ML19345B914

Forwards Review of NRC Seismic Study of Nfs Neutralized High Level Waste Tank.Rept Generally Well Done But Does Not Address Highly Significant Factors.Rept Does Not Give Reassurance Re Health & Safety of Public

ML19345B914
Person / Time
Site:	West Valley Demonstration Project
Issue date:	09/21/1980
From:	Resnikoff M Sierra Club
To:	Haughney C NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
References
17439, NUDOCS 8012020727
Download: ML19345B914 (9)

Text

f h. 7O A,., S~ Ad/

f E2 # " 3164 Main Street'b Buffa }Z.16) 832 9100 9

s u , og b $7 t' DmD n 2 .-

g3\

y<

d September 21, 1980

\.s -

3 g 9  :.5 cc ' " 20 > y,

'

ghney M NMss eprocessing and Recycle BNnch'at s!cucfi R --

M

,

.

'

Division of Fuel Cycle and Mate $i$1, $$fsYy

~

U N 0.S. Nuclear Regulatsry Commissio'tg y & g

-

'{j 1: " 24 Washington, D.C. 20555 Re: Seismic competence of neutrd)ized 3 '@

high level waste facility a6 West 'gliey S

Dear Charlie:

We have now reviewed five reports by the Nuclear Regulatory Comission of the seismic competence of the nuclear facilities at West Valley. The enc-losed report pertains to the neutralized high level waste facility and the NRC-connissioned study," Seismic Analysis of the High Level Neutralized Liquid Waste Tanks at the Western New York State Nuclear Service Center, West Valley, New York", UCRL-52485. In general, we find the analysis well done, but we be-lieve that too many highly significant factors were ignored. We are not re-assured that the hea'th and safety of the public is protected. These factors are discussed in the enclosed report.

There are three aspects that require closer examination that we would like to discuss in this letter.

a) It is clear that much of the discussion of the seismic competence of the HLW tanks is now moot because of the Federal legislation that authorizes DOE to remove and solidify the HLW. It doesn't make sense for the NRC to expend further dollars on such theoretical studies to correct any deficiencies. How-ever, we do believe that further analysis regarding the reprocessing plant and spent fuel pool are appropriate. In particular, we consider it important to provide an analysis of the spent fuel pool building; your studies .only relate to the spent fuel pool itself.

b) In analyzing the spent fuel pool building, we think that there is need for a discussion of the basis for the NRC assumption regarding a surface acc-eleration of 0.2 g. The regulations,10 CFR 100, /,pp. A appear to allow a lat-itude, and that the Staff assumption may not be conservative. According to he regulations, the 1929 Attica earthquake of 0.2 g should be placed at the hest Valley site. Accordingly, Lawrence Livemore Laboratories assumed an earthquake of 0.15 g at the underlying rock at the site, to produce an earth-quake of 0.2 g at the surface. Since the Attice earthquake had an accelerat-ion of 0.2 g at rock outcroppings, it is also consistent with the regulations to assume 0.2'g at the underlying rock at the site, and a greater number at the surface.

c) We were quite surprised to note that the HLW vault side walls are cracked

-

~d;..7" P &.r! 9i @: ..

sierra club radioactive waste carnpaign g 17433 8012020_727  ;

-. . . . . .

• C. Haughney , ,

'

pag 2 two

.

.

due to the thermal gradient, the difference in temperature between the inside and outside of the vault wall. The extent of the' cracking of the vault walls can be determined by Rockwell Hanford with a remote TV camera. We believe it is important to know thc extent of water seepage into the vault, if any.

Several of our concerns about the LLL study might be answered by supply-ing Appendices IIIA through E of the above mentioned study. As far as we can tell, all of the studies released by the NRC contained copies of Appendices I, II, and IIIF, but not IIIA through E. We would appreciate it if.you could supply these.

i.e await your response to the above points. Thank you for your atten-tion to these matters. If any points raised in our review are unclear, don't hesitate to call.

cc: C. :longerson J. Larocca Best,

u. Pyles S. Fink g

• g/ jf R. Vaughn S. Lundine v J. Dingell R. Ottinger

/* arvin fMN/f/k

Resnikoff M. Udall

!.

..

-

l

.

.

.

1

_. _

- - . _ _ __ . . . _ . - .

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

~

~

" ~ ' '

, ,, , wRC S21s"1c Study;

.

.

Page two , , .

.

.

'

.. . .

p' STRUCTURE AND CONTENTS OF THE NEUTRALIZED -

HIGH LEVEL WASTE -TANK AND VAULT  !

A brief review of the structure, history and location of the high level waste tanks is necessary in order to understand the difficulty of accurately developing a model' for seismic analysis. The neutralized waste tanks and con-crete vaults are shown in Figures 1 and 2. The tanks consist of a working tank .

(80-2) and.a spare tank (80-1), both 70 feet in diameter and 27 feet in height

• with a volume of 750,000 gallons. Tank 80-2 contains approximately 560,000 gal-c lons of highly toxic liquid. radioactive waste. The tank-is constructed of car- ,

bon steel plate with a wall thickness of h inch for the sides and 5/6 inch for i the bottom. The tanks sit on Perlite concrete blocks within a carbon steel pan  !

3/8 inch thick. The pan, in turn, sits on pea gravel within a cylindrical re-inforced concrete vault. Each 78 foot diameter vault sits on a concrete pad which rests on a gravel layer. The vaults are 8 feet below the earth's surface.

Each vault has a wall thickness of ih feet and a roof thickness of 2 feet.

For years the government has claimed that the vault sits within rela-tively impermeable silty till soil, a dense clay with occasional lenses of gray-el and sand. However, only one boring was done in the vicinity of the waste

tanks so the silty till theory is based on limited evidence. Furthermore, the boring did not reach bedrock hence the depth to bedrock is unknown. It.is also

'

not known whether the borehole, which could provide a pathway for the migration of radioactive material should the tank leak, was ever plugged. The location and extent of the underground sand:and gravel lenses is also unknown.

The cement slab holding the concrete vaults sits on a gravel Led.

! Water is injected via an 8 inch standpipe in order to keep the clay moist and im--

pemeable. If the clay dried, it would crack and allow water to floa more eas-

ily.

,

The carbon steel tanks are not finnly attached to the concrete vault.

The tanks rest on Perlite blocks and do not slide because of the friction between the blocks and the waste tanks. The purpose of the Perlite blocks between the -

.

saucer and the tank was to insulate the concrete vault from the heat when the j tanks were stress-relieved in the field. In order to stress-relieve the tank, temperatures within the tank reached 10000F, a temperature which would have de-graded the concrete.

l As can be seen in Figure 2, the tank, vault structure is complicated.

by the fact that the tank is not simply' shaped like a flat pancake, but has 6 columns through the interior of the tank. Concrete columns extend from the floor to the roof of the concrete vault. The tank has six 48 inch steel diameter col-

,

umns through which the vault columns pass from base to ceiling. The tank also has other internal columns which support the roof of the tank. In~ addition, var-ious appurtances are attached to the tank and vault such as air spargers or bub-

blers and ventilatton equipment.  ;

iiETHODS AND RESULTS OF ANALYSES -

The method of analysis is simple to describe, but quite complicated to model in detail. dssentially, the weight of the vault, pan, tank and liquid

-

waste and surrounding soil (including the 8 feet of overburden) and certain ef-

fects of the flotation incident (not including the cracks in the vault), were ad-ged to the seismic. loads to calculate the stress components. Thermal loads and

-

l

. .

6 r s ,e . .

v-- - ,- - -i, + s , , * -

~,-,,e= .. g .ww-=.,--

- . .

.

'*. ,

,

-

.

~ 3164 Main Street Butf ato, New York 14214 (716) 832 9100 ,

.

REVIEW 0F THE NUCLEAR REGULATORY COMMISSION SEISMIC STU OF THE NFS NEUTRALIZED HIGli LEVEL WASTE TANK

.

INTRODUCTION In a series of reports 1-5, the Nuclear Regulatory Commission (NRC) Thesehas analyzed the seismic competence of the nuclear facilities at ween May,1978 and March,1979, evaluated the acid and neutralized high level liquid waste tank facilities, the reprocessing building and the spent fuel sto-rage pool. These analyses show that all structures would site.

sustain Most damage under importantly, the stu-the maximum earthquake which could occur on tha dies demonstrate that the concrete vault surrounding the neutralized waste This tank extreme-must be already cracked due to the heat of the high level wastes.

ly disturbing fact has never been published in the media or publically admitted to by policy makers.

In this report, the Sierra Club Radioactive Waste Campaigrexamines the study of the neutralized high level waste tank. The NRC concludes that the vault would crack slightly under an earthquake without damaging the high level

~

waste tank. We find that this re-assuring conclusion is based on an analysis

!

Specifically, the NRC has ignored that has ignored highly significant factors.

the cracking of the vault due to the heat of the liquid wastes and due to the The NRC has also ignored the possible rocking and flotation incident of 1965.

sliding of the tank within the vault due, in part, to the sloshing liquid with-Other factors not given sufficient weight by the NRC in their seis-in tae taic tank.include possible impingement of the tank on the roof supporting in-alysis

-

ternal columns, the effect on the tank of the movement of pipes and ventilation system components, the possibly damaged condition of the Perlite blocks, and the uncertainties regarding tank stresses resulting from the flotation incident of

,

1965.. We conclude that the damage effects calculated by the NRC are seriously estimated. It is not clear that the tank as well as the vault would main-Enderain integrity during an earthquake. Additionally, as each year passes, Tne tank's tne ability l

carbon steel tank continues to corrode and therefore weaken.

to withstand an earthquake can only continue to decline further from its present uncertain state.

What is The waste tank contains a large amount of toxic materials.

While a hazardous dose of strontium-90 is on the ord-meant by a "large amount"7er of one millionth of a curie, the waste tanke contain S r-30. A leak of high level liquid waste into the Cattaraugus Creek watershed

~

could contaminate the water supply of millions of persons in New York and Canada, l because of the possible consequences, we support strongly the move to solidify i the high level liquid waste with all due speed.

,

sierra club  !

radioactive waste campaign .

. . ,. .

'

MRC S;ismic Study

-

page three ,

,

.

and liquid-tank interactions (sloshing liquid) were neglected. If certain maxi-zum stress were exceeded, the tank or vault was deemed to fail. We believe the analysis was competently performed by the contractors Lawrence Livermore Labora-tories, but the uncertainties, described in the next section, do not allow a def-initive conclusion to be made.

r,ertain aspects of the analysis, such as the dead weights of the struct-ures, are quite simple to ascertain. Inputting the seismic load, on the other hand, was not straight-forward. The stiffness properties of the soil, as well as the underlying soil and sand composition, and rock structure, are modelled.

An earthquake is assumed for underlying rock structure and transmitted through the soil colurn to the vault. The earthquake is not a simple shaka. but a ser-

- ies of vibrations in the horizontal and vertical direction extending over sev-eral seconds. One then sums the loads, dead plus seismic, and determines whe-sher the stresses exceed limiting criteria for the vault and the tank. There is some uncertainty regarding the limiting criteria. For tne tank, an empiri-cal formula from test data for cylindrical objects subjected to axial compression, was used. On this basis, the limiting buckling stress for the tank wall was es-Limated at 4,320 pounds per square inch (psi).

The results show that the reinforced concrete vault walls will exceed the limiting criteria with a maximum peak acceleration in the range of 0.13 g to 0.i69 and that the center of the base slab surpasses the limits at 0.189 Con-crete will crack on the inside and the outside of the vault at accelerations less than 0.2 g. On the other hand, analysis shows that the thin-walled tank will

withstand a maximum earthquake without leakage and that, because of friction, the

'

tank will not slide on the Perlite blocks. Thus, cased on the assumptions made above and ignoring the uncertainties we discuss in the next section, the NRC does -

not expect the tank to leak and the highly toxic contents of the HLW facility to enter the vault under the maximum earthquake which could occur on the site.

UNCERTAINTIES IN THE NRC MODEL .

1. Cracked Vault due to thermal loading. As was pointed out in the introduction, it is not well-known by the public-that the concrete vault is al-ready cracked due to the thermal gradient, the difference in temperature between and outside the vault. The thermal gradient is the primary stress on the inside vault. The JRC assumes the high level liquid wastes are at 1800F and the temp-erature outside the vault is 500F. This 1300F difference is sufficient to cause the vault to crack. Between 1968 and 1972, dFS operated a heat exchanger within the tank to boil off excess liquid because the low level waste evaporator in the reprocessing plant did not operate properly. At that' time, the temperature with-in the tank was not 1800F, but 2120F. ..

According to the NRC, the vault has formed cracks on the exterior sur-face due to this thermal gradient. " Based on experience with similar reinforced concrete structures, we assumed the stress relief resulted from creep and for-mation of cracks in the concrete." It is important to determine if the cracks are just on the exterior or extend through the vault. When the NRC observes the interior of the vault with a TV camera, it can be determined whether water is leaking inward due to this cracking.

.

..

.. g .:-

.

em e .=.e- .e. . - -

-*=M e * *y u___- _ _ _ _ _ _ _ _ _ _ _ _ _ . -

.-

- - -

- - . .;RC Seisac stuoy ,-

, .page four

^

.

.

These cracks, which undoubtedly have weakened the concrete vault, were Obvious-not taken into account in assessing the stresses due to an earthquake.

ly, a cracked and weakened vault will undergo more damage during an earthquake than an uncracked and solid vault. We believe this is a major defect which in-validates the NRC analysis.

It is important to add that this cracking due to the thennal gradients occurred because the HLW facility was poorly designed. Had the inside of the vault been ventilated, a relatively inexpensive modificationThis to cool the vault, is a design de-no cracking due to a thermal gradient would have occurred.

fect.

2. Flotation Incident. During construction, afteJ the tank and vault had been cercleted, but before the excavation was backfilled, The vaults,water filled the with tanks and construction pit, apparently to a depth of 30 feet.

saucers within, weighing 2850 tons each, floated 3 to 4 feet off the concrete pad.

The cause of this incident is mysterious. Perhaps a water pipe broke, flooding the pit. It is also unknown to us the effect of this incident on the pipes and other auxiliary appurtenances to the waste facility. While

~

the vaults were float-When the-water in the ex-ing off the concrete pad, mud washed under the vaults.cavation pit was p top and bottoms of the vault cracked. With the present liquids in the two tanke, over 9,000 tons of concrete, liquid and metal are sitting on the damaged conc.ete Since water is flooding the exterior of the vaurcs, pad, only 4 inches thick.this weight is somewhat counterbalanced by a buoyant forc a the single slab for both vaults.

2 The HLW facility was repaired by grouting under pressure the area at the base of the slab. Then the silty till was washed from between the vault and The vaults contir.ue the base slab and cement was introduced to fill the void. Wheth-to rest at an angle; in the case of the alternate tank, the angle isWhether 10 all er the repaired structure is as strong as the original is unknown. -

cracks were filled, whether cement formed unevenly under the slab, and The GA0 what suggest-stresses presently aist on the slab and the vault is unknown.

ed in 19'1 that because of the flotation incident, the integrity of the tank was in quest A and needed to be evaluated. This evaluation has not yet been con-ducted.

The NRC has attempted to model the situation and bound the uncertainties by ring.

assuming, in the worst case situation, that noThegrout reached NRC states,

.

This does not take into account the effect of the vault cracking.

"Had this situation actually occurred (of grout not reaching the bottom girder),

The full ef-the capacity of the vault would have been significantly reduced."

fect of this worst case situation, and its consequences, are not discussed in the NRC report.

It is important to remind the reader that the pan under the tank has It has been reported to us by former a defect, the cause of which is unknown.

workers that the pan was tested during construction and did hold water at the It is quite The defect has occurred during the intervening 15 years.

time.

possible that the flotation incident and/or thermal gradients, has caused the i

l vault to crack further, placing stresses on the pan and leading If thistohypothesis stress corro-sion and the pan defect. The pan is only 3/8 inch thick.

l i

is correct, the pan may be placing additional stresses on the tank and this may l lead to stress corrosion of the tank.

.

• • 5 page, five

. .

,

.

3. Shifting of tank during earthquake. Shifting of the tank would cause the tank to strike one of the six internal concrete columns, possibly causing a tank rupture.. The tank is only held in place within the vault by .

friction between it and the Perlite blocks. In App.III of the NRC study, the NRC has measured the coefficient of static friction and detemined that the tank would not move. However, the NRC analysis has neglected the rocking of the tank. Since the great balk of weight of the tank is the contained liquid, not the tank itself,,it is quite possible that in the event of an earthquake, the sloshing liquid and shaking vault, could together cause the tank to rock. Rock-ing would introduce a large stress on the bottom rim of tne tank. Further, if the tank tilted a mere 1.50, it would impinge the internal concrete columns and perhaps rupture the tank.

4. Seismi'c Effect on tank apourtenances. The tank is not entirely free-standing within the concrete vault. Piping connects the reprocessing plant, the ventilation system and air sparsers to the tank. Piping connects the underly-ing saucer back to the tank and the exterior. The seismic effect on the integ--

rity of the tank must take into effect these appurtenances. This was not done in the NRC analysis.

6. .iaximum earthquake. The NRC has instructed Lawrence Livemore Labor-dtories to use an earthquake of 0.2 g horizontal acceleration at the surface of the West Valley site. The underlying rock outcropping is assumed to accelerate at 0.15 g to produce a surface acceleration of 0.2 g. This is not a conserva-kbe ca,assu m

ption according to theofregulations. Because the West Valley site, and new York, the epicenter a large earthquake, do not lie at the edge of a tectonic plate, the regulations require that the largest earthquake in the

.

region be placed at the site for the purpose of design analysis. The Clarendon-Linden fault is assumed here to be a capable fault. This would require an earth-quake of 0.2 g at any rock outcropping, and therefore greater than 0.2g at the level of the tank. This would make the maximum acceleration greater than ass-umed by the NRC.

.

6. Unpublished NRC studies. The NRC, in a table of contents for '

Appendix III, contracted with Lawrence Livemore Laboratories, for important studies concerning the HLW tank. However, all but one of the six studies is missing from the final report. The missing studies, report numbers and comple-tion dates are listed below:

A. " Review of Liquid-Tabk Interaction Analysis Techniques", UCRL-13835, Dec. ,1977.

B. "is Study of Time Related Properties of Concrete", UCRL-13836, Dec. ,1977.

C. " Environmental Effects on Metal Structures and Connections in NFS Waste Tanks",

UCRL-13837, Dec. ,1977. .

ii. " Plan for Dynamic Testing of NFS Tank and Vault", UCRL-13838, Dec. ,1977.

E. " Evaluation of the Integrity of Existing NFS Waste Tank", UCRL-13839, Gec.,

1977.

F. " Test Report r.oefficient of Friction Between Carbon Steel and Perlite Ccn-crete Surfaci ', UCRL-13840, Dec. ,1977. (completed).

These studies are very important for a complete understanding of the l tank's seismic competence. E.g. , App. IIIA details the known literature on the effect of sloshing liquid on tank dynamics during an earthquake. This would par-tially answer the question of'whether the tank would rock during an earthquake.

The NRC should make these studies publicly available.

l .

'

i

,

. .

'

' " ' '

NRC' Scicmic Study

-

Pags si.r- .

.

.

.

REFERENCES

1. A.M. Davito, R.C. Murray, T.A. Nelson and D.L, Bernreuter,.

.

" Seismic Analysis of High Level Neutralized Liquid Waste Tanks at the Western New York State Nuclear Service Center, West Valley, New ' York", May, 1978, UCRL-52485, Lawrence Livermore Laboratory,

2. C.Y. Liaw, A.M. Davito and R.C.'Murray, " Seismic Analysis of the Acid Liquid Waste TanksLat the Western New York State Nuclear Service Center, West Valley, New York", March, 1979, UCRL-52600, Lawrence Livermore Laboratory.

R.G. Dong and'S.M. Ma, " Structural Analyses of the Fuel Re-

'

3.

ceiving Station Pool at the -Nuclear Fuel Service Reprocessing Plant, West Valley, New York", May 5,1979, UCRL-52575, Law-rence Livermore Laboratory.

4 R.C, Murray, T.A. Nelson and A.M. Davito, " Seismic Analysis of the Nuclear Fuel Service Reprocessing Plant at West Valley, New Ycrk", May 24,1977, .UCRL-52266, Laurence Livermore Lab-oratory.

5. Review letter from Nathan M. Newmark, Consulting Engineering Services to A.T. Clark, Nuclear Regulatory' Commission, dated January 4,1978, under contract AT(49-24).-0116, NMSS

.

.

e

.

.

4 s

.

.

e D

- , _ _ .,

, . . . . . . . . , .

= . . "* . pag

.

+ , , 2 SGV;n

'

.

S in..d.a stando.pe' S er.ft faae ievee .tdh Q 19 stsawlsrae so th leved indcairl

,

• I . T 8 ft ri.n. ag, , ,

"i

.

Stry tiay tuskfde ,

1. 3tv c?'T IWckf=3 3

' ~

c ~'- '~ \ ..$ < ~~ . e -

ju..- .. .a g. m.ne. - . .=.: g..

..

<

.

.

t9 einforud sat. and ya e4 3' concette waisit 5,tjC Cabase steet tank

&lty stay strata I

"

fI #***"'#**

,

J.* / **. .

-Peet.te tdorks JQ.,_ .I . F.'t at .- .-.-... >...

. 1.'< p '" v ***

- --c,a.*

6 6 6 Watse ing.cten heads Pig.f. Sectices view of neutralisert htg't-level waste facility.-earbon steel tant. Pen, and reinforced concrete vestt.

.

- . . - . - . . . . . , ,

.

..

.

.

'

• . .

. .

.

- <

.

'

70 ft t.D. -

,

• A V [y,,,, -

// /// ///// /////

/

' W

,/ '

$

1. T -

/  ? / f ant s

1. -

• 2 f t. 9 is h, /

h h ctes

['

,

'

/ # 48 in. 0.0. pipe

/ / /

/ . / 30 iri. 0.0. col. / /

27 ft -

/ /

/ . / ,

.

/ / / /

/ / / /

-

/C" ,/.. ,/, /~Fe*

( 3 in. f 2 W crew

.e :., Jc r2

,, A

- ... e .:.a -. ., c - - =,, ,e .n.

1$ in.

/ / ,

.j= . <, .

.

,n. s .. .v .y

"-

. . , . . . . . . .

- , _

.

/////O / M3/f t. K4 en.n0.0.;,,(f / D / /

'

v ,

.n.

Fig.2. 'Section view of the high level neutralized waste tank

. and vault detailing the vault colurms.

l

,

'

.

.a., e --

. *

• l-