ML20049A858

Seismic Margin Review Criteria.

ML20049A858
Person / Time
Site:	Midland
Issue date:	07/31/1981
From:	Kennedy R, Stevenson J STRUCTURAL MECHANICS ASSOCIATES
To:
Shared Package
ML20049A857	List: ML20049A855 ML20049A858
References
SMA-13701.01, NUDOCS 8110020346
Download: ML20049A858 (22)

Text

.

e

,f -

SMA 13701.01 SEISMIC MARGIN REVIEW CRITERIA Prepared for:

Consumers Power Company 1945 W. Parnall Road Jackson, Michigan 49201 July, 1981

, Prepared by:

R. P. Kennedy J. D. Stevenson Structural Mechanics Associates 5160 Birch Street 364E Warrensville Center Road Newport Beach, California 92660 C'.eveland, Ohio 14122 (714) 833-7552 (216) 991-8841 8110020346 810925 PDR ADOCK 05000329 PDR

.. . A

.,3-g-~.--ye-,y-- -,- . - . , ,,_w- ,,-.% , , - .~g- , ,r,-- .-<m.,_. _%y-----6-ymyo,y- ,,, - - . , , - - . _ _ , _ , .,,,-.,,.-.,,.y.-y,,yce , , , ,y.p ,- s

1. INTRODUCTION -

The purpose of this criteria is to define the basis by which the Midland Nuclear Power Plant Seismic Category I structures and components, necessary for safe shutdown of the reactor, shall be evaluated to deter-mine the seismic safety margins associated with the site specific earth-quake defined for the Midland site. Furthermore, this criteria shall define the basis by which such structures and compnents shall be selected for review so as to determine seismic safety margins.

2. BACKGROUND The Midland Nuclear Pou r Plant Seismic Category I structures, mechanical and electrical components and distribution systems were designed and constructed to be capable of a safe shutdown of the nuclear reactor in the event of a 0.12g zero period ground acceleration. Also considered in design were the amplified seismic response associated with the essentially mean centered Housner ground response spectra modified as ,

necessary to reflect additional amplification caused by the building structures. The load combinations, damping values and behavior accept-ance writeria used in the design are also defined in the Midland FSAR.

Currently there is being developed a Seismic Margins Earthquake which will be used to evaluate margins to failure of the as designed and constructed Seismic Category I structures, mechanical and electrical components and distribution syste.r necessary for safe shutdown of the reactor. The exact parameters of the Seismic Margins Earthquake have not as yet been established. It is anticipated that the tero period ground I

1 I

acceleration will be increased significant1 as well as a significant increase in response motion in the 2-10 Hz range as a function of response spectral shape based on the use of a mean plus one-sigma (standard devia-tion) spectra rather than the mean spectra used in the design of the plant.

it is well understood that the Safe Shutdown Earthquake, SSE, load used in the design as compared to other applicable load combinations may not be limiting design in all :ases. In such caset some additional margin may exist. However, it should be understood that when the SSE load is the limiting load case the load factor of 1,0 when used with the behavior criteria defined in the FSAR provides no design margin for an increased seismic input. It must be understood that the behavior limits l of the FSAR or the current NRC Standard Review Plans were not intended for use in evaluating a realistic seismic dasign margin to failure but rather to establish the basis for a safe and adequate seismic design.

This distinction between the limiting behavior criteria established for seismic design, which must consider all the variability in construction as well as the uncertainty in design and behavior criteria developed to define margin to failure, has been well identified in Reference 1. This criteria includes a quantification of the principles discussed in Refer-ence 1 relative to damping and nonlinear ductile response behavior.

Again, it must be emphasized the criteria presented in this document is for evalsacio1 of seismic margin to failure of existing plant structures and components and not a redefinition of the criteria used in design.

2

3. SCOPE I 3.1 SELECTION OF STRUCTURES AND COMPONENTS TO BE EVALUATED The Seit !c Category I structure; and systems necessary for safe shutdown are identified in the Midland FSAR. The following specific structures have been selected for review and evaluation:

a. Containment Structure '

b. Containment Internal Structure

c. Auxiliary Building-Control Tower

d. Service Water Pump Structure

e. Borated Water Storage Tank

f. Diesel Generator Building Selected mechanical and electrical components and distribution systems shall be evaluated to determine safety margins associated with a site specific earthquake defined fee the Midland site. The earthquake thus defined is identified as the Seismic Margins Earthquake, SME. The largest earthquake used in the original design is identified as the Safe Shutdown Earthquake, SSE.

The types of mechanical and electrical components and distribu-tion systems to be eval"ated in detail are summarized as follows:

a. Piping Systems

b. Tanks and Heat Exchangers

c. Vertical Pumps

d. Motor Operated Val"--

e. Electrical Panel Boards

f. Electrical Equipment Racks

g. Electrical Cabinets

h. HVAC "omponents

1. HVAC Ouct J. Cable Trays and Conduit 3

_ - _ _. _ . . - ~

Evaluation of components will include consideration of active as well as passive modes of failure.

Detailed evaluation shall be limited to critical components identified by physical observation, a review of existing analysis, and a screening process as outlined in Figure 1 for structural elements and Figure 2 for mechanical and electrical components and distribution sys-tems . First, a minimum set of potentially critical structural elements and components will be selected by physical observation, and a review of existing analyses based upon whether the design seismic stresses are a high percentage of the design capacity or whether the element is judged to be highly vulnerable to seismic effects. These elements and compo-nents will automatically be furthar evaluated. .in addition, non-selected elements and components will be further screened through a very con-servative screening procedure defined in Figures 1 and 2 to provide high confidence of at least a 1.25 factor against failure for the SME. A significant sampling of all elements and components which do not pass this conservative screening process will be added to the list of struc-tural elements and components to be evaluated.

The safety margin against code ultimate capacity (i.e., capacity from the applicable code, qualification test, FSAR criteria, or Standard Review Plan criteria, whichever is most appropriate) will be determined I

and reported for all selected elements and components. In any case where this safety margin is less than 1.0, a conservative margir against failure will also be determineds reported, and defended. It is this safety margin against failure which should be used to judge the seismic i

adequacy of such elements and components. However, the code ultimate margin will also be reported for use by those who consider it to be

[ useful.

l l

l l

i 4

w -w ._ -- w- y y--,+ *.---,

"'*vr T-*r-- * '

'. . j i

4. MODELING AND ANALYTICAL TECHNIQUES

~

4.1 GENERAL In general the same analytical models and techniques developed for the original SSE design and analysis will be used in the evaluation of seismic margins. Deviations from :uch models and techniques will be identified and used only when the original analysis is clearly inappro-priate. This determination will be based on current state of the art knowledge concerning seismic analysis.

All component analysis shall use elastic systems analysis to proportion loading to the component, attached systems, and supports.

4.2 S0IL STRi!CTURE INTERACTION Since soil structure interaction modeling is a complex and some-what controversial subject, this criteria document shall define the soil structure interaction analytical modeling in some detail. The substruc-ture modeling or solution technique as identified in References 2 and 3 shall be used.

A simplified approach involving a lumped parameter model shall be used subject to the following conditions:

i

a. Two control free field site dependent ground responsa motions are defined at the top of new fill and at original grade level. For buildings founded below the original grade level, the original grade spectra shall be used as input at the structure foundation level. For buildings founded in the fill, an envelope of the top of fill and top of original grade spectra shall be input at tite building foundation level.

b. Soil stiffness variability shall be based on a best esti-mate of soil properties plus uncertainty bounds.

i 5

c. Radiation and material energy dissipation (i.e., the soil damping values) are additive.

d. Damping values used in the analysis shall be determined as follows:

1. Material shall be taken as 5 percent of critical.

2. Radiation-translational (horizontal'and vertical) to be taken as 75 percent of theoretical value.(l)

3. Radiation-rotation (rocking and torsion) to b6 taken at 100 percent of theoretical value.(1)

4. Composite modal damping value in excess of 10 percent of critical shall be justified on a case by case basis if used.(2)

(1) As calculated by generally accepted methcds (e.g., Vibration of Soilt and Foundations, b F. E. D.ichart, J. R. Hall and R. D. Woods, Prentice-Hall Inc., 1970 .

(2) For Rigid Body Motion no cutoff is desired.

l 6

i

5. LOAD COMBINATIONS AND ACCEPTANCE CRITERIA 5.1 LOAD COMBINATIONS The seismic margin review shall be conducted for the following load combination:

U = 1.00 + 1.0L + kE sm where:

U = Limiting Load on the Structure or Component D = Dead load L = Operrcing live load during normal operation plus any live load occtrring as a direct result of earthquake loading.

Live loads shall include thermal effects in those cases where thermal effects induce loads which are considered as primary (i.e., vessel nozzles and component supports). An a

appropriate pressure load will be included for containment i systens.

Esm = Safety Margin Earthquake load k = ductility reduction factor (3)

The ductility ratio k will be taken as 1.0 for all code ultimate margin calculations. For computing conservative margins against failure, an appropriate value of k will be selected folicwing the procedure of Appendix A.

i (3) The basis of the selection of the ductility reduction factor k is presented in Appendix A to this criteria.

7 nv , , - 4 , n _ m 4m. - --y- c---

5.2 ACCEPTANCE CRITERIA  :

For passive components (structural and leak tight integrity only) total stresses resulting from the loading, U, shall be limited as defined in the Midland FSAR except as follows:

a. Current ASME/ACI Code limits may be used provided material selection and fabrication requirements are comparable to current code requirements.

b. Actual measured or sampled mean minus 1.65 standard'devia-tion material properties may be used rather than specified minimum yield or crushing strength.

For active components (must operate or change state), total stresses resulting from the loading, U, shall be limited to normal code allowable plus 20 percent, but in no case shall exceed 0.8 times yield or the onset of nonlinear behavior for computing margins against failure.

Code capacity margins will be based on normal code allowables.

Components qualified by test, to seismic inputs equal to or exceeding SME excitation, may be used in lieu of analysis to evaluate components.

6. DAMPING The damping values used in the seismic margins review shall be based on the damping values given in Table 1 of Regulatory Guide 1.61 (Reference 4) with modifications to component damping as shown in Table 1 of this criteria. The modification of damping values used in this criteria are based on the recommendations and data contained in Refer-ences 1, 5, 6, 7, and 8 as discussed in Appendix B to this criteria.

8

i

7. RESPONSE SPECTRA AND EARTHQUAKE COMPONENTS 7.1 FREE FIELD GROUND RESPONSE SPECTRA The free field site specific ground response spectra shall be as agreed between the U.S. Nuclear Regulatory Commission and the Consumers Power Co.

7.2 FLOOR RESPONSE SPECTRA *

• Floor Response Spectra used in the seismic margin review shall be elastic sr>ectra generated by the direct spectra to spectra method as developed by M. P. Singh (Referetxe 9).

7.3 DIRECTIONAL COMPONENTS OF EARTHQUAKE The floor response spectra and the analysis of structurer and components shall assume that the earthquake input is defined in three orthogonal directions. Combinations of directional resultants shall be on the basis of square root sum of squares. The two horizontal compo-nents of earthquake input motion at the foundation level shall be con-sidered equal. The vertical component shall be taken as two-thirds of the input horizontal motion.

l 9

8. REFERENCES -

). Newmark, N. M., and Hall, W. J., Development of Criteria for Seismic Review of Selected Nuclear Power Plants, NUREG/CR-0098, May,1978.

2. Seismic Review Team, "SSRT Guidelines for SEP Soil-Structure Inter-action Review," 3 December 1980.

3. Lawrence Livermore Laboratories, Recommended Revision to Nuclear Regulatory Commission Seismic Design Criteria, NUREG/CR-1161 RD, May, 1980.

4. Regulatory Guide 1.61, " Damping Values for Seismic Design of Nuclear Power Plants," U.S. Atomic Energy Commission, October,1973.

5. Lawrence Livermore Laboratory, Seismic Review of Dresden Nuclear Power Station - Unit 2 for Systematic Evaluation Program, NUREG/CR-0891 April, 1980.

6. Bohm, G. J., " Damping for Dynamic Analysis of Reactor Coolant Systems," Presented at the National Topic Meeting, Water Reactor Safety of the American Nuclear Society, Salt Lake City, Utah, 26-28 March, 1973.

7. Morrone, A., " Damping Values of Nuclear Power Plant Components," ,

Report WCAP-7921 Westinghouse Nuclear Energy Systems, November, 1972.

8. Stevenson, J. D., " Structural Oamping Values as a Function of Dynamic Response Stress and Deformation Levels," Nuclear Engineering and Design, Vol. 60, 1980.

9. Singh, M. P., " Seismic Design Input for Secondary Systems," ASCE Mini-Conference on Civil Engineering and Nuclear Power, Session II, Boston, Massachusetts, April, 1979, Volume II.

l l

l 10 l

l l

TABLE 1  ;

OAMPING VALUES - PERCENT CRITICAL TO BE USED IN THE SEISMIC MARGINS REVIEW FOR PASSIVE COMPONENTS (4)

Structure or Component Percent Critical Damping Code Margio Failure Margin Large diameter piping systems 3.0 4.0(1)

Pipe diameter > 12 in.

S:3all diameter piping systems 2.0 3.0(1)

Pipe diameter < 12 in.

Welded Steel Structures 4.0 4.0(3)

Bolted Steel Structures 7.0 7.0(3)

Welded Steel Components 3.0 4.0(2)

Bolted Steel Components 3.0 7.0(2)

Reinforced Concrete Structures 7.0 7.0(3)

Prestressed Concrete Structures 5.0 5.0(3)

(1) These values are based on test performed by Westinghouse Electric Co. (References 6, 7).

(2) These damping values are consistent with damping values defined for welded and bolted structures and by review of existing test data (Reference 8).

(3) R.G.1.6108E damping levels shall be used as structural damping in generation of floor response spectra where total calculated stresses in the structure for the SME do not exceed one-half yield.

(4) Damping values used in evaluation of active components shall be reduced in the same proportion of OBE to SSE damping values as

! defined in Table 1 of R.G. 1.61.

1 l

11 l

l

All Seismic Category I Structural Elments Required for Safe Shutdown ,

o Select Samplir.; af Critical Structural Elments By One of Following:

1. Design SeismL Stress is High Percentage of Design Capacity

2. Judgement that El ment is Critical and Vulnerable to Seismic

• on

( h e O C

• U Do In-Structure Shear & Moment Diagrams Developed For No Further the SME Exceed Those No Evaluation of of the SSE by a Factor Structural of > 1.25 (1) in the Elments Affected Frequency Range of is Required Interest?

p Yes Select Additional and More Extensive

" Sampling from Structural Elements Which Do Not Pass Above Criteria Calculate & Report Seismic Margin Against Code Ultimate Strength e

l Is Code Ultimate Strength Margin Greater f k No Calculate and Report Conservative Margin Against Failure i

than 1.0? /

l

, p Yes i

No Further Work Necessary (I}See Appendix A for the development of the 1.25 coefficient as the reciprocal of the Ductility Reduction Factor, k = 0.8 FIGURE 1: SCREENING PROCESS TO SELECT STRilCTURAL ELEMENTS FOR SEICitIC SAFETY MARGIN EVALUATION 12

All Seismic Category I Com.nonents and Distribution Systems Required for Safe Shutdown . j U

Select Sampling of Critical Components By One of Following:

1. Design Seismic Load is High Percentage of Expected Capacity

2. Judgement that Component is Critical and Vulnerable to Seismic

1. 4, 6./

o  %, teg Do the Applicable No Further Floor Response Evaluation of Spectra Generated for No Components or Dist.

SPE Exceed those of the -> Systems at that SSE by Factor of > l.25g Floor Elevation within the Frequency is Required Range af Interest?

Yes Select Additional Sample of Cxiponents which Tend to Be Sansitive to Seismic Loading Scale up by the Ratio SME to Floor Spectral Valui t in Frequency Range of 1..*ers ,,,

Calculated Input Seism c '.ox son and Stress or Deformat.on -

Resultants from SSE Loading l

Do Inpu't Seismic Motion or Stress or Deformation No Resultants Exceed Code Ultim,-te No Further Work Acceptance Limits? Re; ort Necessary Margin Assist Code Ultimate Limits.

Yes Calculate and Report Conservative Margin Assist Failure See Appendix A for the devch. ment of the 1.25 coefficient as the reciprocal of the Ductility pea < tion Factor, k = 0.8 FIGURE 2: SCREENING PROCESS TO SELECT COP-7NE.1TS AND DISTRIBUTION SYSTEMS FOR SEISMIC SAFETY MARGln tVALUATION

(

l l 13

\

APPENDIX A USE OF NONLINEAR DUCTILE BEHAVIOR TO MODIFY SEISMIC DESIGN LOADS A.I.0 EFFCCT OF NONLINEAR RESPONSE AND DUCTILITY REDUCTION FACTORS It has long been recognized that the inherent seismic resistance of a well designed and constructed system is usually much greater than that expected based on elastic analysis largely because nonlinear behav-ior is mobilized to limit the imposed forces and accompanying deforma-tions. To consider this effect rigorously it would be necessary to perform time-history nonlinear analysis which is still the fringe of the state of the art for real structures and components. Newmark and Hall have suggested the use of a modified elastic response spectra in Refer-ence 1 to this criteria. In this criteria a Ductility Reduction Factor, k, is developed as suggested by Villanueva(A1) and applied directly to elastic derived seismic load.

Ductility Reduction Factor formulas are mathematically derived by assuming either (1) that displacements are equal whether the structure -

behaves in an elastic or elasto-plastic manner or (2) that the energy absorbed in an elastic system equals that of an elasto-plastic system.

Then, multiply each earthquake force in the member by the DRF or k factor with the earthquake force in the member being computed from a pseudo-static or full dynamic linear analysis. Use these " reduced" earthquake force 3 to design the members. The underlying principle used here is that stress in a member will not continue to increase once it reaches its yield point.

(A1) Villaneuva, A. S., " Ductility Reduction Factors for Earthquake Design," Reprint 3209, ASCE Spring Convention, Pittsburgh, .

Pennsylvania,1978.

14

1 A.2.0 DEFINITION OF DUCTILITY

~

Ductility as used herein is defined as the ratio of the maximum displacement to the displacement at yield point.

max

p. (A.1) y The term ductility generally implies strain values within the elasto-plastic range, and in this range ductility, u, values are greater than one.

A.3.0 DERIVATION OF DUCTILITY REDUCTION FACTOR FORMULAS There are two formulas commonly used to calculate the DRF. The cerivation of these two formulas is shown below. Each is derived using the following respective assumptions:

Assumption 1 - Applicable for frequencies below 2 Hz. For the same strur.ture and loading, the maximum displacements of a nonlinear elasto-riastic analysis equal those obtained if a linear elastic (no yield .4mit) analysis were made.( A2)

From Figure A.1 the actual (yield) force equation is:

6 f = Y a (A.2) y 6,,," max (A2) Clough, R. W. and Penzien, J., Dynamics of Structures, McGraw-Hill Co., New York, 1975.

15

Clearly the ductility reduction factor equation is:

k=1 (A.3)

P Assumption 2 - Applicable for frequencies in the 2-8 Hz range.

For the same structure and loading, the energy absorbed is the same for both the elasto-plastic system and the elastic system. -

Equating equivalent energies Ep=E e as shown in Figures A.2 and A.3, we have:

fK5,8+(4 - 4 )K4, = .K (p,3)

Dividing by the stiffness / and the square of the deformation at yield 6y  :

and substituting u = 6,,x/6 y, the above equation can be rewritten as 2

2u-1= 8[y (A.5) or g = V2u - 1 Sj (A.6) 16

i Since f - K6, the ductility reduction factor is: -

I k= (A.7)

/2p - 1 For frequencies above 8 Hz the value of k increases linearly to the value of 1.0 at 33 Hz.

A.4.0 LIMITIONS ON THE USE OF THE DUCTILITY REDUCTION FACTOR, k. AND THE DUCTILITY, u While the use of ductility reduction factor, k, is a p.'actical means for introducing the effect of nonlinear response and ductility into seismic design, it must be used with care. The ductility reduction factor, k, is based on the global or systems ductility. It is quite possible in a given structural system that local ductility demand can 6xceed the assumed system ductility.(A3) In this criteria the limiting system ductility, u = 1.3, has been conservatively selected to insure local ductility demands can be met. Use of this rather conservative systems ductility factor results in a ductility reduction factor, k w 0.8 for components with dominant frequencies at and below 8 Hz and varies linearly to k = 1.0 between 8 and 33 Hz. In addition, since strain softening behavior which is also characteristic of buckling behavior c* -

result in large local ductility demands, members subject to buckling shall use k = 1.0. Also in cases where a brittle type failure mode associated with shear failures is dominant, k will be limited to 1.0.

Finally, in cases where the effect on local ductility demand has been assessed, k values less than 0.8 may be used.

( A3) Mahin, S. A. and Bertero, V. V., "An Evaluation of Inelastic Seis'mic Design Spectra," Reprint 3278, ASCE Convention, April,1978.

17

A Force, f

f. m _ _ _ _ _ _ _ _ _ _ _ _ _

l I

I

/

g _ _ _ _ _ / Elasto-plastic I l 1 1

' I- l I I I IDeformation,6 I I m Sy 5 max FIGURE A.1: SUPERIMPOSED FORCE - DEFORMATION DIAGRAMS 18

JL Force, f K8 y - - - - - - - ,

, Ep

__ \ Deformation,5 S y Smai FIGURE A.2: ELAST0-PLASTIC DIAGRAM i

l 19

a - w 9

d A Force, f KE man - - - ~ ~ ~ ~ ~ ~ ~ ~ - -

. l

~

E.

N

\ .

o.ro,m.eo.,. .

O max FIGURE A.3: ELASTIC (N0 YIELD LIMIT) DIAGRAM 20

APPENDIX B I CAMPING VALUES USED IN THIS CRITERIA The damping values for structures given in Table 1 of this criteria are the same as those given in Table 1 of Reg.' Guide 1.61.

While there is significant data that has been compiled (Reference 8) since the Reg. Guide 1.61 was published which indicate reinforced concrete structural damping well in excess of 7 percent critical at or near yield of the rebar, the data includes potential soil structure interaction effects which are accounted for separately in analysis. For this reason no attempt had been made to modify structural damping values as defined in R.G. 1.61.

However, for mechanical equipment and piping systems the summar-ized test data (Reference 8) ind!cate the mean measured damping data at stress levels less than 10 percent exceed the damping level defined for equipment and piping in R.G. 1.61 for SSE level stress at or near yield.

For this reason this criteria proposed to use the same damping for bolted and welded equipment as is defined in the R.G. 1.61 for bolted and welded

• structures, namely 4 and 7 percent, respectively. For the same reason it is also proposed to increase the percent critical damping value fer large piping from 3.0 to 4.0 percont and smaller piping under 12 inches in diam-eter from 2.0 to 3.0 percent. It should also be noted that tne Westing-house Electric Co. has been using 4.,0 percent damping for some time in reactor coolant piping based on tests they performed (Reference 6).

Finally, it should be noted that Table 1 of this criteria has made specific provisions for reducing damping values used H the seismic margins review in the generation of floor response spectra and in the analysis of active components when lower stress levels do not warrant the use of the higher damping values associated with the SSE level earthquake.

21

- _