ML20234D855

From kanterella
Jump to navigation Jump to search
Forwards Matl Re Bodega Bay Atomic Park Unit 1 Re Questions Raised in Meetings at Bethesda,Md & Argonne on 630702, Including Cost of Seismic Protection,Conclusions for Final Rept,Comments on Spectra & Working Stresses as Design Basis
ML20234D855
Person / Time
Site: 05000000, Bodega Bay
Issue date: 07/16/1963
From: Williamson R
HOLMES & NARVER, INC.
To: Bryan R
US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20234A767 List: ... further results
References
FOIA-85-665 NUDOCS 8709220215
Download: ML20234D855 (13)
v • d • e


Text

_-- -_-_ . -

./ -

$1..-w

.m C

)p%@ap$?

g j ? W~

)

HOLMES 8 N A RV E R . I N C.Lg yne COPY.

E N C I N E E R 5 C O N S T R U C T O R.5

  • ere souTw riovcnoA sTntrT LoS ANGELES 17 waoisom um LOS ANGELES HONOLOLO I ..

1 l

. July 16, 1963 to 4

Dr. Robert H. Bryan, Chief '

i l

Research and Power Reactor Safety Branch ~ C l Division of Licensing and Regulation [2- S -8 U. S. Atomic Energy Commission 1 Washington 25, D. C. N g

Dear Sir:

t _

6 The following material relating to Bodega Bay Atomic Park Unit No. 1 is transmitted herewith for your information. It relates to questions which have been raised in meetings at Bethesda and 3 others stemming from the Argonne meeting of July 2.  !

Item 1 attempts to shed some light on the question of the costs involved in providing seismic protection, in response to the l l query which arose in the meetings of June 27 and 28 at Bethesda.

I y , Item 2 comprises an initial attempt at some conclusions for the j . final report as requested in the Bethesda meetings of June 27 and

28.

Item 3 considers the advantages and disadvantages of using working stresses as a design basis as compared to using yield point stresses.

Item 4 comments on the spectra considered at the Argonne meeting of July 1.

Item 5 discusses items which the writer feels should be included in.

the revised criteria, which it is anticipated, will be submitted by the Company.

....7-g 3& '

r.m.,.

8709220215 851217 ES -665 PDR h 5106

)

R.H. Bryan 7/16/63 Comments relating to the problem of defining the critical items of the reactor and the special requirements needed to insure their l integrity were submitted to you at Chicago on July 1,1963, In a preliminary draft dated June 29, 1963, titled Special Requirements for Earthquake Resistant Design.

It is believed that the data herein and that previously submitted consider the currently important factors to the extent feasible l under existing circumstances. Accordingly, I will await further direction from you before proceeding further.

l Sincerely yours, HOLMES & NARVER, INC.

R. A. Williamson enc 1.

l l

6 l 1

1 I

l l

l i

  • -- -.w. . p. . . .e.... . , - . . _ , , . . - , . .

a u _-___ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ _ _ _ _

[' dated' July' 16,l 1963 d.

C. T h . ?

1. Cost of Seismic Protection p g ylle Co;7 ( ,4-c 7 ps/

During our discussions on June 27 in Bethesda, Mr. F.N.*, Watson.

indicated an interest in knowing what the cost penalty might be"in de-signing according to the seismic criteria of the PHSR as com, pared to the criteria conventionally used. This is a particularly difficult l question to answer which can be resolved accurately only on the basis of comparative preliminary designs. The following material, based 1

on updated cost information for concrete and reinforcing steel', amends

]

some previous tentative' conclusions.

To provid'e a very approximate answer, calculations were made for the walls of the portion of the reactor building above elevation + 62.

It is in this portion of the structure that seismic effects appear to have a maximum influence on cost. It was assumed that changes in seismic input would be met by variations in the amount of reinforcing steel, without changing wall thickness, and that concrete with approximately 200 lbs. of reinforcing steel per yd. would cost $65/yd . Reinforcing steel cost was taken as $0.12/lb. It was further estimated that under conventional assumptions the steel percentages in each face would be i

! 0. 25, whereas under the seismic input proposed, on the basis of rough f ., calculations, the percentages of vertical steel in each face would be-c ome 1. 0.

1 The calculations led to the conclusion that the more severe seismic forces being considered would increase cost of the reinforced concrete in these walls by about 40% as compared to conventional construction in seismic areas. This is obviously an upper bound which is far too high to serve as an accurate index for the whole structure.

t A major portion of the concrete is located in the' substructure, l i

HoLMis a N ARVER INC j l

i t_ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -. _ __ _ _ _ . . _ _ _ _ .__ _ _ _ _ _ _

. )

where steel requirements and wall thicknesses are dictated by other considerations aside from ground motion. Here the cost penalty will be quite small. A similar condition, to a lesser degree, applies to the turbine generator substructure.

Accounting for these factors, it might be reasonable to Expect that the average increase in the cost of the reinforced concrete features of the facility would be less than 15%. If about two thirds of the cost of the structures and improvements and one third of the overhead con-struction costs, as given in the Company's Amendment No. 3 are chargeable to. reinforced concrete, it appears that the cost increase i 1

in these items might be in the range of $1,000,000 or about 7% of their combined total. l The percentage increase in engineering costs could be expected to  ;

1 be relatively greater due to the design refinements needed to replace the more casual practices usually used. As a completely unsupported estimate, this increase would probably be less than $500,000 (about 18%

of the $2,830,000 allotted to engineering, superintendence, and account-ing in the Company estimate). The combined total of $1,500,000 would be about 10% of the $14,500,000 allotted to structures and improvements

. and overhead construction costs. i

? -

P Again, it is emphasized that these numbers are not to be considered as anything more than an extremely rough estimate of low probable ac-curacy. The only fairly firm conclusion to be drawn is that the cost of the additional seismic protection is very likely to be in the range of 5 to 15% of the cost allotted to structures, improvements, and con-struction overhead.

a m_____-------_---------- - - . - - - - - - _ - - . - - _ - -

)

l I

2. Conclusions for Report i

In the course of the meetings in Bethesda on June 27 and 28, it was requested that the writer submit conclusions for incorporatiori in the final report.

= i Since the report is a joint effort of Dr. Newmark and the writer,.

the conclusions should be a joint effort. However, the writer has for-mulated the following tentative conclusions, which are purposely brief l l

and as non-technical as possible, but could be expanded as necessary.

I

"The following general conclusions' imply the absence of gross differentia 1' foundation movements such as that due to fault slippage g beneath the site.
a. It is feasible to design all features of the reactor which are essential to the prevention of uncontrolled off-site fission product release, to resist the effects of the maximum credible earthquake of the intensity defined herein.

I

[ b. Adequate design will require approaches utilizing the best earthquake engineering information currently avail-able, including methods which evaluate forces and motions

~

based on a suitable estimate of the properties of the earthquake and those of the structure being considered. l f Attainment of the required degree of seismic resistance I l

will also require careful attention to the details of all .

systems needed for containment of fission products."

3. Working Stress Basis vs Yield Stress Basis i

The following considerations relate to the use of conventional work-ing stresses in combination with spectra oflow intensity vs the use of  !

l d & m & w 4 M y l _ _ _ . - _ _ - - _ I

)

yield point stresses in combination with spectra of correspondingly high intensity.

The working stress basis is more conservative than the yield stress basis if the component is stressed significantly by other effects.

f in addition to ground motion, because the ground motion effects theri 2 have relatively less influence. As a simple example consider two components, one of which is unstressed except when seismic forces act. The other carries a stress from other causes of 10,000 psi.

With an allowable working stress of 20,000 psi under seismic condi-tions, the components would be designed as follows:

( No. I _No.2 Non-seismic stress 0 10,000 Seismic stress 20,000 10,000 Total 20,000 20,000 If the yield stress in each component is 40,000 psi, a doubling of the 1

l ground motion spectral values would give the following stresses:

l l

No. 1 No.2 1

. Non-seismic stress 0 10,000 l Seismic stress 40,000 20,000 Total 40,000 30,000 Component No. 2 is less highly stressed than component No.1 and will not be subjected to yield stresses in this particular case until the ground motion spectral values are tripled.

On the other hand, in certain other cases, it is necessary to con-sider conditions at yield in order to be sure of satisfactory performance

1

)

{

l

' i

]

under ground motion input represented by spectra having ordinates .

i double those of the so-called maximum probable earthquake. i One such case involves structural steel components, whe e the ratio of yield stress to working stress is typically more like .1. 6 I

Instead of the value of 2 typical for reinforcing steel. Pr esumably, a similar condition might apply to piping and equipment components;-

however, in these instances the use of higher spectral values mentioned in Item 4 might overcome this problem.

A second case involves factors of safety against overturning. For 1

example, c'onsider a case wherein the seismic overturning moment )

1 is 100,000 ft. kips under working stress conditions, and twice this' '

L value under yield conditions, with an available righting moment (due l

to dead load for instance) of 50,000 ft. kips. It is further assumed i that the customary procedure of providing an overturning resistance of 1.5 times the net overturning moment is adhered to. Then parallel calculations for a design based on working. stress conditions would lead to the following results:

j At Working At Yield I r Stress Level Stress Level r j.

l l Seismic overturning moment 100,000 ft k 200,000 ft k I Righting moment 70,000 70,000 Net overturning moment 30,000 130,000 Resisting moment provided 45,000 90,000 Reserve overturning moment 15,000 - 40,000 In this case the resisting moment provided is' insufficient (structure is unstable) if the seismic overturning moment is doubled.

1 4

e [,

s

)

e A related case applies to spread footings bearing directly on soils.

It is not uncommon to find that under the working stress conditions the ,

influence of the compressive stress in the soil due to verticalloads l ey.ceeds the overturning effect associated with seismic forces, leaving a net compression under the entire footing. However, if a doubling, of the seismic forces causes uplifting over a portion of the base, with the soil being incapable of transmitting tension, the soil stresses can be more than doubled.

Finally, there is the problem of seismically induced waves in fluid container s ;("aloshing"). Here, doubling the seismic input will more l- than double the computed wave height in the fundamental " sloshing"

{ mode. If it is important to prevent spillage of the contents under these conditions, as it might be where the fluid is radioactive, then freeboard 1 i

l j requirements should be based on the doubled seismic input.

l 1

7 , The numbers cited in the examples are not necessarily typical but '

I are chosen merely to illustrate the points invoived. However, it seems 4 1

l apparent, in such instances as those cited and certainly in others, that I the investigation of conditions at yielding may be necessary to eliminate weak links in the system and provide a more nearly uniform safety 1

factor, regardless of whether yielding is supposed to occur at double l

the seismic input, or at some other value. Hence, it is urged that the l AEC take the position that maximum credible earthquake should not cause non-linear behavior in the usual structural sense, nor unacceptable I amounts of sloshing, regardless of whether the Company chooses to design at a working stress level or on some other basis. In the writer's opinion, the advantages of designing on a yield point basir outweigh the disadvantages, in this installation except, possibly, in the case of equipment items.

4. Spectra '

From the results of the Argonne meeting it seems possible that two 6-

?

M E E__.____..____.__.___.____.._._____ - _ _ _ _ ~ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ . _ _ . _ _ _ _ _ . J

)

l 2 spectra will be required in the design; one to be used for buildings and t 1 building components, and the other to be used for. equipment. There l are distinct practical disadvantages to this approach: .

a. There is a hiigh probability of confusion and misinterpretation.

This has already been demonstrated by the variance between . 3 the criteria presented in the PHSR and that which was -

{ intended by Dr. Housner.

1

b. Further confusion arises in attempting to define the inter-

. faces between equipment and basic structure. l L

c. The use of two spectra will certainly not ea'se the difficulties .

to be anticipated in the public hearing.

It is strongly suggested that a single spectra be used and that re-duced stresses be applied in the case of those components which are

considered to require a higher margin of safety. For example, if l spectra corresponding 'to 0.5g are considered appropriate for certain structures, whereas others would require 0. 67g spectra, the allow-able stresses for the latter could be reduced in the ratio 0.50/0.67 .

or 0.75. This would accomplish a result at least as conservative as I that obtained by using two spectra.

l As a result of the Argonne meeting it was agreed that a design equivalent to using 50%g spectra with yield point stresses would be l

considered to be appropriate for the basic structure, this value being l considerably lower than that previously considered to be appropriate.

Dr. Housner's analysis entitled " Acceleration Adjacent to a Fault Produced by the Slip Displacement" appears to substantiate the 50%g figure fairly convincingly. The primary factors in this analysis which might change the picture appear to be variation in the values of gamma i

l

..>,>. . ~

5 and c. If upper bound values larger than those used are a real pos-sibility, then such increases could cause. ground velocities gre.ater than the 2 feet per second given in the analysis, with a corres'p.onding-ly greater damage potential. The writer cannot comment as to the

.. likelihood of this possibility. __

5. Revised Criteria Assuming that revised criteria will be submitted by the Company, such criteria should preferably include information relating to items .

such as those discussed in the following paragraphs.

a. The Company appears to regard the stack as a Class 1 7

structure. If this is considered to be a valid assumption, 4

i ,

it would be well to ask the Company to specify criteria to be used in design of the stack, particularly with regard to such items as the extent to which higher mode responses are to be considered, construction material, whether the stack is to be on ground or carried on a supporting frame, types of lining being contemplate'd, and percentages of ,

j critical damping to be used.

It is not clear whether there will be any elevated tanks in

f. the Class I category, but if this turns out to be the case, t then certain simple but important precautions are needed .

in design to maximize the available seismic resistance to the fullest possible extent. In the case of an elevated tank of the conventional rod-braced type, it is important to be certain that yielding of the rods in critical panels can occur without yielding or collapse of other features. This situation is not necessarily achieved in earthquake resistant tanks designed on a conventional working stress basis and Fenerally u., - ,

I~ _. . . . _ _ _ . _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _

. 1 requires considering conditions at impending failure of the weakest element. The effect on the riser of the large j lateral displacements associated with yielding of the' rods  !

in this type of tank.also needs consideration.

b. The Company should state the criteria to be used fo'f con-sidering the vertical components of ground motion. While i the vertical accelerations are less than the horizontal ac-celebrations, they should not be overlooked.
c. The Company should indicate the type of analysis (computer l or manual) to be used for piping systems in thermal stress analysis and also for seismic analysis. The amount of
damping to be assumed should also be stated. l
d. It is the writer's understanding that the spectra for periods below about 0.2 see or less will be based on an extrapolation.

Since many of the reactor structures and components have periods in this' range and since the spectrum ordinates change

, rapidly with period in this range, this extrapolation may be

subject to considerable error. It is suggested that the

}

Company investigate the feasibility of exter. ding these curves t

) ,

on the basis of additional numerical data auch as that obtain-able from the response spectrum analyzer at Caltech.

l e. It would be desirable to request information from the Company relative to the criteria they propose for evaluating seismic effects due to surges in fluid containers, including the amount l of damping to be considered in connection with the fluid os cillations.

1 i

4

_9 9 m .un.ie e. 49 .

L,_,________________-__ _ _ _ _ - - - -

.' c ~; .

q} )

f. If it is felt that documentation of the seismic resistance of cach critical component should be submitted, then the ,

Company should be appraised of this fact. Perhaps this documentation requirement could be handled as an inter- ,

}

pretation of the statements made on page V-8 of the PHSR, relating to design factors for Class 1 structures.

Ii l

l, l

l I

l 1

i 8!

i

)

3 .-

i 4

l l

l l

l

, July 16,1963

. . _ . - _ . , 4

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

1 t> ~- e n r 4

4*

. . qw . - . .:. -~w r , -w- ~

_.. . , _ ~ . . .m

...+ df;0% M w." j.-. w -ss#g - . . L .-2:wipp6M'%. '

.g,h.-wr.t$h:2.m.

. 9mn$ - 4 '.' * *

. L 9 .. . .

g .

_ .__._ _ .n...e.%. . ~ '....-,....;..... . . . . . . .. . _ . . _ . _ _ . _ ,

oArt or pecumeurs oarc ncesiveD No.: ,i f'>nowz_-

i M

e

~ '

bN . c o, .),oR,, hn$,(

x (& enel)

R. A. hi liwapa .

TOz OksG.: CCs _

OT H Eks .

Eryan x -

ACTION NECES$ARY Q CONCURRENCE O DATE ANSWERED:

e Y, NO ACTION NECESSARY Q COMMENT O

-~

CLAGSIF s POsi OFFICE FILE CODE:

U 50-205 (swp1 only)

REG. Nor DL &CNIP flON: (Must Be Unclassefaed) nEFERRED TO DATE RECElvED BY DAtt Ltr trar.s the fu1 Lowing natarial relab- ~

izy to Ocdega imys g, , g t/ sus. file sy.

ENCLOSUhlBs .

Docunsat, relating te gecations rainsd h sesetine at. batnerda and othen stern frca tre Arscnna su ti ns

[ '

of ? -b3-oost of seismic pvte m en, ..

etc. . .

"'"^"' Lv %(tj l g

M {* '

~

& (t) eA < x _. D ) ~

' Sk,% v.r$"...&,,, 6 dCU -r .

  • .....m......,,... . - . =

c.s. ATOMC .:htRGY COMESSION MAIL CONTROL FORM romu uc.sses (s.co)

=,-c.. 2.cc-- ----  ; . x .

~~

7 i 4 7./.W .s

. [c..: w.;c  :. g'.

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

c y * ; t..%;.> ,,p::.3 %;;4 ng.s.-9w".p&gvc.;;am . _ m ,- _ . _

(s-_so) ' '-l.

.  ; m- v. . - .. .. ; ._ .g.u.~ . ,.gH .. ;...

"2.

g,-_.a.cg . .yg yg.3.,.j.grg.g- =

n .

m ..

~

' . . . .*[ ' ? .M

  • w'*,, '

%Y T' e *'

,R , X*"a.ne.a.ss"$&&WM A ? *****'******'* ^- 4e'y-[***F..

. y =N , ^ * * .4 '~7 i1 s .' . ( ,, ,. j

~.00, ,,. $_ .fh, ',,,,O~. ., '.c',' ' ". ~AR ^'

? Q.h 4 y. ~ . * '

,K- *k .$5 l ....

. .y. , n2 g w ,; ....,..g, w

.y.g-;. .}.gc_% e

~ Q ; . i('; a :_ _ - - .

.qY: *

~

- .L i ^ y = , .~.y,, -( -h

, ---. y - y ..f yn .- . ,q., .,--z.--;.. m. 4 . , .. _ 7

_ ,. . . . . . :. .~,. , .

..,.. .- - a -. 2";*.* ..

.,.'u. ::,. 3 _ .. - . -;

.n.

. ,. s -

t . .

_ .gp, 'M .-

4 g

  • e'= wp

.g .

    • ,,, O + - "+,

1 1

1

  • . j

. j I

l  !

nh -

.I..; .

OIp ph.8 m_

Y 4

1 I

l J

g

- _.____ .________ _.._ _._ _ _ _ _ _

Retrieved from "https://kanterella.com/w/index.php?title=ML20234D855&oldid=3194114"