Research on Streamlining Seismic Safety Evaluation of Underground Reinforced Concrete Duct-Type Structures in Nuclear Power Stations - Part-1. Scope, Objectives and Major Results of the Research

Research on Streamlining Seismic Safety Evaluation of Underground Reinforced

Concrete Duct-Type Structures in Nuclear Power Stations

-Part-1. Scope, Objectives and Major Results of the R e s e a r c h -

Aoyagi Yukio 1~, Kanazu Tsutomu 1), Endoh Tatsumi 1) and Okaichi Akihiro 2)

1) Abiko Research Laboratory, Central Research Institute of Electric Power Industry, Chiba-ken, Japan 2) Kansai Electric Power Company Ltd., Ohsaka-shi, Japan

ABSTRACT

In Japan an extensive review has been made on the earthquake safety evaluation of civil engineering structures since the Great Kobe Earthquake occurred in January 1995. Also, underground important reinforced concrete structures in nuclear power plants are critically scrutinized in line with the streamlining the seismic design. Electric power industry has been jointly conducting research project on this specific type of nuclear power related RC structures since 1987, dividing the research into two phases, the first from 1987 through 1991 and the second from 1997 through 2001. The paper introduces the historical background and the achievements of the research and the concept of seismic safety assessment guideline to be concluded within 2001. The paper serves as an introductory briefing on the series of papers, which follows as Part-2 through Part-6.

INTRODUCTION

Despite the world trend of diminishing nuclear energy, in Asia, especially in East Asian countries including Japan, Korea and China, nuclear power development is still gaining momentum in the supply of electric power. In those earthquake prone countries, availability of convincing seismic design for nuclear plants is of prime importance in gaining public acceptance. On the other hand, streamlining the earthquake resistant design is another crucial factor to keep the nuclear energy competitive with other energy sources.

This series of papers deals with the research on interactive dynamic behaviors of soil and embedded reinforced concrete

(RC) duct-type structures with a view to streamlining the seismic design and structural safety evaluation of such structures. The research has been supported and participated by all the major electric power companies in Japan since 1987.

B A C K G R O U N D OF THE RESEARCH

In Japan the first design specification of RC structural members based on limit states was officially announced in October 1986 by JSCE (Japan Society of Civil Engineers), which allowed us to apply limit states concept in place of allowable design utilized up to that time. The new limit states design, which basically consisted of partial safety factors, was considered to be effective also in rationalizing and economizing the design of nuclear power plant related RC structures. To materialize the design method applicable to what we call civil engineering RC structures in nuclear power stations, a working study committee was organized, being participated by professors and design engineers specializing in this field. The research started as a common project across the electric power industry in Japan as early as 1987. The committee is headed by Prof. Okamura, then Professor of Civil Engineering at University of Tokyo.

After five years' strenuous experimental as well as analytical works, in September 1992 a safety assessment guide" Safety Evaluation Manual for Earthquake Resistant Design of Important Civil Engineering Structures in Nuclear Power Plants"[ 1 ] was published. The target structures were RC underground ducts for accommodation of emergency cooling water pipes, intake water pits, emergency cooling water channels etc, (Fig.- 1). The basic philosophy of seismic design was that two levels of design earthquakes (S 1 and $2) were considered in the safety check-up, in which the structural ultimate capacity of the section checked shall be satisfied against $2, while the section shall remain within elastic limit or the stress in steel shall not exceed yield stress against S 1. The economical advantage of applying this limit states design was that we could reduce the section depth and in some cases the area of reinforcement by taking into account the reduction in section stiffness caused

SMiRT 16, Washington DC, August 2001 Paper # 1294

by cracking and/or yielding in steel.

In the meantime in January 1995 Great Kobe Earthquake occurred leaving devastating damages and claiming thousands of human lives. But what shocked us most was the collapse of center pillars in a metro station in Kobe (Fig.-2). We had to face the fact that possibility of collapse in underground RC hollow structures could not be ruled out. The opening eyes event motivated us to restart the second phase of the research in line with the review and revision of earthquake resistant design of civil engineering structures attempted as an urgent task by JSCE. The focal point of the second phase was clarification of dynamic interactive behaviors of soil and RC structures embedded. Also, the development of analytical method achieved during the past ten years should also be incorporated. Large scale laminar shear box tests were conducted in this phase, which are discussed in the papers to be followed in the series.

In the framework of JSCE, concept of seismic design has been shifting from that of limit states design to that of performance based design. The structural safety performance must be checked in the light of two or three levels of design earthquakes maintaining the functions demanded depending on importance of the structure.

The second phase of our research, which will be completed in March 2002, aims at drafting a seismic safety assessment guideline for important underground RC structures in nuclear power plants, considering the performances required during their expected life spans.

i CONCRETE

CULVERT

11

I2LERuTcTuRE

TURBINE

BLDG. REACTOR

BLDG.

.... - ... ~ I:ii~i~;~-~!.i;~!:!~i;:::-;i~il;i;;i;:¢:~"!~!~l i l L ~ ' ~ : ~ ~ , '

tofQ

CIRCULATING WATER PIPE SUPPORT

Fig.-1 Important Civil Engineering Outdoor RC Structures to be classified as As and A Classes in Terms of Seismic Safety Assessment

(Yellow colored structures)

a) Caving-in of ground surface b) Collapse of center pillars

OUTLINE OF EXPERIMENTAL STUDY

In phase-1 of the research, which finished in 1991, static laminar shear box test was conducted, in which two- box-type RC duct models with a scale of 1 to 4 with respect to actual structures were buried (Fig.-3). The RC models were forced to deform back and forth in shear through the deformation of the dry sand in the laminar box. The maximum shear deformation angle in RC models of 2% was attained, rendering tensile reinforcements well beyond the yielding strain. In this experiment it was confirmed that the relative shear stiffness of the RC structures with respect to that of the displaced soil played a decisive role in determining the deformational modes of the structure embedded. In other words, consideration of stiffness reduction caused by cracking in concrete as well as yielding in steel is a key issue in rationalizing the safety evaluations against severe earthquake effects [2] ,[3].

• ,~, ...

Surcharge (steel ingots)

Sand fill + i /

i I (drysand) | i / Load applied i

/

/

• ' ~

Loading

wall

Reinforced concrete duct

a) Test scheme

2

0

(D

-1

(D

-2_2

I Measured

] No.2 model I"-~.'~'..~i'-/~" ... \

I Measured: I ] / ~

-1 0 1 2

Shear deformation angle (10 " 2 )

\

Initial rigidity(No. 1 model)

Initial rigidity(No.2 model)

No.1 model; less stiffer J model without haunches and tensile reinforcement of 0.2%

No.2 model; stiffer model J with haunches and tensile reinforcement ratio of 0.44%

b) Relative shear deformational patterns of stiffer and less stiffer models

Fig.-3 Static Laminar Shear Box Test with embedded RC Duct-type Models

Another important item of study in phase-1 was the shear strength estimate in the comer regions of RC ducts (Fig.-4). Experimental results explicitly showed that in the comers of RC frame structures the shear strength expressed in nominal shear stress substantially exceeded that predicted by the provisions specified in JSCE Limit States Design. The design formula was basically derived from the tests on simply supported beams. A number of tests were conducted on RC specimens simulating the conditions in the comer regions. The conclusion obtained from the experimental study contributed to the reduction of design stresses, which in most cases dispensed with the arrangement of shear reinforcements [4].

behaviors between soil and RC box- type structures, and to check the validity and adaptability of the dynamic analytical methods available for practical use. The experimental procedures and the test results are described in Part-2 [5].

4.9m

acks . . ~ ~ 1 Hydraulic j _ ~_~- ~..~... ~...~ ~ ~ ~

~. ___~ :iii! iiii i!ii ~"~ SPecimen

==~ iii::i:: Hydraulic jacks

r / 2

. , . . ~

o %

O o

5

Overly conservative

JSCE CODE

~

Proposed method

a) Test scheme for actual size specimen, b) Comparisons of experimental results with JSCE code and the formula proposed by the authors.

Fig.-4 Shear Loading Test in Corner Regions of RC Duct-type Structures conducted in Phase-1.

ANALYTICAL METHODS

The first important step to analyze the dynamic interactive behaviors between soil and embedded RC structures is how to properly model the two materials, that is, soil and RC structural members. In some cases, joint models between the structure and soil need to be incorporated especially when substantial slippage is expected at the interface layer.

Two types of hysteresis dependent models are conceivable for soil; One is a total stress model, which is often represented by Ramberg-Osgood model [6], and the other is an effective stress model such as that proposed by Nishi [7]. The latter is considered more sophisticated in that it is capable of taking into account the elasto-plastic properties including dilatancy of soil.

For RC structural components, two types of models are also proposed; One is the so-called "macro model" by which hysteresis dependent restoring force characteristics are expressed by moment-curvature relationships of RC flexural members. This model is useful only when flexural deformation is predominant. Takeda model is proposed dependent on the level of axial forces [8]. The other is what we cail "micro model" such as developed by Maekawa et.al [9], which is based on non-linear constitutive equations defining the structural behaviors of reinforced concrete elements.

The combination of the model types for two material components leads to four kinds of analytical variations, which are tabulated in Table-1 as methods A, B, C and D. These methods are discussed in Part-3 thorough Part-5 [10, 11, 12]. Method D has not yet been tried since it is still too tedious and requires much computational time, which makes it inappropriate for practical safety assessment analysis at least for the moment. Each method has its pros and cons, the mutual relationships of the methods being illustrated in Fig-5.

Table-1. Characterization of Analytical Models

Method Soil RC members

A

D*

Total stress.

Hysteresis dependent. Ex. Ramberg-Osgood M.

Total stress.

Hysteresis dependent. Ex. Ramberg-Osgood M.

Effective stress.

Elasto-plastic & dilatancy. Ex.Nishi-Kanatani M.

Effective stress.

Elasto-plastic & dilatancy. Ex.Nishi-Kanatani M.

Macro model Restoring force

Characteristics Ex. Takeda M. Micro Model. Material non-linearity Okamura-Maekawa. Macro model

Restoring force Characteristics Ex. Takeda M. Micro Model. Material non-linearity Okamura-Maekawa D*: under development

Method A

-Flexural deformation of members -Behaviors of contact plane / I -Dynamic earth p r e s s u r e ( ~ , , ,

\ b , \

7 [ " The effect of ~ \ \

[/

[ of non-linear model Exoeriment dilatancy on the j] /\ for flexural RC _~ _ flexural

/]

~ e m b e r s . ~ ~ f o r 2 a f i ; n r 0 f ~ /

Method B [ Method C

-steel strain I-up-down movement -concrete cracking [-vertical settlement of -section forces I ground surface

I-dynamicearth pressure

Fig.-5 Relative Relationship among Analytical Methods A, B and C.

OUTLINE OF SAFETY ASSESSMENT PROCEDURES

The target structures classified as "As" and "A" in terms of seismic design are divided into two categories depending on the function of the structure. The first group includes the structures to support emergency safety equipment and the second one refers to the structures, which are important by themselves. RC ducts to accommodate emergency cooling water pipes, intake water pits to support emergency pumps etc, belong to the former, while the emergency intake water channel is an example of the latter. The prime requirement for the first group is to keep the function of the supporting components or equipment during and after earthquakes. Seismic importance of the structures is classified into two categories; As and A, for which design earthquakes $2 and S l are to be considered, respectively.

The performances demanded are expressed in rather qualitative manner, but the target performances are specified in more quantative fashion as shown in Table-2. The flow of seismic safety assessment is given in Fig.-6. Application of dynamic interactive analysis is mandatory, in which up-down acceleration equivalent to half the horizontal design value must be considered simultaneously in such a way as to give rise to the m o s t unfavorable condition.

Table-2 Demanded and Target Performances for Important Civil Engineering Outdoors RC Structures Types of structures

Structures supporting As class equipment or piping.

-Intake water pit

-RC duct for emergency cooling water pipes

Structures classified as As by itself.

-Emergency cooling water intake channel

Structures supporting A class equipment or piping.

-Emergency gas treatment duct.

Demanded Performances Maintaining the functions required for supported equipment or piping in case of emergency.

Assuring passage of required water flow.

Maintaining the functions required for supported equipment or piping in case of emergency.

Target performances To be checked against $2 1) Prevention of collapse for

supporting structures. 2) Limiting the overall shear

deformation angle, ultimate shear capacity etc

No collapse or excessive deformations against $2.

To be checked against S 1. 1) Prevention of collapse for

supporting structures. 2) Limiting the overall shear

(1) Maintenance of Demanded Performances -Identification of Limit States including the Equipment to be supported

/

Structures not to support equipment I Structures to support equipment

i

1(2) Prevention o f Structural Collapse I Assurance of space for

i

installation of equipment, Aspect of seismic performance structural integrity etc,

(3) Stability of Structural Behaviors (Prevention of Brittle Failure, Drastic Softening etc,)

Physical phenomena

(4) Prevention of Cover Concrete Spalling ( Checking Buckling of Compression Reinforcements) - Limitation of Currently Available Analysis -

(5) Prevention of Shear Failure

Limit State for Flexural Failure

1(4) Prevention of Cover Concrete Spalling 1

I

i

General case For duct-type structures

( 6 ) Compressive Strain in D fi "' - "

* (7) Structural Shear

_ _ < _ _ i e ormanon Angle < 001

Cover Concrete 0 O1 i

, " [ Or < given by Eq. 1

/

... "T

(8) Corresponding Values obtained from Non-linear Interactive Dynamic Analysis

(Micro Model)

(9) Corresponding Values derived from Dynamic Interactive Analysis based on Member Non-linearity ( Macro Model)

Fig.- 6 Flow for Performance Assurance for Important underground RC Structures

Fig.-7 Verification of Target Performance with Respect to Analytical Results for the Case of Flexural Failure

Analytical models should be selected from the confirmed ones, keeping in mind environmental conditions, characteristics of the structures etc. The dynamic analysis should be started with the stresses due to permanent loads as initial conditions.

Target performances of the structure in terms of safety evaluation are checked as described in the flow chart (Fig.-7). The target performances are given in two ways; one is average compressive strain in the cover concrete in compression zone. The limiting strain is specified as 1%. To compute the strain, method B or D is recommended. As is discussed in Part-6 [ 14], this level of compressive strain assures enough margin of safety before the collapse of the structural member as long as the failure mode is basically flexural tension. The other verification is to ascertain the shear deformation angle in the case of duct-type structures. If the calculated shear angle of the duct does not exceeds 1%, the structure is unconditionally judged as safe. This is based on the pushover analysis conducted for the conceivable cases. In Fig.-8 the analytical results of the pushover analysis are plotted, which were obtained for the cases covering the conditions to be encountered in the practical design. Except for a few cases, the limiting shear deformation angle of 1% satisfies the 1% strain in cover concrete with a comfortable allowance even when axial forces are present. Analytical methods A and C suffice for this shear angle check.

= .~ 0.05 o

,

~ 0.04

~ 0.03

t D

~ o 0.02

= ~ o 001 e~0 o

• r / ~

• ~"~ ~ 0 • " ~ O

o

" - r - - ' f - - " r ' - r - " " - - r - " r - - T - - r - ' - - r - T - - T - r - - - - r ' - T - - 1 - - - r .. . . r--1--'T'-T-- - - T - " v - r - r -

O

. . . . . .

% '.

. . . 9. . . . ! . . . ! . . .

0 ' '

o o o i :

o o i !

0 ~ , . ',0 o i ; '

;~---i ... o-i ... d ' ° ... i ... i ...

-T~O ', ~ ', O0 ',0 , . , ~ :

~ o : : oB ::o ®b o~ :: i

' o ~ i ® ~ : : ° o : : % ~ o :: o ° ~ ~ . . . i . . . " . . . ~ .. . . ~ .. . .

, , , , , , , , i , , , , , , , , , , , , , , , ,

0 10 20 30

Ranges in Properties of RC Duct-type Structures Analyzed.

Intemal length of duct 2.0 to 15.5m Wall thickness ...0'3 to 2.25 m .

Height/Wall thickness ratio 1.9 to 15.1 Over burdening soil cover

Tensile reinforcement ratio Concrete compressive strength Yield strength of steel

0 to 30 lI1 0.25~2.5% 20~50 N/mm 2 200~500 N / r a m 2

Embedment depth (m)

In case the calculated angle exceeds 1%, the limiting value, which is given as a function of structural properties in Eq.-1, is allowed to increase if the specific section parameters are met as described in Part-6 [ 14].

R = R ( h , t , fc, fY, o,, or0) Eq.-1

R : Limiting deformation angle

h: Height of wall, t: Thickness of wall, fc: Compressive strength of concrete, fy: Yield strength of steel, o t: Tensile reinforcement ratio, cr 0: Axial stress in wall

Durability of the RC structures is also a crucial issue to be considered. The structure must withstand the design earthquakes even at the last moment of its design life without impairing the demanded performance. Fig.-9 compares the load-shear deformation angle relationships of intact and severely corroded RC single box ducts with actual dimensions [ 15]. The sound structure exhibits higher loading capacity, but for the corroded one ductility was somewhat improved in spite of reduction in ultimate loading capacity by about 20%. Two criteria should be satisfied with respect to durability.

1) Concrete should be crack-free due to corrosion of steel by ingress of chloride ion as well as the effect of carbonation. 2) The ingTedient materials and proportions of concrete should be so selected that the concrete would not deteriorate due

to freeze-thaw effects if any.

500

400

300

200

~ '

100

Q

0

o -100

i~

' r A

- oo

-300

v

-400

I

-500

.,-?

2.~5m

-4 -2 0 2 4 6 8 10 12

Shear Deformation Angle (%)

intact Specimen with Hanches

Severely Corroded Specimen with Hanches

Intact Specimen without Haunches

Fig.- 9 Shear Deformation Angles as affected by Corrosion in Steel and Presence of Haunches

FUTURE W O R K S

To our regret civil engineering RC structures in nuclear power plants are currently designed by old allowable stress approach. We hope our proposal will be implemented in seismic safety assessment and contribute to the streamlining the design and safety assessment.

A C K N O W L E D G E M E N T

The foregoing study is a part of the common research entitled "Development Study on Verification Method of Seismic

Pelformance of Underground Re#forced Concrete Structures in Nuclear Power Stations (part-2)" which is supported by

Electric Power Industry in Japan and extends from 1997 through 2001. The managing company is Kansai Electric Power Co. Ltd. The authors are very grateful to the concerned of the power industry. They also appreciate valuable advices given by the committee organized in JSCE and chaired by Prof. Hajime Okamura, President of Kochi Institute of Technology.

R E F E R E N C E S

1. Nuclear Civil Engineering Committee, JSCE " Safety Evaluation Manual for Earthquake Resistant Design of Important Civil Engineering Structures in Nuclear Power Plants" (in Japanese) Sept. 1992

2. Aoyagi, ¥., Endoh, T., and Katahira, F.," Experimental Study on Soil-Structure Interaction of Underground Reinforced Concrete Ducts Subjected to Earthquake Loading," Transactions of the l lth International Conference on Structural Mechanics in Reactor Technology, Vol. K, paper K14/6, August 1991, pp. 387-392

3. Tohma, J., Iwatate,T., Ohtomo, K., Satoh,. and Katahira, F., "Determination Method of Critical Seismic Loading on Outdoor Important Civil Engineering Structures," Transactions of the l lth International Conference on Structural Mechanics in Reactor Technology, Vol. K, paper K14/4, pp. 381-386, August 1991

4. Aoyagi, ¥., and Endoh, T., " Shear Strength in Comer Region of Reinforced Concrete Duct-Type Structures to be embedded in Soil," Transactions of the 12th International Conference on Structural Mechanics in Reactor Technology, Vol. H, paper H09/3, pp. 289-294, August 1993

5. Ohtomo, K., Suehiro, T., Kawai, T., and Kanaya, K., "Part-2. Experimental Aspects of Laminar Shear Sand Box Excitation Tests with Embedded RC Models" To be published on Transactions of the 16th International Conference on Structural Mechanics in Reactor Technology, Vol. K, paper number 1295, August 2001

6. Jennings, P. C., "Periodic Response of a General Yielding Strucrure," Proceedings of ASCE, EM2, pp. 131-163, 1964 7. Nishi, K., and Kanatani, M., "Constitutive Relations for Sand under Cyclic Loading on Elasto-plasticity Theory," Soil and

Foundations (Japanese Society of Soil Mechanics and Foundation Engineering), Vol. 30, No. 2, pp. 43-59, June 1990 8. Takeda, T., Sozen, M.A., and Nielsen N.N., "Reinforced Concrete Response to simulate Earthquake," Journal of Structural

Division, ASCE, Vol. ST12, pp. 2557-2573, 1970

9. Okamura, H., and Maekawa, K., "Non-linear Analysis and Constitutive Laws for Reinforced Concrete Structures," Gihoudou-Publishing Company Ltd. May 1991

10. Matsui, J., Ohtomo, K., Kawai, T., and. Okaichi, A., " Part-3. Analytical Simulation by Simple Macro-models for Soil and RC Structures" To be published on Transactions Of the 16th International Conference on Structural Mechanics in Reactor Technology, Vol. K, paper number 1296, August 2001

11. Matsuo, T., Ohtomo, K., Matsui, J., and Okaichi, A., " Part-4. Analytical Simulation by Sophisticated RC Micro-model and Simple Soil Model," To be published on Transactions of the 16th International Conference on Structural Mechanics in Reactor Technology, Vol. K, paper number 1298, August 2001

12. Kanatani, M., Kawai, T., Matsumoto, T., and Okaichi, A., " Part-5. Analytical Simulation by Sophisticated Effective Stress Soil Model and Simple RC Macro-model," To be published on Transactions of the 16th International Conference on Structural Mechanics in Reactor Technology, Vol. K, paper number 1389, August 2001

13) Aoyagi,¥.,, Minh, N.N., and Kanazu, T.," Nonlinear Dynamic Interactive Analysis of Embedded RC-Box Culverts in Laminar Box Filled with Sand" Proceeding of the GEOTECH-YEAR 2000, Development in Geotechnical Engineering, Vol. 2, pp. 291-300 November 2000, Bangkok, Thailand

14) Miyagawa, T., Matsumoto, T., Aoyagi, Y., and Kanaya, K., " Part-6. Verification of Ultimate Load and Ductility Capacities of RC Ducts," To be published on Transactions of the 16th International Conference on Structural Mechanics in Reactor Technology, Vol. K, paper number 1307, August 2001