ML19208B490
| ML19208B490 | |
| Person / Time | |
|---|---|
| Site: | West Valley Demonstration Project |
| Issue date: | 06/22/1979 |
| From: | Coffman W COFFMAN, W.A. |
| To: | |
| Shared Package | |
| ML19208B487 | List: |
| References | |
| NUDOCS 7909200432 | |
| Download: ML19208B490 (17) | |
Text
-.-Estimates of the Scope and Resources needed to perfor:
Seismic Analyses of Facilities at the Western New York State Nuclear Service Center June 22, 1979 b7'Jayne A. Coff=an 4405 Bel Pre Pad Rockville, IG 2C353 , ..,i,l.-9200-
...!I
SUMMARY
A series of analyses pertaining to facilities at the Western New York State Nuclear Service Center have been considered. Specifi-cally, the tasks are in the two =ajor areas of high level neutralized liquid vaste storage tanks (HLW) and the fuel receiving and storage
~facility (FRS).
Available reports concerning previous analyses by LLL and LASL vere reviewed to deter =ine the scope and methodology of concluded seismic analyses for related facilities at this site. Some seventy Cravings were reviewed of the design of these facilities to gain familiarity with the physical characteristics of the structures and their surrou-ding soil.
A narrative is presented in the subsequent sections of this report for each task. This includes a discussion of the approach which is expected to be most appropriate and productive. Cost elements it.cludi=g structural engineering effort, computer time and travel are esti=ated on a task basis.
A su-... q of the estimated resources required to perform the var.aus tasks follows. The estimates tend not to contain conserva-tism or contingency for inefficiency due to delays, etc.
1:s ~j!s30
-Problem Method Analysis Comp Travel Time Time (days)($)($)Neutralized vaste tarJc-vault concrete spalling Lnd calculations 10 0 0 tank rupture toughness vs. particle shape, size, co=pressive
'strength Neutralized waste tank:
piping Modal analysis 25 500 113 Utility room equipment Modal analysis 20 350 225 FRS rack Time history 40 3,500 0 analysis FRS vall rework
- 1) base fix on hand 25 - 30*500 112 calculations and study of LLL re-sults (detailed)
- 2) reanalyze by best available ethod
.Totals 120 -$4,850$450 125 days* Depends on availability of LLL analyses
<} j, )~'2 ,t, ,
..II HIGH LETIL WASTE TANKS 1.Concrete Sealline LLL has suggested it is possible concrete chunks, presurably without any rebar, could break loose from the top of the vault during a seismic event and freefall and i= pact the top cf the
-tank. They expressed concern in reg d to the potential for rupture of the tank top.
The first activity is to esti= ate the size range of the con-crete chunks.
A second consideration is to dete:=ine the probability the concrete chunks, perhaps in ter=s of various size classes, would have of reaching the top of the tank without being significantly impeded by the network of girders which acts as a superstructure on the top of the tank and which in turn is supported by the six colunns.For the particles which can reach the tank and give up their kinetic energy about three size classes should be considered.
The large size vould have more mass and thus more total kinetic energy but may well have a less sharp i= pact surface than a s= aller concrete masa.
The drop distance is only 4.0 feet and therefere the specific kineticenergyofaconcretechunkis4.0(ft-lb/lb). A cethod drawing upon the Charpy and I:cd i= pact toughness tests methodology will be used to help assess the integrity of the tank top in t a of i*,s toughness and siribrity of the i= pact with other articles 3 c./b 34d'
._i for which inpact toughness data is available. Effort should be made to account for the possibility of concrete crunbling on i= pact vise tank penetration.
The freefall time is about 1/2 second and the impact velocity will be about 16.1 fps which is in the range between that for the Charpy (17 5 fps) and the I:od (115 fps) tests.
, Material property data for the concrete and steel vill be needed. It =ay turn out that the energy associated with a possible direct impact is sufficiently low so that penetration and rupture of the tank is entirely impossible. Alternatively, if the impact is in the range near where fracture is a feasible consequence, more effort vill be required to refine naterial representations.
Hand calculations should be =ade to investigate those cases expected to be representative of li=iting conditions. The effort required to perform this analysis is esti=ated to be ten days of a structural engineer's time. The task is not expected to involve any charges for computer time or travel.
s h 4
-2.Firine The neutralised high level vaste tanks have a network of underground piping which connects the two tanks together and to process equipment. The quantity of pipe is perhaps as much as 600 feet and =uch of it is of about 3.0 inch diameter. There is also other smaller instru=entation piping. In regard to vaste containmt, part of this piping =ay not need to be included in the analysis. A detailed review of the functional character of the instrument tubes should be made and ones with insignificant failure consequences excluded from the detailed seis=ic analysis procedure.
The LLL analysis (1) of the vault, tank and contents vas done in a way which treated the dy- H c effects of the soil surrounding the vault as though it were equivalent to a static pressure loading on the walls of the vault. The motivation for this was simplicity (and the authors stated that inclusion of soil dy"-ics was beyond the scope of their study). The justification was that the soil had a natural frequency (fundamental) which was =ach 1cuer than that for the steel tank structure.
The static pressure loading method is likely to be much scre applicable to the vault than to the pipe network because the stiffness of the pipe network is lower (because the lateral bending stiffness is relatively lov for a pipe) than that for the vault.
The consequence of this is that the pipe is expected to have several frequencies which are below the lowest vault frequency. Therefore, the pipe and soil are more likely to be dynamically interactive.
5 i. 3 9_hc ,, l g-.
.-This suggests that the dynamics of the soil probably can't be treated as being equivalent to a static pressure load using the soil pressure theories of Mononobe and Ckabe (2, 3).
For this arnlysis, it is suggested that the finite element model include:
1.pipe network in detail as 3D pipe elements
-2.u cc.ird49gsoil as 3D solid elements or equivalent springs and mass elenents 3 the vaults and their contents would be =cdeled as a s=all =u=ber of rigid casses, perhaps as few as three spring and cass pairs on each end of the pipe network for the liquid, the 'ank and the vault (for one particular degree of freedo=).
Item 3 above would allev inclusion of the cross influences of the vaults / tanks umn this p.ig;3, integrity inquiry vanout attenpt-ing to analyze subregions of the vault / tank system. Tre soil surrounding the vaults would be treated either as having no d. ramie effect (as LLL did) or as having a spring stiffness characteristic for its interaction with the surrounding soil. The =odel choice would depend on an independent estimate of frequency disparity (vault-soil).
Tne analysis should be rade using the response spectra method and a linear model for all of the ec ponents (and caterials).
If the modeling indicates the soil is highly nonlinear (uniarial tension and ce=pression) then a tine history solution vould be considered. It is not anticipated that this is likely to be needed.
6: L l,\ J Ultj
..- - -. _ _ _., 1 In piping systets the mechanical loads can be very sigaifi-cant. For e ple, undesirable and high loads can be applied to force misaligned runs of pipe into align =ent so connections or attachments to structural supports can be =ade.
Also, settling of fill can cause a net downward load on the buried pipe. Other loads result frem gravity, internal pressure and ther=al cycles associated with caterial transfer through the pipes.
ANSI s+nMW 31.7 applies to nuclear piping (for power plants) and can be used as a guide to suggest many elements of good practice in this analysis as to considerations such as connecting technique, inspection, load eccbinations and their
, stress linits.
It is estimated the seis=ic modal analysis can be acco=plished in 25 days. A site visit is not definitely required. An initial effort should include sketching and iscretric of the pipe network (with di=ensions) and explicit definition of pipe supports (location, type) and whether the pipes are in an open channel or one back filled with soil.
Total esti=ated resources are:
1.structural analysis 25 days 2.computer time
$500 3 travel$113 (.5 probability of trip te 'a'est Valley) 7.l !, 9 t ,; e<s-
.3.Ecuirment Auxiliary power generating plant equipment incl" rid ag diesel engine, generator controls and sete power conditioning equipment as well as other support systems are located in the utility Icom.
This equip =ent is intended to provide power in the case of corner-cial power outage for priority functions such as ventilation.
.The analysis of this equipment in regard to a seismic event vill be rather straightforward. The necessary activities include the develop =ent of a relatively si=ple but representative mass and stiffness model of the components and their method of attach-ment to the floor of the utility building. Tne rassive concrete pedestals vould also be included in the d W ~ie model. Tne engine and generator would typically have ' shock-counted' links in their load paths to ground. These seg=ents have lov stiffness and serve to attenuate the high frequency vibrations of the stiff eachine parts which result fres *- M h -ce of the high speed rotating nachine.
The = ass-stiffness medel can likely be accomplished with the use of some 40 = asses and 100 springs. Hand calculations vould be used to dete:--d e ' equivalent' spring constants and : ass values.
Also, in various regions heavy valled pipe elecents =ay be used to represent generator rotor, housing and perhaps the engine as well.Thesa elements senetimes allow a : ore straightforward representation of the epatially distributed cass and stiffness of a cceponent such as a generator and they can be easier to use.
Control boxes and batteries can be : deled ac peint nasses.
--B< , _,-
.A critical aspect of this control room equipment nedeling is the nature and strength of the attach =ents.
Details in regard to the equipment is not shown on the utility roca drawings. A visit to the plant will be necessary in conjunction with develepcent of the =odel for this task. The equipment foundations appear to be independent of that for the b"Wi g.
The buming analysis is
, not included herein.
-Nonlinearities are not expected to be i=portant in magnitude or nu=ber. The =odal analysis method vould be used to obtain the response to the seis=ic event per the Regulatory Guide 1.60 spec-toral shape.
It is esti ated that the resources needed for =aking this analysis are as follows:
1.structural 20 days 2.computer time
$350 3 trip to West Valley
$225 9-, ,/\,'
.%III FUEL RICEI7 DIG STATICN 1.Fuel Rack The fuel rack and canister syste= involves nonlinearities due to the sliding of the canister in the rack (slip-stick =otion under the applied force syste=). Since the canisters are not absolutely contrained in the N-S direction against sliding, a significant departure frc= linear elastic behavior is possible. The todal analysis =ethod is dependent upon the e:dstence of a co=pletely.
=atheratically linear (physically elastic) structure. The force and motion syste= associated with slip-stick behavior under the predo-4" ce of friction (no slip condition) is linear. There is no =otion until a threshold force is achieved and then =otion tends to continue without the develepnent of increased resistance.
The =cdal methed could be applied on a U'4 t analysis or try-and-see basis. This vould involve:
1.Assu=e the canisters are each bolted to the rack and do not =ove.
The modal analysis vould give the load on the bolt.
2.Asst =e the canisters have negligible resis ,ance to sliding.
The individual canisters can be assu=ed to be connected to the rack by weak springs in the II-S direction.
The submerged weight of the canister and its contents veuld be taken into account to assess the weight which when :::ltiplied by a friction coefficient for vet surfaces vould yield a sliding force.
This vould be done by a hand calculation.
If the force in the bolt of cass 1 above is greater than this sliding feree then analysis 10..#Il,q--
- 1 is an upper 1Mt in regard to tie down lead, canisters vill slide but we haven't determined hov =any vill slide out. The second analysis vould be used to provide information on the narf m ecmpression of canisters in the N-S direction. The first analysis would also give information on the lov frequency interaction in the E W direction in ter=s of m+m deflection and rack strength.
, The straightforward method of h- M
- g the fuel rack problem is to do a transient analysis with a time history forcing function.
It is anticipated that a single rack with 26 canisters would be modeled. Finite element =cdels are generally not co=prehensive of the physical constraint that two objects can't occupy the sa=e space. Nonlinear interface (gap) elements are used to si=ulate the E-W clearance between racks.
'F." 6-s These same general element types are used to si=ulate the canister's tendency to stick to the rack until a threshold sliding force is achieved. A very weak spring constant is used to establish a unique displacement at any given ti=e vise constant force notion (sliding).
This transient analysis would be ec:prehensive of:
1.rack tie down stresses 2.canisters sliding out of rack (scuth end) 3 canister interaction in 3-3 direction
' -L's40 11 nI ,J 4.interaction of rack with artificial fixed walls in LW direction to si=ulate interaction cf adjacent racks 5 strengtt of rack due to E-W =otion 6.overall rack strength Deflections (E-W) would be checked to see if they tend to open up enough space to allow a canister to ' drop through'.
The transient analysis is more likely to produce -e 4 nsful results. The =odal analysis would be both incoalete (the extent of fuel sliding out would not be known) and pese a risk in regard to providing sufficient infor=ation to the analyst so 'ht con-clusive results and effects can be s'ated.
The transient nanlysis would require 40 days of structural analysis, about $3,5C0 of computing time and no travel.
.12).
--!2.FRS Wall Da age
'Dong and & (UCRL-52575 5/5/78) say the results of their detailed amlysis indicate that valls F/N and C/S are most severely stressed (ref. 4). Also interior valls F/C, W/F and W/C vere highly stressed. They discredit the applicability of their ther=si stresses calculated for a 35 F. S 7 and elastic concrete on the basis of the4T not SPP 7 ng in the lover part of the pool, stress l1 relmtion opportunity during life and cracks providing stress relief. The authors didn't report any stress =agnitude data for the load cases such as ther=al, seis=ic, water and soil pressures.
Without the ther=al stress the vall F/N, which is the north vall of the pool, exceeds the allouable criteria in terms of vall moment at 0.16g seis=ic (sero period) spectrum excitation. The most severe location is at the top near the intersection with the W/F vall which partitions the pool and the water treat =ent cell.
Appreciation of the analysis is i=peded by the unavailability of responses to the independent load cases and by not reporting any stress values at all.
If the results were in terms of stress, results could readily be thought of in ter=s of =aterial strength. This is lost when only 12 values of =cment are reported.
If it is accepted that the task is to bring the Ng jH to a mini =um value of unity in vall F/N for every ele =ent then:
1.structural change need be =ade 2.evaluation of the stress =ust be rade to deter =ine whether the i= prove =ent is adec;; ate.
s , h,)Ib*0 iJ 13
-.-For a prismatic and isentropic beam ca.6ng a given =ccent the bending stress is inversely proportional to the square of the thickness. Also, the bending moment capability increases as the square of the thickness. Thus an increase of thickness of 10%
S4- m ly, if the concrete /
vould decrease the stress to 82.6%.
r steel vall had its bending rigidity increased by 13 or more per-cent the g/g g should approach unity.
It is not clear from locking at the drawings what method of reinforce =ent would be most appropriate. It vill be assumed that one of the following is possible:
1.vire mesh reinforcement across the entire N face (outside)
.,---.-, O C e t 2.steel plates epoxy bonded with devels into the vall l, , ' ,',-,, F tL n I..">, L%3.reinforce =ent of the top of the walls with steel plate caps (epoxy devel bonded to concrete)
.../...., . . .
.n.~,.: c_.-a.u L ,.Constraints nay be identified during the course of the reverk which vould eliminate any of the above. This list is only suggestive and not conplete.
/{'i f)M
.~;-.i In order to proceed with a solution to this task it is neces-sary to: 1.obtain LLL results in regard to load case, location of stresses, force.: or 2.develop an independent model and repeat their task.
F *irg a new load path sometimes constrains the existing
, structure so that the reactions are significantly redistributed and the original structure c:ay still be overstressed. Therefore some analysis should be done to evaluate load path readjustment.
If 1 is possible the fix should be structurally analyzed in M days provided decks and documentation is forthcoming.
The LLL SAP 4 model decks should be reviewed and LLL should be asked for their detailed verk in regard to determining the resulting response (figure 24, p. 36, ref. 4).
, If 2 is necessary the cost will incretse by about 60 to 70 percent of LLL pool analysis cost (excluding the cask drop portion).
There is at least a 0 5 probability a site visit vill be neces::ary with a travel cost of $112. The cceputing time cost is estimated to be $5C0.
'15 , t ~,'h2'-.
.'..References 1." Seismic Analysis of the Acid Liquid Wste Tasks at the Western New York State Nuclear Service Center," 'TCEL-52600, LLL, arch 1979 2.N. Mononobe and H. Patsuo, "On the Deter-bation of Earth Pressures During Earthquakes," Proceedings of World Engineering Confe enca 9, , p. 176, 1929 3.S. Okabe, " General Theory of Zarth Press.:re," Journal of Japanese Society of Civil Engineering, vol.12, no.1,1926.
4.R. G. Dong, S. M. E, " Structural Analysis of the Fuel ReceiW_ng Station Pool at the th2 clear Fuel Services Reprocessing Plant, West T Valley nee York," UCRL-52575, LLL, m y 1973.
p / l,^} h ';'16 n .~ e: n .
'- b .J /. .j