-2000 apply.

P W

, , , - , , _ - _ _ , - r- m- y y '-'""-^7

- ~-"-"" " 7#" " * '

(

4 o 5.1.2 Design Requirements

-For the design of energy absorbing support parts that perform by sustaining strains beyond elastic limits, the following rules shall apply in lieu of those in NF-3140 and NF-3300.

Fatigue Design Energy absorbing support parts may be designed by load rating. The ' load rati.ng of these parts shall provide for a relationship between imposed strains and #11owable number of cycles.

Cumulative fatigue effects associated with all strain cycles expected for the life of the energy absorbing support parts resulting from all service loadings should be evaluated. The sum of the ratios of the expected number of thermal cycles and dynamic load cycles to their respective allowable number of cycles from the design fatigue curve shall be limited to one.

Design fatigue curves based on testing of prototypical samples of energy

- absorbers shall be developed and used in the fatigue design. Samples tested shall. represent the material used for construction, shapes used, and methods '

l r~ of fabrication. The fatigue design curves shall show a relationship between

• ' the amplitude of strain (c) and the allowable number of cycles (N). The ordi-nate shall show the value of strain, c, and the abscissa shall show the allowable number of cycles,. N. The allowable number of cycles, N, shall be calculated as the smaller of Nmean/1.5 or Nmean-oX2, where Nmean is the number of cycles to failure of specimens tested at the strain value, c, and a is the maximum deviation.

Dead Weight Consideration Loads induced by the dead weight effects of a supported piping system .

should not exceed one-fifth of the elastic yield load capability of the energy' absorbing supports. .

Steady State Vibration Consideration "

Operational steady state vibration effects on energy absorbing suppor -

parts, if present, shall be maintained within the stress limits for the applicable fatigue design curves of Section III, Division 1, Appendix I.

Piping Design Consideration The inelastic characteristics of energy absorbing support parts-shall be considered in satisfying all design and service loadings for which the supported piping is evaluated.

Evaluation The staff determined that the design of the Point Beach energy absorbers was in conformance with the Code Case requirement. Design fatigue curves were developed based on testing of prototypical samples. The appropriate safety .

f actors on allowable numbers of cycles as defined above were applied. ,

l I

t .

Cumulative fatigue effects associated with all specified OBE cycles, SSE cycles, and thermal cycles were considered and evaluated in accordance with the Code Case equation. The highest cumulative usage f actor was .04 compared to an allowable value of 1.0.

The results of the piping analysis provided by the Licensee indicated that the loads induced by the dead weight did not exceed one-fifth of the elastic yield load capability of the energy absorbing supports.

After installation of energy absorbers, the system will be reviewed for normal operating vibration upon startup. The scratch plates on the energy absorbers will be used to verify that the vibrations are within the endurance limit of each energy absorber plate. The limits are marked on each scratch j plate. In addition, the scratch plate will record the occurence of '

unanticipated dynamic loads, such as waterhammer, even though the staff is i aware that no such occcurences have been recorded in this particular pipe run  !

(attachment 1, item 3). 1 The Licensee has committed to perform a walkdown (attachment 1, item 19) I of the modified Main Steam Bypass Line in order to determine whether there is j any apparent adverse interaction with. adjacent safety-related equipment due to increased piping displacement during the seismic event. Results of the

' walkdown should be documented and such records should be maintained. ,

Dead weight and thermal analysis of the piping system utilized the elastic r' stiffness characteristics of the energy absorbers. The results of these analyses confirmed that the energy absorber displacements remain within the

[ elastic range. The seismic analysis of the piping system considered the

• ~

inelastic' characteristics of the energy absorbers through the use of an equivalent linear analysis method. Discussion on the staff evaluation of the seismic analysis is provided in Section 5.3.

5.1.3 Fabrication Requirements Welding shall not be permitted for fabrication and installation of energy absorbing support material. All other requirements of NF-4000 shall apply. -

Evaluation -

Based on the staff review of the technical specification, it was .

determined that the fabrication requirements are met. Plates are assembTed using high strength bolting material. No welding to the plates is allowed in

• the assembly or the installation process. All other requirements of NF-4000 apply. '

5.1.4 Documentation Requirements The Code Case number shall be shown on the data report.

l

?

I

Evaluation The staff determined that the documentation requirement is not applicable to Point Beach because the Code of Record is ANSI-831.1,1967 Although the i

energy absorbers will meet the material, design and fabrication requirements of the ASME Code Section III subsection NF, the Licensee is not required to obtain a Code stamp or generate a Code data report. '

5.2 Regulatory Guide 1.84 Requirements

) Regulatory Guide 1.84, Revision 24 states that Code Case N-420 is i

acceptable provided that each applicant desiring to use the Code Case provide the following additional information:

1. Indication of systems in which energy-absorbing supports are to be used

2. Fatigue design -

3. Piping system analysis results considering inelastic behavior of supports

4. Plans for inservice examination of energy absorbers .

{ ,

Evaluation 4- -

• The Licensee has indicated that energy absorbers will be installed on the

-- Unit drawing 1 Main (Ref. Steam7). Bypass Line at the locatirons shown on the piping isometric This staff evaluation is limited to this specific application snubber and does not imply generic acceptance of energy absorbers as replacements.

N-420. Fatigue design of the energy absorbers meets the requirements of Code Case acceptable. Based on the staff review, the fatigue design is considered 1 Piping system analysis results indicated that the ANSI-B31.1 pipe stress

code allowables for normal, upset, faulted and thermal expansion are met.

Deadweight and thermal analyses were performed using standard linear piping i analysis linear techniques elastic spring acceptable elements. to the staff. Energy absorbers were modeled as. -

absorber displacements remain in the elastic range.The The seismicresults analysis.was.confirmed that t based on an equivalent linear analysis method.

the seismic analysis is included in Section 5.3.A discussion and evaluation of The inservice inspection program is based on ASME Code Section XI rules as discussed in Section 3.3 The Licensee provided additional information on the program in References 5 and 6.

Energy absorbers will be given a visual VT-3 examination installation. onAfter an annual basis for a period of three years after initial three years, the energy absorbers will be placed on a .*

rotating 10-year period.ISI program such that each one receives a VT-3 examination within a be replaced. Plates determined to be cracked by visual examination would In addhion, scratch plates will be inspected to determine if any unanticipated Icadings have occurred to the energy absorber.

absorbers are de9gned in conformance with ASME subsection NF, the staff foundSince en that the VT-3 examination method with inspection intervals as specified above is practical without and active conpatible with code requirements in Section XI for supports parts.

acceptable. The staff finds this inservice inspection pIrogram However, should plates be found to be cracked, the Licensee .

should report the occurence to the NRC and provide information regarding root cause, staff implications to piping integrity, and proposed corrective actions for review. i considered in such event.More frequent insW ctions and expanded ISI samples may be I

_= . . . - . .. . - - - -_ __

5.3 Seismic Analysis

{

l The seismic analysis of the Point Beach Main Steam Bypass Line was based j on ihe equivalent linear modal response spectrum analysis method developed by i Bechtel and incorporated into the modified ME101 coguter program as discussed in Section 3.2. The method approximates the effects of the energy absorber nonlinearities by the use of equivalent stiffnesses and equivalent modal 4

damping ratios which are dependent on the energy absorber displacements.

Modal damping ratios used in the Foint Beach analysis varied between 0.5% and 17%. Since t,he daging ratios exceed the values used in the original design of the plant, the licensee has committed to amend the FSAR to include the use i of higher damping values in the Main Steam Bypass Line caused by the use of i energy absorbers, t

. Evaluation -

i The Bechtel equivalent linear analysis method is a newly developed l r

- technique not previously reviewed or accepted by the staff for generic use. ,  ;

Therefore, the staff requested its consultant, Brookhaven National Laboratory l to perform a confirmatory seismic analysis of the Point Beach piping system

, , using conventional nonlinear analysis methods. BNL developed a piping system' model using the ANSYS general purpose finite element coguter program. The 4

"model was similar to the mo.dified ME101 model with the exception that nonlinear elastic-plastic elements were used to represent the inelastic hysteretic behavior of the energy absorbers. The pipe run remained a linear model. A direct integration time history analysis of the piping system was performed by applying three independent seismic acceleration time histories which correspond to the design spectrum in the vertical and two horizontal

, directions.

The results of the BNL time history analysis were compared to the modified A

ME101 results. Corresponding displacements were cogared at points of maximum' response and at energy absorber locations. The cogarison showed the BNL results to fall within a range of 40% below to 60% above the ME101 j displacements. Some of the exceedence, however, could be atributed to the ," 3 i conservatism of the acceleration time history input and of the damping within a certain frequency range. . .

ANSI-B31.1 equation 12 code stresses were also coquted from the BNL analysis. These stresses include the combined effects of pressure, weight and SSE loads. Stresses predicted by the time history analysis fell within a range of 40% below to 20% above the ME101 stresses at corresponding locations. However, the time history analysis predicted lower stresses in the areas where ME101 predicted maximum stresses. The highest equation 12 code stress coguted from the time history analysis was 21,100 psi at the seismic portion of the pipe run cogared to an allowable stress of 27000 psi (1.8Sh) of A-106 Grade B carbon steel. Since the calculated maximum stress is far below the yield stress (35000 psi) of the piping material, the use of a linear model for the pipe run in the time history analysis is justified. .

i I

2  ?

f ,

Based on the results of the confirmatory analysis described above, the staff concludes that for this application the energy absorbers will provide adequate control of the piping response during an SSE to keep piping stresses within faulted condition code allowables.

6.0 Conclusion .

The staff concludes that the proposed replacement of snubbers on the Point Beach Unit 1 Main Steam Bypass Line by Bechtel designed energy absorbers is acceptable. Staff acceptance is based on the following:  ;

1. The material, design and fabrication requirements of ASME Code Case N-420 were met. (Section 5.1). .

2. Additional information required by Regulatory Guide 1.84 was provided, including ISI program, fatigue design, analysis for inelastic seismic response, and details of pipe run design and configuration. The -

information was reviewed by the staff and found acceptable. (Section 5.2) 4 - .3. .The seismic analysis of the piping system with energy, absorbers provided by the Lincensee was based on the equivalent. linear modal response

._ spectrum analysis method incorporated in Bechtel's modified ME101 computer

, program. A staff confirmatory nonlinear time history analysis verified t

. that during an SSE, pipe stresses will remain below code allowables with

! ample margins. (Section 5.3)

4. Possible occurences of unanticipated dynamic loads, such as waterhammer will be monitored. (Section 5.1.2 and attachment 1, item 3) ,

5. A walkdown of the modified pipe run will b'e performed and documented to

, _. insure no adverse interactions with adjacent safety-related equipment.

, (Section 5.1.2 and attachment 1, item 19) ..

However, the staff feels that cracks of energy absorbing plates, if '. .

observed during future ISI, should be reported to NRC along with root causes, implications to piping integrity, and proposed corrective actions. (Section 5.2)

Since a wide variation in results was. noted between the modified ME101

, scismic analysis and the confirmatory time history analysis, and lack of

! verification that the former results are always bounding, the staff does not accept tne generic application of the modified ME101 analytical methodology at  !

this time. The staff will continue to review any future proposed applications l

! of Bechtel-designed energy absorbers in the area of seismic analysis on a j case-specific basis until a generic evaluation of Bechtel's methodology is

gonpleted. -

i i

l' - -, .- - - --- .- - -

. , - . . - - - - - - - - - - - - - - - - - - - - - - - - ~ - - - ~~

a ,

l l

7.0 References

1. M.Z. Khlafallah and H.M. Lee, " Technical Basis for the Use of Energy Absorbers as Supports of Nuclear Power Plant Piping Systems," Bechtel Power Corporation, January 1985, PDR Docket No. 50-277 for Peach Bottom

, Plant Unit 2.

2. WEPC0 letter VPNPD-86-207, NRC-86-40, C.W. Fay to H.R. Denton, " Energy

Absorbers as Replacements of Snubbers for Seismic Support of Nuclear Piping Systems, Point Beach Nuclear Plant, Unit 1 and 2" dated May 13, 1986.

3. WEPC0 letter VPNPD-86-396, NRC-86-89, C.W. Fay to H.R. Denton, " Response to Request for Additional Information" dated September 3,1986.

4 WEPC0 letter VPNPD-86-453, NRC-86-107, C.W. Fay to H.R. Denton, " Energy '

Absorbers, Point Beach Nuclear Plant Units 1 and 2" dated November 11, ,

1986.

~

' 5.

• WEPC0 letter VPNPD-86-491, NRC-86-114, C.W. Fay to H.R. Denton, Response L --

lto 22 questions requested by the staff (attachment 1), dated December 4, 1986.-

'~

6. WEPC0 letter VPNPD-87-99, C.W. Fay to G. DeGrassi, " Energy Absorber Information Point Beach Nuclear Plant" dated March 11, 1987.

7 Bechtel drawing P-107, Rev. 5, Job No.10447,' " Point Beach Nuclear Plant, Main Steam Outside Containment to HP Turbine Control Valves and to Condenser, Unit 1," April 18,1983. -

9 e

@ O e

b i

l 0

?

0 6

e ATTACHMENT 1

~

Licensee Responses to NRC F st for Additional Information on Ener absorbers

~

i ., . . o a

e

• OO h

l l

Analytical Methodology

1. Provide step by step description regarding application of the analytical methodology to the sample problem in Reference 2. Provide detailed input information (geometry, material properties, loads, etc.) and all calculations required for each step in the process from snubber removal and energy absorber (EA) installation to final evaluation.  !

Answer: The design process of piping systems supported with_ energy absorbers is essentially similar to that for systems supported with rigid supports and snubbers with few exceptions. In thermal and other static analyses, the stiffness characteristics of ener.y absorbers are accounted for. For seismic analysis, ,

.. the inelastic action of energy absorbers is accounted for in an equivalent linear elastic method of response spectra type analysis. All other normally applicable seismic analysis methodologies, procedures, and code allowables used with linear elastic analysis of piping systems are applicable.

Dissimilarities between energy absorber supported systems and traditional linearly supported systems are primarily limited.

, , to the design of the energy absorbers themselves and the method 4- -

used to calculate their equivalent linear response via the con-

, . cept of modal damping ratios.

s In the context of snubber elimination and replacement analysis as used in the sample problem, the following steps summarize the process utilized:

i) Bechtel standard computer program ME101 was used to 1

_~ code the piping geometry, material properties, loads, I

support configuration, etc., for the system in

_ accordance with the existing system configuration.

ii) All snubbers were assumed deleted. The remaining spring supports and rigid supports were incorporated in the reanalysis of the system with energy absorbers.

System reanalysis for the Ibad cases considered was then performed as described in Enclosure 1 of l Wisconsin Electric letter VPNPD-86-396 dated l September 3, 1986. '

l iii) A number of energy absorbers were added to the system '

at some locations where snubbers existed. Load case analyses and prudent iterations thereof were then performed, resulting in an optimum sizing and number of energy absorbers. Results of the analysis were i

used in the evaluations of the pertinent system and support parameters to satisfy the existing project criteria.

i 1

)

i

. . . . - , - , , . . - . , - . -..-,-- , -.-,.,_n .-., , ,... . . . - - - - - . . . . - - - , , . . -, -

i l

iv) Two linear response spectra dynamic analysis methods may be used for energy absorber supported systems.

The first method involves specifying an equivalent stiffness, based on which an optimum sizing of energy absorbers is determined by the computer program ME101.

This method is described in Section 6.2a of Reference 4.

The second method involves specifying predetermined sizes of energy absorbers, the characteristics of which are used in an iterative ME101 linear analysis. The details of this method are described in Section 6.2b of Reference 4. The second method was used in the example problem.

.. v) The energy absorber loading curves (characteristics) which are incorporated into the ME101 program were used in the load case analyses. . The loading curves data for the energy absorber sizes used in the sample

- problem are shown in the ME101 input and output data.

_ Energy absorber sizes selected for this problem are P . in accordance with Bechtel standard sizes.

vi) e _. -

Thermal analysis was performed using the elastic stiffness characteristics of energy absorbers. The output was confirmed to remain within the elastic

.. range of the energy absorber displacement.

vii) Pipe stress, support loads, nozzle loads, and all other pertinent response parameters were evaluated to the existing project criteria.

Steps to be completed prior to installation of the energy absorbers are as follows:

1) Support detail drawings reflecting the snubber deletions and energy absorber replacement will be completed in accordance with standard practice and project commitment.

11) The energy absorbers will be fabricated in accordance with the Bechtel fabrication specification and Code Case N-420 requirements.

iii) An installation specification will be prepared in accordance with project requirements and used in the installation of energy absorbers.

The following sample problem specific information was provided to the

, NRC and Brookhaven National Laboratory by Wisconsin Electric letter VPNPD-86-453 dated November 11, 1986.

1) ME101 Input Listing I

2) ME101 User's Manual _

l l

3) Energy Absorber Loading and Hysteresis Curves for Size E4BS-5

4) Earthquake Artificial Time Histories (Horizontal and Vertical)

5) Piping Isometric P-107 (Main Steam Outside Containment, Unit 1),

Revision 5.

Additionally, the earthquake artificial time histories were provided to BNL on magnetic tape and diskette, along with the isometric key nodal point information, by Wisconsin Electric letter VPNPD-86-490 dated December 1, 1986.

t 1

~

• /,

d- ..

e e

.h g

"me O

O I

e 8

l I

l 1

2. Provide additional information on the method for extrapolating design response spectra curves to higher damping values including the method used to generate the response factor curves in Figure 6.2 of Reference 4.

Answer: The most accurate way to obtain all required response spectra curves for various damping values would be to generate them from the original time history. In many cases this time history is not obtainable. Therefore, two conservative methods are available in the ME101 program to generate the required curves without time history data. The first one is a simple inter-polation from existing bounding spectra curves. The second one is an extrapolation of existing specta to higher damping values. Since response spectra generally decrease inverse- ,

.. exponentially with the increasing damping value, the linear interpolation method will provide conservative results. For the logarithmic extrapolation scheme, a decaying rate has to be determined. When spectra for two or more damping values

, are available, plant specific decaying rates can be determined rather accurately. When only one spectrum is available, a T

conservative estimation of the rate of decay is necessary.

. In this case the basis for the rate of decay adopted in Figure 6.2 is based on Regulatory Guide 1.60. All of the above methods are appropriate for use.

_, The amplification factors for horizontal design spectra, and their ratios relative to the values for 0.5% damping are listed below:

1 Amplification Factors for Control Points Percent Relative Ratics of Acceleration Displacement Critical I Damping A B C D (33 cps) (9 cps) (2.5 cps) (0.25 cps) A B,C D l

i 0.5 1.0 4.96 5.95 3.20 1.0 1.000 1.000 2.0 1.0 3.54 4.25 2.50 1.0 .714 .781 5.0 1.0 2.61 3.13 2.05 1.0 .526 .641 ,

7.0 1.0 2.27 2.72 1.88 1.0 .458 .588 '

10.0 1. 0 1.90 2.28 1.70 1.0 .383 .531 For the vertical design spectra, although the amplification factors are different, the relative ratios are identical to those listed above. These relative ratios are termed ' response factor' and plotted in Figure 6.2 of Reference 4. In general, these ratios are appropriate for ground spectra but will be conservative for higher elevations. In Reference 4, control point A is set at zero period acceleration, control point C is at the maximum acceleration response, and contrni point D is at

___ _ - - -_ _ . . - -- .- . _ _ . . . . ~. . _ , .

2. (cont'd) the boundary of constant displacement response. Control point B is set at the geometrical mean of frequencies A and C, which matches the B frequency of the horizontal design spectra almost exactly.

The extrapolation procedure of Figure 6-2 was used in the sample problem. Figure 2-1, attached, shows a comparison between the damping ratios extrapolated by ME101 for the sample problem and those generated from an artifical time' history which is generated from the broadened 0.5% response spectra curve. The relative conservatism of the extrapolation methodology is demonstrated. .

.. Figure 2-2 shows similar results on another generic spectra.

Relative to actual spectra generated from time histories, the extrapolation method of Figure 6-2 has been demonstrated to be conservative overall, and particularly conservative at the peak

~ ~

areas. Although this method has the potential for slight under-shooting in the valleys of the response spectra, any such in.

. , stances would be of no numerical significance.

s ._ .. .

ee O

e

• e O

,. . 9 . .

.5% -> VARIOUS DAMPINGS

~

. P.B. SSE '

FA - 33.3, Amax = 3.46g, B = 0.1 2  :

1.9 s 1.8 1.7

$ 1.6 1.5 1.4 1.3

' +

1.2 +

+ + + +.

.ta,*, 2, 2 %.

+ + +,,.. s. .n. . + J.. , %. ,. * ,. ,,.,. ,. ,.,. ,.,., ,. ,.

+ +

• 0.9 + .

A1 +

0.8 '

0.7 s ~

ll 1 L 8 ^*  !! \\ A .. u m w ll L g'\WytKJ

\\ Y o3

/ r f

v igg _

0.1 Wi \]

y W -

n- -

O i i i i i i i i i i i i i i i i i O 2 4 6 8 10 12 14 16 18- 20 FREQUENCY (CPS) .-

FIGLHE 2-1 Ratios of extrapolated accelerations at higher danping (EA) to those generated frun time histories (SA) at nodal frequencies of thin Steam bypass line of Point Beach Plant.

~ '

-TB- N S-B .5% -> 1,2,3,^,'5,7,15,20% .

+' V.

1.9 - i

+% d,,i,'"h,1

.,. - 4 ,,, .,..

-u *,*. # +. ..... .

18 _

t

'T' 2.

+ h4 __ _f -IMIMF+ . 4d .4 ,

-$.+.+.+.+.+t11_

.;; t' 2""""'%. tf.. a,+ $ L * +44++4+++-_ -

y a g A mn,,m,xmmuguiuts+ J L+ . ..

+ + + + + + + + + --

o ti W;!Smia*v 4@!7 4 % +.* =. -

E+++++++++~~

tt if @? + ggmunmgimittiin+,i';g.fg7 .;+ C : ; g .

u a a+

(6 """~+4 . #x*=

~ ~

E TV _. humuu+ d+ ++ + ++ + + + -

_ ,_ ij

1. 9 F '""'"- d ' ... " " "" i ..4muuu3%niddEP st-

+3+++' :

+ ~'-

H H:+

y' '

0.9 -  %

(j i

5g 0.8 *-

O

• * ~tif .I h

F

!,_, n '

o.57028t1s)s a La i O.6 7 g p

T o 0.5 Y t

2 -

c.

\

0.4

'['

\ '

0.3 -

0.2 f \

x r l^ '

~_

O.1 -

O J .i,i "

i i i i i i i i i i i i i i i O 2 4 6 8 10 12 14 16 18 20 FREQUENC( (CPS)

Figure 2-2 Itatios of extrapolated spectral acceleration at higher danping (Im) to those generated from tine histories (SA), at all significant frequencies, for an exanple spectra.

~

W

3. Describe the methods that are_used to analyze other dynamic loads

. such as normal vibration and water hammer.

Answer: Other dynamic load cases, including water hammer, for which a force, displacement, or acceleration time history is-specified, would be analyzed using the non-linear time history analysis capability of ME101. Non-linearity only results from the applicable energy absorber characteristics.

For the sample problem, steam or water hammer loads were not included in the original design of the system. In evaluating

, the inputs for.the reanalysis of the system with energy absorbers, Wisconsin Electric considered the possibility of these transients.

.. Based on operating records and discussions with operating per-sonnel, the system has not been subject to these types of tran- r sients and therefore did not include a steam or water hammer ,

load case in the reanalysis.

~

Steady state vibrations are typically evaluated by field obser-vations and econciled by system physical changes rather than.

by detaile6 analysis. Normal steady state vibrations are j e -- -

verified in the field to be within the endurance limits of the energy absobber plates. The Bechtel design of energy absorbers

-- incorporates features aimed at facilitating field determination

__ of steady state vibrations. Endurance limit displacements of energy absorbers are well above those typically experienced during steady state vibrations.

1 i -

\ a 5

l 4

I e

l

\

4. Section 6.4 of Reference 4, states that EA's may yield under thermal expansion conditions. Can they also yield under dead weight conditions?

Will the nonlinear stiffness characteristics be modeled for the piping static load cases?

Answer: All load cases specified for system design are analyzed, including deadweight and thermal.

Deadweight effects on energy absorbers are limited to 20 percent of the yield displacements as required by ASME Code Case N-420.

Thermal expansion effects are typically maintained within the ,

.. range of the yield displacements of energy absorbers. However, this is not mandatory. If thermal expansion analysis of a given system results in yielding of energy absorbers, appro-priate considerations of the non linear stiffness and fatigue will be made in accordance with Code Case N-420 requirements.

5I In the example problem, the thermal displacements at all energy absorber locations are within the range of their yield 2 ., .. displacements.

s b

.e e

O 3

O f

+-

l

5. Define the stress-strain and hysteresis curves used to define the EA properties. Are these curves best estimate, upper bound or lower bound? Are they based on specific EA material tests?

Answer: At present only two types of materials, SA182-TP304 and SA516-Gr60, are used in the fabrication of the energy absorber plates. Actual load-displacement curves derived from testing of prototypes are used in the analysis in lieu of stress-strain curves of the material. Therefore, these curves are best estimates which assure a more accurate consideration of stiff-ness effects. -

The sensitivity studies described in Section 4 of Reference 4 ,

.. demonstrated the acceptability of reasonable variations in material properties. The fabrication specification will specify a tolerance on pertinent X-shaped plate material properties such that variations in loading curves are kept within reasonable

. limits similar to those assumed in the sensitivity studies.

__ Periodic testing of production samples is intended for verifica-r tion and incorporation in the existing data base. l 1

A ._ .. <

B g

l S

l 1

l 1

6. In computing piping system frequencies, is the effect of damping included? Can this be significant for large values of damping?

Answer: Piping systems modal frequencies are calculated without consideration of modal damping ratios, i.e., in an undamped state. In the equivalent linear analysis methodology of the Bechtel ME101 computer program, a maximum equivalent modal damping ratio limit of 30 percent is built-in. The actual model damping values from calculations are not expected to exceed 20%. The undamped natural frequency will shift down-ward by less than 5 percent if -the maximum damping ratio of 30 percent is accounted for, and by less than 2% for 20 percent damping. A maximum shift in natural frequency of ,

.. these magnitudes is considered negligible.

h.

e ed e e

.e f

e l

l 1

i i

7. Since damping is highly dependent on the accurate calculation of dynamic displacements at EA locations, will additional safety factors be included to address uncertainties in material properties and modeling techniques?

i

Answer: The equivalent linear analysis methodology with energy absorbers as utilized by ME101 is conservative when compared with non-linear analysis and actual testing results. Exten-sive sensitivity studies, as described in Section 4.6 of Reference 4, have demonstrated low system sensitivities to reasonable variations in design parameters, which include variation in energy absorbers material characteristics and I other system variabilities. These sensitivity studies were performed using the non-linear analysis method. Their con-clusions equally apply to systems analyzed with the equivalent linear methodology.

- This conclusion can be verified by examining the results_of the analyses of the sample problem provided in Enclosure 1 9-

• ' of our letter dated September 3,1986. As summarized in

'-- ~ ' Table'1, two revisions of the system configuration were analyzed using the equivalent linear analysis method of ME101.

One analysis was on Revision 2 of the isometric drawing which corresponds to the original calculation of record. The

~ second analysis was based on Revision 5 which included minor s

changes which were judged to be acceptable relative to the  !

original calculation of record. The Revision 5 analysis . '

• also included a change in the sizes of some energy absorbers.

A review of the comparative results'from the'two calculations indicates minor differences in the system responses.

~

Because of the high damping provided by energy absorbers, the system response is generally less sensitive overall to changes in damping values. In a linear system, a change of damping values from 1% to 2%, or vice versa, will cause a significant variation in system response. In an energy absorber supported 1

system, on the other hand, a change of* damping value from 10%~

to 1S%, or vice versa, will cause less significant changes in the system response. Most of the basic methods and procedures used in the analysis with energy absorbers are identical to those employed in traditional linear analysis. .Therefore no additional safety factors need be included.

e

)

i 3

I

l l

' Fatigue Considerations'

8. Considering the wide scatter of test data normally observed in developing fatigue curves, why is a safety factor of only 1.5 considered sufficient for the design fatigue curves? The ASME code Appendix I fatigue curves use a minimum safety factor of 20 on cycles.

Answer: The fatigue design of energy absorbers is in accordance with Code Case N-420. The specified minimum safety factor of 1.5, coupled with the fact that the. fatigue design curve for energy absorbers is derived from testing of prototypical samples of actual units, results in a fatigue' design safety factor con-. .

. sistent with many pressure boundary components under current ASME rules. Note that the Code Case specifies determination of allowable number of cycles for a given strain level to be the smaller of (N mean/1.5) or (Nmean - 20), which is consistent

. with the normally accepted statistical design basis. Refer to Section 8 of Reference 3 for additional discussion on the

? _ fatigue testing performed on energy absorbers.

e ...

Periodic testing of fabricated samples from production runs will be performed to verify acceptable correlation with

-- the fatigue design basis curve.

o h

l l

J

, i i

i O

l f

1

-, 4 y ec, , -- . r . , , - , e r-- -,,---e.% _- -- y .,-.w--+--,.

9. Will normal operating vibration loads be considered in the fatigue evaluation of EA's?

Answer: All load cases will be considered in the fatigue evaluation of energy absorbers. In the case of normal operating vibra-tions, no analysis will be performed but the energy absorber deflections will be inspected and limited to within the endurance limit of the material. Refer to the response to Question 17 for a discussion of proposed vibration inspections.

6 m.-

D i ._ . ,

s e

'e

=

10. Section 1.2 states that the EA's allow piping systems to accommodate higher than design earthquake loads. In light of the low safety factor on fatigue, how can this be justified?^

Answer: If higher than earthquake design loads are known, specified, and quantified, they would be analyzed in the same manner as design loads, including evaluation of the fatigue effects. The statement in the report is intended to illustrate an inherent advantage of energy absorbers to accommodate high unanticipated loadings not considered in the system design, which enhances reliability. Energy absorbers-are capable of accommodating a significant number of cycles if such unanticipated loadings should occur. Any unanticipated loading on the EA would be .

. recorded on the scratch plate showing the maximum displacement experienced.

mm.

S g ., .. ,

d e

~

4 I

E

i i

i I

Damping

11. Equivalent viscous damping factors calculated by equation 5-4 of Reference 4 are based on a steady state sinusoidal motion assumption.

Justify the application of these equivalent damping factors to transient loads.

Answer: The equivalent viscous damping factors calculated by Equation 5-4 are based on peak responses. When peak responses are of interest, this equation is appropriate and is consistent with the general concept employed in linear modal analysis methods. If responses lower than the peaks were of interest, then other methods may be more appropriate. For example, ,

, damping based on root-mean square (RMS) response may be used to better predict the RMS response. In the development of the equivalent linear methodology, evaluations using calcu-lated damping based on three methods were made; peak, 70%

. of peak, and RMS. Evaluations based on peak responses pro-vided the best overall correlation with test and non-linear II -

methods, e., .

a h

e

'e l

• ~

12. Will different modal damping values be developed for OBE and SSE load cases?

Answer: Yes. Different modal damping ratios are calculated for OBE, SSE or any load case combinations specified for system design.

1. 3 e

O 4- -

-m e

em O e

9 i

4 t

l l

. . _ . . _ _ _ _ ,, _ . m _ -. _

. e

13. How is damping due to EA plastic deformation combined with piping material damping?

Answer: Medal damping ratios due to energy dissipations in energy absorbers are calculated and added to the generic system damping to determine equivalent modal damping ratios. . Refer to Section 5.2 of Reference 4 for the calculation of modal damping ratios due to energy absorbers. The generic system damping used is determined from the plant licensing commitments.

eo O

e6 e

e 8

y - . - - ,-- .,. . , * , ,

l l

l l

14. How do changes in the hysteresis curves due to cyclic thermal loads affect EA damping?

~

. Answer: One of the criteria used for selecting suitable energy absorber materials is the ability to exhibit essentially constant hysteresis curves throughout the design life. This is determined from actual tests on prototypical samples of the materials used. If the hysteresis curve degrades signifi-cantly for a given material, then the design hysteresis and the number of cycles used in the fatigue calculations will be limited to the number of cycles where the hysteresis is con-stant. Tests have demonstrated that the selected material for the energy absorbers proposed for the Point Beach Nuclear .

.. Plant application (SA-516) has an almost constant hysteresis curve throughout its design life. Additionally, cumulative fatigue testing with variable strains has demonstrated no impact on the hysteresis or fatigue endurance for this material.

__ In most applications energy absorbers will typically be L -

designed to remain within their range of elastic displacements under thermal loads. This is the case for the sample problem.

a ._. .. ,

ee em o

j

~

- - . - - - - . . . , , - - - , , - . . .. , , ,.-v. . . - _ . - , _ , - , - _ . , -

Desion Considerations

15. Will high temperature piping affect the operating temperature of EA's?

If so, are EA material properties affected?

Answer: 1) Energy absorbers are typically located at the building structure end on the restraint line of action. They are connected to the structure on one end and to the pipe through a strut, a rod, or a transition piece and pipe clamp on the other end. Heat transfer between the high temperature pipe wall and the energy absorber is dissipated by the connecting hardware. Therefore, no long term temperature effects on the plates from the .

.. process pipe temperature itself are expected.

ii) Normal ambient temperature effects on the material characteristics will be accounted for in accordance with the published ASME data on yield strength. Higher

, _. temperatures would tend to enhance the material ductility l .- . and will not result in a reduction sf fatigue endurance.'

iii) The maximum normal operating ambient temperature is

~~

typically 120'F. At this temperature, the change in the energy absorber plate's material properties as compared to room temperature is negligible.

e

• e 4

i

. -.,-.n-.

. - - c-- w - - _ , , - - , - . _ , n. .

t

16. Will EA's be exposed to long term radiation which could affect their material properties?

Answer: No appreciable effects on the material characteristics due to normal LWR ambient radiation levels are expected. This is consistent with the general use of ASME type materials in nuclear plant applications.

P ee 6

O f d *.,

• s b

O 3

4 8

I

17. Since snubber removal will change the dynamic characteristics of the piping systems, will each system be monitored for changes in normal operating vibrations?

Answer: After modification to an existing piping system, which includes replacement of snubbers with energy absorbers, the system will be reviewed for normal operating vibration ~on startup. The scratch plates on the energy absorber will be used to verify that the vibrations are within the endurance limits of the energy absorber plates. These limits are marked on each scratch plate. -

Inservice inspection requirements for these devices would be *

~ consistent with those of other standard component support hardware, i.e. visual examination. A cutout in the sides of the energy absorber box design has been provided to accommodate this visual inspection. Plates determined to

, be cracked by visual examination will be replaced. Scratch plates will be inspected to determine if any unanticipated

{~ _

loadings have occurred to the energy absorber.

i.. .

d M

b 4

g e o

e

n '.. , l l

18. Can multidirectional loads be transmitted through EA's? If so, how are the off direction loads considered in the analysis?

Answer: Energy absorbers are designed for single direction loading.

• The design incorporates pin connections for attaching to the piping and building structures, which permit free swing in the 2

unsupported directions. Thus, no off-direction loading from the piping.is transmitted. See the figures below for concep--

tual energy absorber support arrangements. Deadweight loads

. due to the energy absorber itself are considered in the support ,

design if significant. l 4

h e

--: 6 k ... --

. 0 -l y

j ( . . . - -

b Scheme A O

j . . .

.r 4-g

~ ~ ~

) g s

}~ ,

i l

Scheme B 4

s l

l

_. _ _ . _ - . _ . - . _ _ , _ . . - _ , - - _ _ _ _ . ~ . _ _

i .

o e. .

19. Since a piping system with EA's can be expected to have larger seismic displacements than one with snubbers, will potential interactions with adjacent equipment and structures be checked?

Answer: Yes, all potential interferences due to increased seismic displacements will be checked by a physical walkdown. If potential interferences with adjacent equipment or structures exist, they will be reconciled.

1 O

~

a ._ .

a e

1

+-

99 h

e 3

4

I Regulatory Guide 1.84 Requirements

20. Provide a list of piping systems in the plant in which the EA's will be used.  ;

Answer: The initial application proposed for energy absorbers at Point Beach Nuclear Plant is on the mainsteam bypass line for Unit 1. Further applications will be requested on a case by case basis unless generic acceptance of such devices as snubber replacements is granted by the NRC.

O t

i - -

a m

I w

• I l

i 4

8 i

r--

9 ---- - - - < , - - 5 -t - 'm -- r- --

/*., ,

l Test Correlations

21. Section 4.5(b) of Reference 4 demonstrates the need to calibrate the I frequency and damping parameters of the piping model to achieve '

accurate results. Will the calibration be applied on a production basis? If not, provide justification.

Answer: Calibration of production runs is not required. Calibration of frequency and damping parameters was used during the development stages to arrive at an analytical model that correlated to test results as accurately as possible. Such a step was useful to serve as a baseline for later sensitivity studies and other investigations. It was found that very .

. accurate correlations were possible with the energy absorber systems with relatively minor calibrations due to their smooth performance characteristics. Such an accurate corre-lation is not possible with traditional linear systems which contain snubbers or gaps. .

, Section 4.6 of Reference 4 demonstrates that energy absorber

' '~ " ' '

supported systems are not sensitive to reasonable variations in system p.roperties due to the high damping provide by energy absorbers.

3

's a

[

. . - ._ L .- - . - -.- . .

.4s.

22. Figures 5.5 and 5.9 show that linear analysis can underpredict response.

How will this potential underprediction be accounted for in production runs? Do the sample problems represent worse cases or can more signi-ficant underprediction be possible?

Answer: Figures 5-6 (this is the correct reference instead of Figure 5.5) and 5-9 show the relative, instead of actual, distribution of the MODS /PKLV ratios. In this relative comparison of MODS /PKLV ratios the following was used:

i) Rayleigh damping was used in the direct' integration time history analysis program. The damping coefficients were set such that the Rayleigh damping equals the generic ,

.. system damping at both the fundamental frequency and the cut-off frequency. This resulted in the effective damping being lower than the system damping, and consequently higher PKLV response values.

p ii) Raw, unbroadened spectra were used in the MODS analysis.

e- -

. / iii) System modes higher than the cut-off frequency were not considered in the response spectra method and thus resulted

s. in lower MODS values than actual.

, iv) Numerous earthquake excitations were used. Therefore, the i results indicate of the general trend.

i The scatter indicated by the figures is typical of similar MODS to time history comparisons. In actuality, analysis

).

~

with standard MODS methods, which involve broadened spectra, ZPA effects, etc., will result in virtually all points being above the ratio of 1.

, Review of Figures 5-6 and 5-9 illustrates the effect of

, energy absorbers in lowering overall system response sensitivities. This conclusion can be drawn by observing the trend indicated by the two figures.' Figure 5-6 shows that for OBE events where damping is lower, the system response approaches that of a linear system. The MODS /PKLV distribution in this case should approach the distribution of linear systems. Figure 5-9 is for SSE events where damp-ing is relatively higher. The distribution is markedly shifted toward the right, reflecting the conservatism of the equivalent linear analysis methodology and the effect of energy absorbers in lowering response sensitivities.

l

,_, Distribution Copies:

' Docket Files

.NRC PDR Local PDR PAD #1 r/f PAD #1 p/f TNovak, Actg. DD NThompson, DHFT OGC-Bethesda EJordan l BGrimes JPartlow Glear PShuttleworth TColburn FRosa ACRS (10)

LFMB l

l j