ML19329D534

Rept to Aec,Adequacy of Structural Criteria for Crystal River 3 & 4, Draft

ML19329D534
Person / Time
Site:	Crystal River, 05000303
Issue date:	02/28/1968
From:	Hall W, Hendron A, Newmark N NATHAN M. NEWMARK CONSULTING ENGINEERING SERVICES
To:
Shared Package
ML19329D533	List: ML19329D532 ML19329D534
References
NUDOCS 8003160133
Download: ML19329D534 (14)

Text

. ,*,.... . ...

('v.') i J A 's" I l T. I I !.1, i J C ',*/ s.1 A R R ss CO;;CU: Til:0 Ci Cl;;;;r.it;G CCRVICC3 1114 CIVIL EliGITJCCrilt;G CUILDi?.G URDANA. ILLit;Oi3 G1001 CRAFT REPU3T TO AEC REGULATORY STAFF ADE(U"CY CF TbE STRUCTURAL CRITERIA FOR CIiYSIAL RIJER UNITS 3 AtiD 4 f;UCLEAR GEtiERATIt;G PLAliT Florida Power Corporation .

t (AEC Cocket i;os . 50-302 and 50-303) by ti . M. t'ewrra r k ,

W. J. Hal1, and A. J. Hendron, Jr.

l l

l l

February 1968 t

8003160I3 3

~

7_; ,

..'"..,,.. ,m ADEQUACY OF THE STRUCTURAL CRITERIA FOR ThE CG STAL UllITS 3 AND 4 NUCLEAR GENERATIh9 PLANT by it. !1 Net. r.o r k , W. J. Fall and A. J. Hendron, Jr.

INTRODUCTION This report cencerns the adequacy of the containnent s tructures and comporgnts , reactor pipir.g and reactor internals for the Crystal River U. ;s 3 ar.d 4 Nuclear Generating Plant, for which application for a construction permit f.as been made to the U. S. Atomic Energy Commission (AEC Cockets ha. 50-302 and 50-303) by the Florida Po.ser Corocration. The facility is to be lc atee in the northseestern portion of Citrus Count y, Flor ida, on the C u l f o f l.2xico betw2cn the rc.suins ci t he Withlaccochee and Cr ys tal Rivers , and apuronima tely 7i mi l e s I.'.i c f C r y s t a l R i .e r , and 70 miles N of Tampa, Florica. Specifically this report is concerned with the evaluation of the design criteria that de se rrai ne the ability of the cantainment s ys tem, piping, and reactor internals to withstand a design carthquake acting simultaneously with other applicable leads forming the basis of the design. The facility also is to be des igned to withstanJ a maxirc.ca car thquaka s itaul eancously wi th other applicable loads to the extent of insuring sa fe shutdo.in and conta inment. This report is based on i n f or ma t ion a nd criteria set forth in the Preliminary Safety Analysis Report (PSAR) and supplc.Tants there to as listed at the end of this report. '.le have par t ic i pa t ed in discuss ions with the AEC Regulatory Staff, and the applicant ar.J its consultcnts, in v.hich cany of the design criteria were discussed in detail.

DESCRIPTICM OF THE FACILITY The Crystal River h clear Generating Plant is described in in: PCAR cs c pressuriced teater reactor nuclear steam supply system furnish:d . ; v.2 Lc t :c ;.'t cod '.liican Cc..pany and des igned for an init ial po tar cutput of < C. .'-

.e

m.. _ . - - - .

N (03 5 li..'O nJc) ict ccch unit. The reactor coolant system cons ists of the rcccice vessei, cccica puxps , s team generators , press ur i zer , cr.J i r.; c rccane c t i ng piping. For cc;h rec tor there are two steam generators. The rec :ce vessel uili have ca inside dicmeter of about 14.3 ft., a height of 41.7 it., is des igacd icr a pressure of 2500 psig and a ten.perature of 650 F, cnJ is m:Je c/ S/.-302, Grcdc 0, steel clad with type 304 austenetic stainiess steci.

- The cor.c o i rc.c nt for this plant cons is ts of two s ys tems cs fci lc. 3 :

(1) the recctor Luilding which provides biological and missile shicidir.g, cnd t.hich ccatains tac energy cnd material that might be releascu by cn accident; c r.) (2) the er.gineered safeguards systems which limit the maximum valta of the er.crgy reicascJ by an acc ident.

Tne rec cer malicir.g, which encloses the reactcr cnd steem car.arctors, con:ists cf c steel 1 ed concrete shell in the form of a reinf ccced ccr.;rcte vert ical cylinser with a flat base and a shallow dome roof. The cylindrical structure of 130 ft. ir. side diar.nter has side walls rising 157 feet frca the top of the icunJatica slab to the spring line of the donc roof. The ccr. crete s ige walls of the cylinJer and dena will be approximately 3 f t. 6 in, cr.J 3 ft. O in. in thickness, respectively. The foundation not will Le cpproxima tely S f t. thick with a 2 ft. thick concrete slab over the bottom liner plate.

The fcundat ica 310a uill be reinforced with convent ional s teel re inforcing.

The cylindrical un'is vill Le prest ressed with a pos t-tens ioning sys tcm in the vert ica l c..J .'.cr iccat a l d i rec t ions. The doca roof will be prestress:J c t i l i n i r.) a 1. re c-uc / r cs t- tens ioni ng s ys t em. The inside suricce ci U.a rccctoc builcir.g uiii te lined with a carbon steel liner 3/3 in. thich for I ti.: cfiir.dcr c..J d w.: cnd 1/4 in. thick for the base. The rccctor buiiving is c ss er.t ia l i / th: scme cs the containnant buildings for the Turkey foin;. C ac a., l cr.J 1hree iliic Isicna ;)lcats.

, .b' . ' .:*

-3, Personnel and equipment' access hatches are provided for access to the reactor building. In addition there are other penetrations for piping and electrical conduits.

The engineered safeguards for each nuclear unit consist of the energency core injection system and the reactor building atmosphere cooling and. washing system.

Other Class I components and systems whose des ign must include .

consideration of seismic ef fects are listed in Appendix SA and . include such items as the spent fuel. cool ing sys tem and shutdown cool ing s ys tem, reactor cont rol room and equipment , and the oost-incident air filtration system. Some of these items are located totally or par,tially outside of the reactor building.

The facility includes a cooling water intake and pump structure located at the foot of the intake canal about 400 ft. from the reactors.

The bedrock at this site is located approximately 20 ft. be ne a t h the present. ground surface The surface overburden consists in the upper

' layers- of approximately .3 to 5 f t, of surface fill, followed by the natural soil cover cons isting of -deposits of -thinly laminated organic sandy silts and clays interspersed with marine deposits, and in turn overlying a residual limy

" oil unit derived from the decomposition of the underlying bedrock. The bedrock consists of biogenic carbonates of Tertiary Age. The uppernost bedrock member is that identified as the Inglis ' member which is characterized by a cream-colored to an occasionally tan, porous, granular, biogenic limestone and dolomite deposited in a shallow marine environment, iThe closest evidence of possible faulting occurs at a distance of three miles to-the east of the site. Studies of the site show no evidence of exis tence -of subsurf ace f aul ts.

,- v -- - ,

.k.

SOURCES OF STRESSES IN REACTOR BUILDING AND CLASS I COMPONENTS The reactor building is to be designed for the following loadings:

dead lodd; live loads (including roof loads, pipe pene trat ion reac t ions, and crane loads)'; internal pressure due to loss of coolant accident of 55 psig; test pressure of 63.3 psig; negative pressure of 2.5 psig; accident temocrature of 2819F and operating temperature of 110 F; wind loads corresponding to roof line load of 35 psf; tornado wind loading (300 mph tangential wind veloc ity, external vacuum of 3 psig and missile loading); prestressing loads; and seismic loading as described next.

The seismic design of the reactor building is based on the response to a maximum horizontal ground acceleration of 0.05g. Also, the design is to be checked to insure no loss of f unct ion for an earthquake based on a maximum horizontal ground acceleration of 0.10 9.

The piping, internais, and vessel support design procedure is outlined in answer to quest ion 9.11 of Supplement No. 1. Therein i t is noted that these 4

items will be des igned for various loading combinations as listed in Table 1, 1

j including the design' load, the design earthquake and pipe rupture loads.

In addition a discussion of modes of deformation of reactor internals, and the allowable deforcations are presented in Table 2.

As noted in Appendix 5A, all class I s t ructures, components and systems will be designed for primary steady-state stresses combined with the appropriate seismic stresses, and where applicable, in accordance with the appropriate codes. In the case of primary steady-state ' stress combined with the seismic stress resulting from the maximum earthquake, the response is to e

e r w

be limited so that the funct ion of the component, s ystem, or s t ructure shall not be impaired to prevent a safe and orderly shutdown.

COMMENTS ON ADEQUACY OF DESIGN Foundations The applicant has proposed to found the mat foundation for Crystal River (Jr11ts 3 and 4 on 3 structural fill composed of crushed l i ne s tone. The base of' the structural fill is planned to be at about elevat ion 73 and will extend up to about elevat ion 80. Quality cont rol of the crushed linestone fill and the 98 percent maximum Modified censity (ASTM Test Designation 01557-66-T) requirement as noted in Ref 3(a) will be adequate to assure a structural f il l with sat is factory s t ress-s t ra in proper t ies.

Because the e>ploratory investigation revealed the presence of both open and filled solution cavities in the- l imes tone bedrock 'benea th the s i te, the applicant proposes to undertake consolidation grout ing beneath the reactor building to about elevation 30 and beneath other structures to about elevation 60. From the informat ion presented in the founda t ion grout ing repor t on Unit 2 (Ref. 3(b)) and the report on the test . out ing program for Crys tal River Units 3 and 4 (Ref. 3(c)), it appears that the mod i fied spl i t-spaced hole procedure utilized on Unit No. 2 will be adequate for the foundat ion of Units 3 and 4 The effectiveness in providing a curtain wall around the area to be grouted is illustrated quite clearly by Fig. 5 of Ref. 3(b) which shows a graph of hole order versus unit grout take. The graph illustrates that the grout takes-approach reasonable limits in the Tertiary and Quaternary

-holes. ~It is understood that the grouting specifications for the grouting contract are flexible to the extent that the decision on the hole order at which grouting will be stopped is to be decided by the field engineer. It

94. 6 .

would be our recommendation in the appilcation of this procedure that the unit grout takes be reduced to 0.5 to 1.0 cubic feet per l'i~neal foot f hole be fore grout ing is s topped. From all the available data for the site thus

~

far (Refs. 3(b) and 3(c)) It appears as if this result will be accomplished on either the Tertiary or Quaternary consolidation grout holes at 8 ft.

and 4 ft, spac ings , respect ivel y, if a cur.tain wall is first establ ished around-the area to be grouted. We believe the proposed structural fi11 and grouting program will be adequate to prevent excessive dif ferential supplement of the reactor buildings and appurtenant structures.

Seismic Desion All s t ructures, components , and sys tems class i f ied in Class I are to be designed for a design carthquake based on a maximum horizontal ground acceleration of 0.05g. Such items are also to be designed for a maximum corthquake based on a maximum horizontal ground acceleration of 0.109 so as not to impair or 1.revent a safe and orderly shutdown of the plant. These desis levels are in agreement with those proposed by ttie U. S. Coast and Geodetic Survey (Ref. 4) and we concur in these design criteria.

The response spect ra to be employed in the design are given in Fig. 3 of Appendix 21. The response spectrum shown is for five percent gravity, the design earthquake, and we concur in the use of the spectra as shown on tbc assumption that at periods greater than 1.0 sec. (not shown) the spectra do'not drop sharply but remains essentially at the spectral velocity levels at which the present plot is cut off. We assume that the response spectra for use in desig,n for the maximum earthquake loading condition will be twice the values of the spectra just . described.

~

'~

The vert ical component of earthquake excitation will be taken as two-thirds of the horizontal component and will be assumed to occur simultaneously with the horizontal component. We concur in this criterion.

It is noted in Section 6 of Appendix SA that "The respective vertical and horizontal seismic components at any point on the shell will be added by summing the absolute values of the response (i . e. , s t res s , s hea r , moment ,

or deflection) of each contributing frequency due to vertical motion and ad'ing d the resultants to the corresponding absolute values of the response of each contribut ing f requency due to horizontal motion." The seismic stresses are then to be added directly to the dead load, live load, operating loads, and accident (pressure and temperature) loading conditions in accordance with the loading express ions presented in Appendix 50. From this one can infer that the scismic stresses are added linearly and directly with the other applicable stresses, and on the bas is of this assumpt ion we concur in the design approach.

The damping values to be employed in the dynamic analysis are given in Section 5 of Appendix 5A. These values are to be employed for both the maximum and design earthquake. As noted in answer to Question 9.3 of Supplecent 1, a damping value of 5 percent of critical will be used for both the design and maximum earthquake for rocking ef fects for the foundation. We concur in the use of these values in tae design.

The general method of dynamic analysis will be either a modal analysis or-will be carried out in accordance with the procedure outlined briefly in Section 6 of Appendix SA. The discussion presented in Section 6 suggests that for systems such as piping systems which are highly complex geometrically, g

that the analys is may be carried out as for a s ingle-degree-of-f reedom system.

We do not corcur in this approach in general, and it is our recommendat ion that

c

_g.

a formal dynamic analys is be per formed for Class I s t ruc tures aquipment, piping, and reactor internals as appropriate, especially for those s ys t ems which are vital to safety of,the plant.

Further information on the dynamic piping analysis is included in the answer to-Quest ion 9.12 and provides sone clarification to the discuss ion presented in Sect ion 6 of Appendix SA. Howe ve r , it is noted that the description given only applies to the dynamic analysis of piping systems supported at '

fixed points. The applicant is requested to provide additional information' concerning the methods of dynamic analys is that will be employed for the piping sys tems. Addit ional comments on the analysis of piping sys tens appear later in this report.

The method of analysis to be employed for the reactor building is described in Section 2.2 of Appendix SC and we concur in the approach as outlined there.

All structures and components classified as Class II are to be designed for a ground acceleration of 0.05g in accordance with the procedures of the Uniform Building Code. We are in agreement with this approach.

General Desion Provis*ons '

The load combinat ion equat ions to be employed in the des ign of the reactor building are presented in Sect ion 1.3 of Appendix 50. We are in general agreement with the combinations to be employed with one except ion, namely that of load exoression'"c" wherein it is our belief that a term reflecting the accident pressure load is missing. Clarification of this point by the 1

applicant is requested. '

The design stress criteria for the reactor building are presented I in Appendix 58 and St. It is noted therein that'the load deformation behavior .

J i

• 'l

9 of the structure is one of clas t ic, low strain response. The building will be checked for the f actored loads and load combinations, compared with the yleid strength of the structure, and the load capacity is to be defined as the upper limit of the elastic behavior of the effective load carrying s tructural mater ials. The deformation of the structure is to be such that the conpress ive s t rain in the steel liner does not exceed 0.005 in/in, nor to cause average tensile strains to exceed that corresponding to the minimum

• yield stress. Membrane tens ien will be limited to 3 /f' and it ,i s no t ed C

further that when principal flexural tension exceeds 6 [f' due to therral gradients through the wall, non-prestressed reinforcing will be added to resist thermal stresses. It would be our recommendation that no ne t membrane tens ion be permit ted in the containnent shell but on the assumpt ion that the lat ter s tatement refers to the combination of membrane tension combined with flexural tension arising from pressure or thermal effects we concur in the general des ign provis ions noted.

The reinforcing steel to be employed in the plant will consist of e ther ASTM A-15, A-408, A-431, or A-432 It is noted in Appendix 50 that arc welding for reinforcing splices will not be employed and that Cadweld splices will be used when required. We are in agreement with this approach.

l The liner is to be designed so that the critical buckling stress will be greater than the proportional limit of the steel. Present analysis, according-to the PSAR Indicates that the basic accident conditions produce a strain of approximately 0.002 in/in, in the liner. The liner is to be analyzed as a flat plate and the liner anchors, which will be vertical angles, are to be spaced horizontally at 18 in. center to center. The liner anchors are to be designed such that the welds connecting the anchors to the liner 4

, ~,'*..

~. ..

will fail before.the liner is breached. Generally we concur in this design approach for the liner, although it is not clear how this type of attachment may affect the buckling strength and long-term service performance of the liner.

A discussion of the general design criteria for handling differential settlements and relative motions under seismic response is presented in Appendix SA and we are in agreement with the general concepts presented there.

The post-tensioning stressing system to be employed will consist of either the S.E.E.E. or the BBRV system. In general, the design concepts to be employed in the prestressing are similar to those employed in other plants designed by Gilbert Associates such as Turkey Point, Oconee and Three Mile Island. The reactor tendons, which are unbonded, will be protected from corrosion by insertion of a protective coating in the tendons. The steel portions of the plant will be connected electrically to provide protection against stray currents. It is noted in the PSAR that the tendon inspection program could be made if it appeared desirable. It is our recommendation that a reasonable inspection program be implemented, especially in view of the location of this plant near a salt water environment.

Pipina,-Reactor Internais, Reactor Ves s e l a nd Ves se l Sucoorts internals, The design approach to be employed for the piping, and reactor which also would include for the most part the design of the engineered safeguard system, are to be designed for general criteria as outlined in the PSAR, namely in accordance with applicable ASME codes and procedures outlined A further more detailed discussion of the design in AEC Publication TID-7024 1.

approach is presented in answer to quest ion 9.11 of supplement The possible; modes of deformation of reactor internals are summarized in Table 2 of the answer to question 9 11 and involve values labeled " allowable"

'8

( g= + . . gs m.

and "no loss of function." It is noted in the discussion preceding the table that the "no-loss-of-function" deformations could cause safety problems, and that the " allowable" deformations are those that are used as design limits. It is not clear whether these design limits refer to those associated with the design earthquake loading condition or the maximum earthquake loading condition or even a combination of seismic. )ading with other applicable loadings. Clarification of this point by the applicant is requested,,in orper that a better judgcent on the margin of safety inherent in the design can be made.

The approach to be employed for the piping appears to be patterned after that presented in ASME applicable codes and in Westinghouse Electric Corporation Report WCAP-5890 Rev. 1, 1967 However, the approach presented is limited in that it relates solely to the margin of safety with regard to stress levels and does not provide information on the nergin of safety with regard to permissible strain or deformation. With regard to the presentation encompass ing poss ible s t rain hardening, no informat ion is presented to form a J udgaent as to whether the stress analysis conforns to real property mate rials, and moreover whether localized stresses or deformations are included in the analysis. Further information concerning the design criteria to be employed for the piping, particularly with respect so the maximum earthquake loading c o r.d i t io n , is requested.

i Ins t rumentat ion and Cont rols The design of the control instrumentation for seismic ef fects is discussed in ans,wer to question 9.13 of Supplement 1. Therein le is noted

'that "the -components in the reactor protect ion system and safeguard evacuation d

S

. .s .~ .

system will suf fer .no loss of funct icn at accelerat ions of 0.19 horizontal and 0.0679 in vert ical condi t ion." A similar comment is given concerning the batteries and battery mounts. We can not concur in_this approach, for an analysis may show that the instrumentation can be subjected to larger accelerations. Also, will the inst rumentat ion function under condit ions of moderate tilting?

Floodinq i Information concerning pose ble flooding of the site is presented in Appendix 2C and in answer to Qu- t ion 9.12 of Supplement 1. The protection provided against flooding appears adequate to us.

Cranes The polar crane in the reactor building is a Class I comoonent and is noted in Appendix SA that the design will be made to insure stability during an earthquake. It is noted in answer to Question 9.10 that other handling bridges which are not considered Class I equipment are also provided wi-h anti-derailing devices. The design criteria for the cranes are acceptable to us.

Penetratinns It is noted in Section 7 cf Appendix SB that the penet rat ions will be designed for the load combinations applicable to the reactor building and will be anal ~yzed by using the finite element technique developed by the Franklin Institute Research Laboratories. Steller penetrations will be des igned in accordance with published and accepted procedures as noted in the discussion presented in Appendix 58 We are in general agreement with the design approaches outlined briefly in Section 7 of Appendix 58.

.5 - ,- v CONCLUSIONS In line with the design goal of providing serviceable structures and components with a reserve in strength and ductility, and on the basis of information presented, we believe the design criteria outlined for the containment and other Class I components including the reactor internais, and piping, vessels and supports, can provide an adequate margin of safety for seis$icresistance. However, in arriving at this conclusion we have.noted; in the report several items for which additional infornetion is required from the applicant, narely informat ion concerning the analysis of the piping under dynamic lo; ding, stress criteria for piping, design criteria for the reactor internals, design of instrumentation and controls, and clarificat ion of the load combination expressions.

REFERENCES

1. " Preliminary Safety Analys is Report , Vols. 1, 2, 3, and Appendices,"

Crystal River Units 3 and 4 Nuclear Generating Plant, Florida Powe r Corporation, 1967 P

2. " Preliminary Safety Analysis Report, Amendments 1 and 2," Crystal River Units 3 and 4 Nuclear Generating Plant, Flcrlda Power Corporation, 1968.

3 " Preliminary Draft Reports (to be filed with AEC)

-(a) " Foundation Investigation -- Proposed Nuclear Power Plant -- Florida Power Corporation," Woodward, Clyde, Sherard and Associates, February 7, 1968.

(b) " Foundation _ Grouting Report - Unit 2," Crystal River Plant, Florida Pcnver Corporation, by Gilbert Associates, Inc. , Report No.1657, January 30, 1968.

(c) " Test Grouting Frogram - Units 3 and 4," Crystal River Units 3 and 4, Florida Power Corporation, Gilbert Associates, Inc., Report No. 1658, )

January 30, 1968

4. " Report on the Seismicity of the Crystal River Site," U. S. Coast and

' Geodetic Survey, Rockville, tbryland, .

\

1

. i l