ML20090H805

Reliability Assessment of Indian Point Unit 3 Containment Structure Under Combined Loads

ML20090H805
Person / Time
Site:	Indian Point
Issue date:	05/23/1984
From:	Hwang H, Kawakami J, Shinozuka M BROOKHAVEN NATIONAL LABORATORY, COLUMBIA UNIV., NEW YORK, NY
To:	Ashar H NRC
References
BNL-NUREG-34861, NUDOCS 8407270196
Download: ML20090H805 (21)

Text

-

^

. o - .. -

50-2 %

PDL BNL-NUREG-34861 Reliability Assessment of Indian Point Unit 3 Containment Structure Under Combined Loads H. Hwang, M. ASCE*, M. Shinozuka, M. ASCE**, J. Kawakami*, M. Reich

• Abstract In the current design criteria, the load combinations specified for design of concrete containment structures are in the deterministic for-mat. However, by applying the probability-based reliability analysis method developed by BNL to the concrete containnent structures designed according to the criteria, it is possible to evaluate the reliability levels implied in the current design criteria. For this purpose, the re-liability analysis is applied to the Indian Point Unit No. 3 contain-

< ment. The details of the containnent structure such as the geometries and the rebar arrangements,. etc., are taken from the working drawings and the Final Safety Analysis Report. Three kinds of . loads are considered in the reliability analysis. They are, dead load, accidental pressure due to a large LOCA, and earthquake ground acceleration. This paper ,

presents the reliability _ analysis results of the Indian Point Unit 3 containnent subjected to all combinations of loads.

1. Introduction .

Concrete containnent structures in the United States are currently designed according to the ASME code [1] and other supplementary require-ments such as Standard Review Plan (SRP)[12], etc. The load comb'inations specified in these criteria are in the deterministic format and the re-liability levels implied in the load combinations are not stated ex-plicitly. However, it is important to evaluate these reliability levels to insure the safety of the nuclear structures. ,

9

• Brookhaven National Laboratory, Upton, NY 11973.

• • Renwick Professor of Civil Engineering, Columbia University, New York, NY 10027.

B407270196 840523 PDR RES 8407270196 PDR 4-2-1 Hwang

t The Structural Analysis Division of Brookhaven National Laboratory

-(BNL) has been developing a probability-based reliability analysis meth-odology for nuclear structures, particularly for concrete containment st ructure s. [4,7,10,11] An important feature of this methodology is that the finite element analysis and random vibration theory have been incor-porated into the reliability analysis. By utilizing this method, it is possible to evaluate the safety of nuclear structures under various static and dynamic loads in terms of limit state probability.

By applying the reliability analysis method to the concrete con-tainment structures designed according to the criteria mentioned above, it is possible to evaluate the reliability levels implied in the current design criteria. For this purpose, the reliability analysis is applied to the Indian Point Unit No. 3 containnent structure. The results of the reliability analysis are presented in this paper.

2. Containment Description The Indian Point Unit No. 3 nuclear power plant employs a pressur-ized water reactor (PWR) nuclear steam supply system furnished by West-inghouse Electric Corporation. The containment structure consists of a vertical right cylinder with a hemispherical dome on the top. The cy-linder-dome system is built on a basemat of thickness 9'-0" (2.75m)..

Thus, the cylinder-dome system is considered to be fixed at the base in the present analysis for simplicity. The thickness of the dome is equal to 3'-6" (1.07m), whereas the thickness of the cylindrical wall is 4'-6" (1.40m). The inside radius of the dome and the cylinder is equal to 67'-6" (20.59m). The height of the cylindrical wall is 148'-U" (45.14m) and the total height of the containment is 219'-U" (66.80m). The con-tainnent wall is reinforced with hoop, meridional and diagonal rebars.

The details of rabar arrangement for the cylinder and the dome of the containnent are tabulated in Tables 1 and 2, respecitvely.

The containment structure is subjected to various static and dy-namic loads during its li fetime. In this study, only three types of loads are taken into consideration. They are: dead load, accidental .

pressure and earthquake ground acceleration. From reviewing the 4-2-2 Hwang ,

drawings and the Final Safety Analysis Report (FSAR)[8], it is found that there is no live load acting on the containment structure.

The dead loads arise mainly frcm the weights of the dome and the cylinder. The weight density of the reinforced concrete is taken to be 150 lb/ft3 (23.55 kN/m3). The accidental pressure is assumed to be caused by a large Loss of Coolant Accident (LOCA). The accidental pres-sure is considered as a quasi-static load and is uniformly distributed on the contaiment wall. The design value of the pressure is 47 psi (0.32MPa). From FSAR, the design value of the ground acceleration for the Operating Basis Earthquake (0BE) is 0.1 g applied horizontally and 0.05 g applied vertically. Additionally, the ground acceleration for the Design Basis Earthquake (DBE) is 0.15 g horizontally and 0.10 g vertically.

3. Containment Modelling In order to utilize the finite element analysis results in comput-ing the limit state probabilities, the contaiment modelling should be ,

made in such a way that the local. coordinates of the elements have the same directions as those of the rebars. This is very easy to achieve in this study, since all the rebars are in the hoop and meridional direc-

' ' I tions. ' .

The finite element utilized in the analysis is the shell element as described in the SAPV computer code. A three-dimensional finite element model is used for the structural analysis of the contaiment. A de-tailed cross-sectional view of the containment model is shown in Fig.1.

~ As' can be see from this figure, the contalment is divided into 23 la-yers. Except at the top of the dome, each layer has 24 elements such that the nodal points are taken every 15' in the circumferential direc- ,

tion. This discretization requires a total of $53 nodes and 540 ele-ments. The boundaries of the elements are made such that it matches the change of the reinforcements. Hence, the amount of reinforcements in most of the elements is the same as shown in . tables 1 and 2. However, the meridional rebars in the dome portion are varied in dif ferent ele-vations, an average value of the meridional reinforement in these ele-4-2-3 Hwang

ments is used. It is noted that the diagonal rebars, which provide in-plane seismic shear resistance, are not included in the present analy-sis. The liner is also disregarded as a load carrying structural com-ponent in the analysis. Furthermore, such other complications as pene-trations, personal lock and equipment hatches are not included in the study.

4. Material Properties In order to perform a reliability analysis, it is necessary to de-termine the actual material properties. In the present study, the mean values of the material properties are used in the analysis. The varia-tion of material properties will be included in the sensitivity studies in the future. The properties for the concrete and rebars are summarized as follows:

Concrete The minimum compressive strength of concrete at 28 days used for the Indian Pont Unit No. 3 containment is 3000 psi (20.7MPa). However, the mean value of the compressive strength is estimated to be 4896 psi (33.78MPa) from test data. The weight density of the concrete is taken ',

to be 150 lb/ft3 (23.55kN/m3 ). Young's modulus and Poisson's ratio . '

are 3.1 x 106 psi (213'90MPa) and 0.2, respectively.

• Reinforcing Bars As can be seen from Tables 1 and 2, No.18 rebars are the main re-inforcement used in the contaiment structure. Hence, the statistics for No.18 rebars is used to represent all other types of rebars.

Young's modulus and Poisson's ratio are taken to be 29.0 x 106 psj ,

(200100MPa) and 0.3, respectively. From the test data the mean value of the yield strength fy are estimated to be 71.8 ksi (495.42MPa).

b. Probabilistic Models for Loads Various static and dynamic loads act on the contaiment structure 4-2-4 Hwang

I .

during it's li fetime.

These loads may be caused by normal operating, environmental and accidental conditions. Since the loads atrinsically involve random and other uncertainties, an appropriate probabilistic model for each loadinust be established.

Dead Load As mentioned in Section 2, the dead load primarily arises from the weights of the contairrnent wall. It is noted that there are some uncer-tainties as to the actual magnitude of the dead load.[6] For the pur-pcse of this analysis, however, dead load is assumed to be detenninistic and is equal to the design value, which is computed based on the weight density of reinforced concrete as 150 lb/ft3 (23.55kN/m3),

Accidental Pressure The accidental pressure is considered as a quasi-static load and it is uniformly distributed on the containment wall. The accidental pres-sure is idealized as a rectangular pulse and will occur in accordance ,

with the Poisson law during the containment life. Under these assump-tions, three parameters are required to model the accidental pressure:

the occurrence rate Ap (per year), the mean duration udP (in sec-onds) and the intensity P. For the Indian Point Unit No. 3 contairinent,'

the mean duration is taken to be 1200 seconds. This value is obtained*

f ran the approximation of the time history. According to the Indian Point Probabilistic Safety Study [9], the mean occurreace rate for a large LOCA is 2.16 x 10-3/yr. The intensity of the accidental pres-sure is treated as a Gaussian random variable. The consensus suivey of _

the nuclear structural loads, which was carried out by BNL, indicates that the ratio of the mean value to the design value is 0.89 and the coef ficient of variation is 0.12.[5,6] Sirce the design value of the accidental pressure is 47 psi, (0.32MPa), the mean value, P and the standard deviation, op, are 41.83 psi (0.29MPa) and b.02 psi (0.0345 MPa), respectively.

4-2-5 Hwang

p.

Earthquake Ground Acceleration The earthquake ground acceleration is assumed to act only along the global x direction. The ground acceleration is idealized as a segment of a stationary Gaussian process with mean zero and a Kanai-Tajimi spec-trum. The Kanai-Tajimi spectrum has the following expression:

1 + 4c2 (,j, )2 1 - (w/wg)2 2, C (,,j,g)2 99 0 2 where the parameter So represents the intensity of the earthquake.

wg and cg are the dominant ground frequency and the critical damp-ing, respectively, which depend on the soil conditions of a site. The soil condition of the Indian Point Power Plant is determined as rock.[8]

For such a soil condition, Ref. 3 reccmmends that og and cg in Eq. 1 are taken to be 8nrad/sec and 0.6, respectively. The mean duration pdE of the earthquake acceleration is assumed to be 15 seconds. The peak ground acceleration A1 , given an earthquake, is assumed to be A1=pogg where pg is the peak factor which is assumed to be 3.0. The standard deviation of t.he ground acceleration, o ,gcomputed by integrating the Kanai-Tajimi spectral density function with respect to e,is 1*

I '

a

= Mnw + 2c9 ) C0 (2C (2) 9 9 g The peak ground acceleration A 1 , given an earthquake, can be rewrit-ten as A1 = ag /S7 (3) where o =P'fnwg(2 + 2'g) (4) g g If the earthquake occurs in accordance with the Poisson law at a rate

>E per year, it is easy to show that the probability distribution '

FA (a) of the annual peak ground acceleration A is related to the 4-2-6 Hwang

probability distribuiton FAg(a) of Al in the following f ashion.

F(a)=exp{-X[1-FA(a)]f A E i

i

" 1 + Q , u FA (a[

F A1 Therefore, if a o indicates the mir.imum peak ground acceleration for any ground shaking to be considered an earthquake, FA1(ao) = 0 and hence, AE = - in FA (a o ). Assuming that AF (a) is of the ex-trer.e distribution of Type II, i.e.,

FA (a) = exp[-(a/p)-"] (6) where a and y are two pararreters to be determined. By least square fit-ting to a harzard curve given in the PRA study [9], we find a = 3.14 and y = 0.0135.

From Eqs. 5 and 6, we obtain:

FA7 (a) = 1 - (a/ao )-a a4ao (7)

Under these conditions, one finds that AE = 1.64 x 10-2/ year for I,

ao = 0.05 g. Combining Eqs. 3 and 7, and writing Z for (, we fur .

ther obtain the probability distribution and density functions of Z in' the forms, respectively, FZ (z) = 1 - (ag z/ao )-a for z >, a o

/ ag (8) f Z(z) = a(ag/ao )(agz /ao)-("+1)

The information about the maximum earthquake ground acceleration, 4 a max, which represer.ts the largest earthquake possible to occur at a particular site, is needed in order to determine the limit state prob- -

l ability. In this study, amax is chosen to be equal to 0.7? 3 l

t .

4-2-7 Hwang

F ,

6. Finite Element Analysis Static Analysis As mentioned in Section 5, dead load and accidental pressure are considereo to be static loads acting on the containment. Using the fi-nite element model described in Section 3, a static analyi ; of the con-tainment due to dead load alone is performed. Similarily, the analysis of the containment due to unit accidental pressure alone is also carried out. These results are going to be used in the reliability analysis.

Dynamic Analysis For dynamic analysis of structures, modal analysis is erployed.

Hence, the dynamic characteristics of the structures are represented by the natural frequencies and associated mode shapes. Using the model de-scribed in Section 3 and one half of the stiffness of the uncracked sec-tion, the first 20 natural frequencies and corresponding mode shapes are evaluatep. It is important to choose the significantly participating modes for the reliability analysi.s. In this study, only the first and second pairs of bending modes are included in the analysis.

I

7. Limit State For The Containnent I

A limit state essentially represents a state of undesirable struc-tv al behavior. In 9eneral, it will depend on the characteristics of the structures and the loadings that act on tne structures. For a particular structural system, it is possible that more than one limit state mey be considered. Limit states must also be related to the response quantities obtainable from the selected structural analysis method, e.g., the finite element method adopted in this study.

In this paper, the flexural limit state for containments is defined according to the ultimate strength theory of the reinforced concrete.

It is described as follows: At any time during the serivce life of the

~

structure, the state of structural response is considered to have reached the limit state if a maximum compressive strain at the extreme 4-2-8 Hwang

fiber of the cross-section is equal to 0.003, while the yielding of re-bars is permitted. Based on the above definition of the limit state and the theory of reinforced concrete, for each cross-section of a finite element, a limit state surface can be constructed in terms of the mem-brane stress and bending moment, which is taken about the center of the cross-section. [2] A typical limit state surface is shown in Fig. 2.

In this figure, point "a" is determined from a stress state of uniform compression and point "e" from uniform tension. Points "c" and "c'" are the so-called " balanced point", at which a concrete compression strain of 0.003 and a steel tension strain of f /Esy are reached simultane-ously. Furthermore, lines abc and ab'c' in Fig. 2 represent compression failure and lines cde and c'd'e represent tension failure.

8. Reliability Analysis Results The reliability analysis methodology used in this study is sum-marized in Ref. 4. Based on this reliability analysis method, the structural model, loading conditions and the limit state described in the preceding sections, thp reliability analysis for the Indian Point Unit No. 3 containment structure .is carried out. The results are pre-sented in this section.

Dead Load and Accidental Pressure (D+P) .

The load characteristics of the dead load and the accidental pres-sure due to a large LOCA are described in Sections S. The conditional limit state probability of the critical elements is 3.99 x 10-7 The critical elements are the elenents denoted as 289 to 312, which are located in the first layer of the dome section just above the spring

l. line. Since the structure and the loads are both axisymmetrical in this case, all the elements located at the same level have the same limit state probability. The limit state is reached as the hoop rebars in the critical elements are yielded. When the ccaditional limit state proba-l bility is multiplied by the expected number of such simultaneous occur-rences during the containnent service life of forty years, the uncondi-tional limit state probabilities P (D+P) f under the simultaneous action '

of D and P is obtained, and is equal to 3.46 x 10-8 4-2-9 Hwang

(

Dead Load and Earthquake Ground Acceleration (D+E)

The analytical idealization of the earthquake ground acceleration is presented in Section 5. For the combination of the dead load and earthquake ground acceleration, the critical elements are elements 6, 7, 18 and 19. These elements are located in the lowest finite element layer and immediately adjacent to the global x axis. In this case, the limit state is reached as the meridional reinforcing bars in the criti-cal elements are yielded. The locations of the critical elements and the nenner in which the limit state is reached are obviously consistent with the structural and loading symmetry with respect to the x-axis un-der this particular load combination. The lower and upper bounds of the conditional limit state probability P(D+E) ace found to be very close and equal to 1.02 x 10-8. Multiplying by the expected number of sim-ultaneous occurrences of D and E, the unconditional limit state prob-ability during the 40 year lifetime is 6.72 x 10-9 The fragility curves are used in seismic probabilistic risk assess-ment (PRA) studies for nuclear power plants. The fragility is defined as the conditional limit state probability for a given pea'k ground ac-celeration. Using the BNL method for generating fragility curves [4], a fragility curve for the flexural limit state of the containment is pre-sented in Fig. 3 and Table 3 shows the corresponding numerical values. I '

Since all the data used in the analysis are taken to be best estimate , >

values (or mean values), this fragility curve may be interpreted as the mean fragili ty curve. It can be seen from Table 3, the peak ground ac-celeration corresponding to the median of the curve is 1.29.

Dead Load, Earthquake Ground Acceleration and Accidental Pressure (D+E+P)

The probabilistic characteristics of the loads are described in .

Section 5. Under the combination of these loads, the critical elements are found tu be elements 294, 29S, 306 and 307, which are located imme-diately adjacent to the global x-axis (when projected cnto a horizontal plane) at the first layer above the springline. This is the same level at which the critical elements are found under the D+P load combination.

The manner in which the limit state is reached is also the same as for 4-2-10 Hwang

the D+P combination (i.e., yielding of the hoop rebars). The lower and upper bounds of the conditional limit state probability P(D+E+P) under this load combination are also very close and the average is 1.62 x 10-5 Finally, multiplying by the expected number of simultaneous oc-curence of these loads, the unconditional limit state probability Pf (D+E+P) under this load combination during the containment life of forty years is 8.8b x 10-13, Comparing the limit state probabilities under the load combination D+P with those under the combination D+E+P, it is observed that (1) the mode in which the limit state is reached (yielding of the hoop rebars) in the critical elements is the same and (2) the critical elements under the load combination D+E+P comprise the four elements which are most stressed by the additional earthquake load among those critical elenents under the load combination 0+P. These observations suggest that the accidental pressure P is a dominant factor in controlling the condition-al limit state probability. The substantial reduction in the uncondition-al probability P (D+E+P),

t as compared with P f (D+P), is primarily attributable to the fact that the probability of simultaneous occurrence of three loads, D+E+P, 5.46 x 10-8, is much smaller than that of two loads D+P, 8.64 x 10-2, Overall limit State Probability .

The limit state probabilities evaluated in the preceding section are those at the critical eierents within the containment under various load combinations. While the limit state probability of the contairment as a whole, or the system limit state probability, under a certain load combination is always larger than that of the critical elements, the au-thor's experience in structural reliability analysis suggests that the difference between the systea limit state probability and the limit state probability of the critical elements is tolerable for the type of load-structure system under consid'eration. Therefore, for the sake of analytical simplicity and computational economy, the present study ap-proximates the contairment limit state probability under each load com-bination by the critical element limit state probability. Under the as-sumption that the contairment will not fail under dead load alone, the 4-2-11 Hwang

overall contaiment limit state probability pr is then obtained as the sum of the limit state probabilities under all these (mutually exclu-sive) load combinations. Hence, the contaiment limit state probabil-ity is 4.13 x 10-8 for its lifetime of forty years. The reliability analysis results of the contaiment are summarized in Table 4.

9. Concluding Remarks This paper presents the reliability analysis results of the Indian Point Unit No. 3 containment structure under dead load, accidental pres-sure and earthquake ground acceleration. The reliability analysis meth-od for concrete containment structures developed by BNL is employed in the analysis. This is the first attenpt to carry out the reliability analysis for the existing containment structures in order to evaluate the reliability levels implied in the design criteria. It is noted that the estimated reliability levels are affected by the judgenents made in the design process and the assumptions made in the reliability evaluation. In order to reasonably assess the reliability levels im-plied in the design criteria, it is recessary to continue ti? efforts by carrying out the reliability analysis for oth~ re existing containment stnJctures designed according to these criteria.

Acknowledgements .

The autho.1 wish to express their appreciation to Mr. H. Ashar of' the Nuclear Regulatory Commission for his advice and support during vari-ous phases of this study. The authors also wish to thank the Power Au-thority of the State of New York and United Engineers and Contractors for providing the drawings and test data. Thanks are due to Ms. Diana Votruba for the typing of this paper.

NOTICE This work was perfonned under the auspices of the U.S. Nuclear Reg-ulatory Commission, Washington, D.C. The views expressed in this paper are those of the authors. The technical contents of this paper have neither been approved nor disapproved by the United States Nuclear Reg-ulatory Commission and Brookhaven National Laboratory. .

4-2-12 Hwang -

I

s e 1

Table 1. Cylinder Reinforcement.

Meridional Elevation Hoop Primary Seconda ry Di agonal 0 to 25'-0" 2#18 014 in 1#18 @ 12 in 1#18 @ 12 in 1#18 @ 30 in 25'-0" to 45'-5" "

1#11 0 12 in 45 '-5" to 50"-3" 2#11 @ 36 in 50'-3" to 54'-10" " "

1#110 36 in 54'-10" to 110'-6" 110'-6" ,

to 148'-0" 1#18 @ 60 in+

1#14 @ 60 in Hwang 4-2-13

i Table 2. Dome Reinforcement.

Angle From Spring Line (Degrees) Hoop Meridional Diagonal 0 81/4" outside 0*-9.5* 1#14 j 1#18 0 0.796* 1#18+1#14 @ 4

( @ E.0" inside (horiz. dist.)

9.5*-18.5 1#14 0 8.0 in 18.5*-35*

1#11 0 2 (horiz.

4 dist.)

35*-55' 55*-60* 1#14 @ 9.0 in 60*-75* 1#18 @ 1.593*

75*-83* 1#18 0 3.186*

83*-86* 1#18 @ 6.37*

36*-90* 1#18 @ 12.74 4-2-14 Hwang

l Table 3. Fragility Curve.

PGA(g) p(D+E) PGA(g) p(D+E) 0.60 2.15 E-6 1.205 0.50 0.65 3.79 E-5 1.25 0.61 0.70 2.45 E-4 1.30 0.71 0.75 1.11 E-3 1.35 0.80 0.80 3.86 E-3 1.40 0.86 0.85 1.09 E-2 1.45 0.92 0.90 2.58 E-2 1.50 0.95 i 0.95 5.34 E-2 1.55 0.97 1.00 9.84 E-2 1.60 0.98 1.05 0.18 1.65 0.99 1.10 0.27 1.70 1.00 1.15 0.38 1.20 0.49 Table 4. Limit State Probabilities.

f Expected Number Conditional Unconditional Load Com- of Limit State Limit State Critical binations Occurrences Probability Probability Elements

! D+P 8.64x10-2 3.99x10-7 3.46x10-8 289,290, l

' ...,312 D+E 6.56x10-1 1.02x10-8 6.72x10-9 6,7,18, j 19 D+E+P 5.46x10-8 1.62x10-5 8.8Sx10-13 294,295, 306,307 l

4.13x10-8 i

Ove rall --- --- ---

NOTE: Assuming the containment will not fail under dead load alone.

4-2-15 Hwang

l

% e --

3 ;2 oo, @o i e o sn I y &e p

$07 s fl 5\, tA g . 14.5 . S 6Y'S u /

s / 17 a'

/

/ /9 337 ' 3

' i

• i

/s[M'5 4o (

265 - 288 289 - 312

,5.O' l

241 - 264  ;

8.O' -

1

- 217 - 240 10.O' 193 - 216 y 15.O' ,

y -

ELEMENT O l 169 -192 b 30.O' l a >

% i

'oas -

,k ,

,I E 145 -168 25.O' ,

=,-

4 10.0, 121 - 14 4 "g ---

p 97 -120 f$ ,-

20.O' 73 - % hj 10.O' 49 - 72 5 ~

5.O' 25 - 48 N _' 5.O' .

1 - 24 8 __ [5.O' .

a sso m a w.xv  !

Fig. 1. Cross Section of Containment Model. 1 4-2-16 Hwang

i

+ , t m C 8

0 1

x i 2

d b 3 3 O

l O

l x x O 2 5 Tn __ -

e ^ c e

a f.

r u

S e

t a

t S

t i

i m

L b 2 I .

g d i F

s

• 0!

1 x

2 C

- 9 asNIwN T$a l l ll'! , l ' .!:!* I ' *ir l: ! i!i . k ii

i m i l l l I I I l l 1 -

i i

b

_ Q -

- _ T -

- S s

- _ g w

a w

U N_ W a

z a

O 8 .

u t

_ _ om t.a.J

_ _ gl ul a

U D.

o b'

- .1 I I l I i i i I i 'g:

.e.. -

O w W q N O

" d d d d gl .

l .

l l 3'20 01300 A1Ill8V90Bd 31VIS IIWIl 7VN0IlION03 E!

4-2-18 Hwang l _. , . - - - _ , _ - . - . - _ _ - . _ - - -

\

l References

1. ACI-ASME Joint Technical Committee, " Code for Concrete Reactor Vessels and Containments, ASME Boiler and Pressure Vessel Code Section III - Division 2", 1980.

2. Chang, M. , Brown, P. , Kako, T. , Hwang, H. , " Structural Modeling and Limit State Identification for Reliability Analysis of RC Contain-ment Structures", SMIRT-7 Conference Paper M 3/2, August 22-26, 1983.

3. Ellingwood, B., Batts, M.E. , " Characterization of Earthquake Forces for Probability-Based Design of Nuclear Structures",

BNL-NUREG-51587, NUREG/CR-2945, Sept.1982.

4. Hwang, H. , Shinozuka, M., Brown, P. and Reich, M., " Reliability Assessment of Reinforced Concrete Containment Structures",

BNL-NUREG-51661, NUREG/CR-3227, Februa ry,1983.

5. Hwang, H. , Wang, P.C. , Shoaman, M. and Reich, M., "A Consensus Estination Study of Nuclear Power Plant Structural Loads",

BNL-NUREG-51678, NUREG/CR-3315, May 1983.

6. Hwang, H. , Wang, P.C. and Reich, M., "Probabilistic Models for Operational and Accidental Loads on Seismic Category I Structures", BNL-NUREG-51682, NUREG/CR-3342, June 1983.

7. Kako, T. , Shinozuka, M. , Hwang, H. and Reich, M., " FEM-Based Random Vibration Analysis of Nuclear Structures Under Seismic ,

Loading", SMiRT-7 Conference Paper X 7/2, Chicago, IL, August 22-26, 1983.

8. Power Authority of the State of New York, " Indian Point Nuclear Power Plant, Unit 3, Final Safety Analysis Report", Docket 50286.

9. Power Authority of the State of New York, " Indian Point Probabilistic Safety Study", Docket 50247, March 1982. .

Hwang 4-2-19

l

. 10. Shinozuka, M., Kako, T. , Hwang, H. and Reich, M., " Development of a Reliability Analysis Method for Category I Structures", SMiRT-7 Conference Paper M 5/3, Chicago, IL, August 22-26, 1983.

11. Shinozuka , M. , Kako , T. , Hwang , H. , Brown, P. and Reich, M. ,

" Estimation of Structural Reliability Under Combined Loads",

SMiRT-7 Conference Paper M 2/3, Chicago, IL, August 22-26, 1983.

12. USNRC, Standard Review Plan, Section 3.8.1, NUREG-0800, Rev.1, 1981.

i 1

. s 4-2-20 Hwang

F- ~

. A i NHC # OHu GA UI. ff UCLC AR FWGul AIORY COMMISSION 1. RE POR1 NUM B E R (if earf g O

y. m '..*e*..n*"a*

rn s- #'* *' '.s*a CnM 32'1 ~

' 'D. *'*

• ENT.-NUPrG-34861 I r"u. *en*, c* n*.i e

PUBLICATIONS RELEASE FOR UNCLASSIFIED 2. DisT RisUTiON C AT E oORv NO. 1,ns e,. e,,. ....e n.-s e,--

NRC CONTRACTOR AND CONSULTANT REPORTS , '" a:' [~,'e, 7.'s,',*;'"'"*"

e (Plc.use Type or hint) s~u*tGesMV i

3. TITLE ANO $UB T6TLE (5 rete en f fr ee she a en des wmeng Reliability Assessment of Indian Point Unit 3 Containment Structure Under Combined Loads '

i. A u T H O a s i,, , , , hen ,h,ee ,.e.e r.,s, . .h., ,.u. e, ., a nd .,he, y H. Hwang, M. Shinozuka, J. Kawakami, M. Reich

f. N AME OF CONTR ACTOR M AILING ADDRESS (N,eber and se,eet siry stese end e.p sede) T E L[ PHONE NO.

Brookh3ven National Laboratory Upton, NY 11973 666-2577

g. D AT E M ANUSCRIPT 7. N HC f HOG h AM LPONSORIT E CHrdlC A L MON 110R COMPL E T E D l TE LE PHONE NO.

May 23, 1984 H. Ashar 443-5892

p. CON I R ACT O AT A

e. C ON1R AC1 OR F EN NUMrst R lDo not last DOf sonsrett noonberl, -- ~

D. It CONIHACIOR 65 AUIHOH8e'l D B O PHINI, P4 i ASE PHOvet)E THL f OL LOWING INF OHM A110N N.assit.as of Cogeien Psinted f st;as.eted Cornnesition Cost f ssin.sted Printer.g Cast

9. TYPE OF DOCUMENT (ChefA 8PP'C)P"8fe 00K/

e. It CHNICAL RLPORT O)FOf4 MAL __ . _ _ _

(2) $NIE RIM

t. CONF E RE NCE P APE R _

m TiTot Or cow e ntNCt PArt a: ASCE_CoMLrenCe. _ - _ _ _ __

.2, O ATus> Or CONI t Ht NCt: Sept. 10-12, 1984 (JD LOCATION OF CONI E HE NCE.: Raleigh, NC

c. OTHER (Ind.sete type of oseon a p. shesos sprectL jouonal erusle gu de etc)

'1 %. SPt C1 AL On51 RIBU1lON $$ted est sepues no she D.senteven Serv.ses Boonset D.vos.on el Trehn*selinformeuen end Decoment Conrool) 15pessiv spes.*Iinssowst oes such as

• hteEr see*Ieble only es speso!selle appro.ed tv progrom oftas A* er " Send se arreshed oddressees * $winen eds**essed ese. Lag letals for spearel d ssntenen Conknee enssouchons en reeerse or orporese sheet of necesseoy) 9 I

11. PAT E NT CLC t.R ANCE (1/ #ppl, cable / 12. SUHMI T If O BY

• • " " " " * "' " "'R m R mriC m OR m u ms OR nr,e.,,na, e .. .e d s .n .i. .d. sh,n.d N H C e s .o 4 2c A io,niher

.th st.e esisted dos wenosos foe re love.

TO App.opetete resent Counoed

e. PAf t NT CLE AR ANCf NOI Hf OutHL D b. Of F tCI AL*5 ORC A NIZ ATION AL UNIT ts. P AT I NI CL f Afi ANCE Ofl Afd i f D

c. r A nyN reiAHArpsiHesosD . _ DepL of MUCledr_ Energy _ _ _ _ . _ . _ _ _ _

1 P All N T LOUN5E L*5 Lit.PJ A l uM E DAIt c. SIG rJ A T U 8t E f e v e^e.. sed s enbesse, en.s.el en h;*C A8eawf DATE W.Y. Kato, Deputy Chairman 'i

([g 5 gj I-

. ]

..