NRC Generic Letter 1979-66

NRC Generic Letter 1979-066: Information Regarding New Fuel Cladding Strain and Fuel Assembly Blockage Models and Compliance with 10 CFR 50.46

ML031320434
Person / Time
Site:	Beaver Valley, Millstone, Hatch, Monticello, Calvert Cliffs, Dresden, Davis Besse, Peach Bottom, Browns Ferry, Salem, Oconee, Nine Mile Point, Palisades, Indian Point, Kewaunee, Saint Lucie, Point Beach, Oyster Creek, Cooper, Pilgrim, Arkansas Nuclear, Three Mile Island, Prairie Island, Brunswick, Surry, North Anna, Turkey Point, Vermont Yankee, Crystal River, Haddam Neck, Ginna, Duane Arnold, Farley, Robinson, San Onofre, Cook, Yankee Rowe, Maine Yankee, Quad Cities, Humboldt Bay, La Crosse, Big Rock Point, Rancho Seco, Zion, Fort Calhoun, FitzPatrick, Trojan
Issue date:	11/27/1979
From:	Eisenhut D Office of Nuclear Reactor Regulation
To:
References
NUREG-0630 GL-79-066, NUDOCS 8001070276
Download: ML031320434 (32)

. V _.I- --

' - L-?

UNITED STATES

NUCLEAR REGULATORY COMMISSION

WASHINGTON, D. C. 20555 November 27, 1979 GI-- 77_66 (LETTER TO ALL OPERATING LIGHT WATER REACTORS)

Gentlemen:

Letters, dated November 9, 1979, were sent to all the licensees requesting information concerning the new fuel cladding strain and fuel assembly blockage models and compliance with 10 CFR 50.46. Subsequently further information was provided by the NSSS vendors and fuel suppliers regarding the impact of cladding heating-rate dependent burst temperature effects on rupture time and rupture strain/blockage and consequently on calculated peak cladding temperatures. The vendors and fuel suppliers supplied additional information by letters to the staff.

Copies of these additional letters are enclosed. This new information should be used in preparing your response to the November 9, 1979 request.

incerely, Darrell G. senhu ting Director Division o perating Reactors Enclosures:

1. Letter from Babcock & Wilcox, dated November 20, 1979

2. Letter from Combustion Engineering, dated November 16, 1979

3. Letter from Exxon Nuclear Co., dated November 16, 1979

4. Letter from General Electric Co., dated November 16, 1979

5. Letter from Westinghouse Electric Co., dated November 16, 1979

6. Letter from Yankee Atomic Electric Co., dated November 20, 1979 C?)e) -6-1-7

Baboock& icox v -24 PO. Rox3260. 4=?:*s, Va 24505 Terephioxe: MU 384-51 13 November 20, 1979 Hr. Darrell G. Eisenbut Acting Director Division of Operating Reactors Office of Nuclear Reacter Regulati"

U.S. NucleUr Regulatory Commission Vashington, D.C. 20555 SubJect: ClAdding Svefling and Rupture Models fcr LOCA Analysis Dear Mr. Eisenbutt On Wovebear 14, 1979, Kr. R F. Denise of the Division of Systeas Safety contact I&W with regard to the Burst Temperature Curve approved for nze by BE&W in LOCL analyses. li. Denitse re- quested BW to considcr the effect of ruloying the Staff's ramp rate correzltion. a contained in Draft KMMO 0630 to deteridne the Burst Temperature Curve for use in WC& analyses.

B&' has exained the ramp heat up rates calculated prior to rupture for B&W WSS syste= which have either OLs or CFs granted under 10 CW 50.46. (Documented in EA1-10102, Rev. 2, EW-10103AI, Rev. 3, ind EA-1010S, Rev. 1.) Ilterpolating from the Staff's r heat up rate versus imp stress and failre temperature referenced above, BSW has found that the Staffis correlation predicts the fuel cladding to rupture at the same or higher tamperatures for al1. cases, except the 4-foot coe elevation for the £77-Fuel AssSemby raised-loop pnt (SAW-

10105, Rev. 1). The razp rate prior to rupture fI this case is approxinately 12 C/s, whle the ertrapolation of the NkEC

curves to the EW Burst Temperature Curve at tbat &ae stress Indicates & 22 Ca hte t up rate. SW has estimted tbat the effect would be an earlier rupture, and, thetere, additional oxidation due to metal-water reaction, rulting in an increase of appr"imtel 6O1 In the peak cladding tupeerature (PCT).

The original analysis showed a peak cladding temwrature of

2073F. The addition of 80? wod result iu w peak of 2153'F

and not violate tie requirements of 10 CFR 530.4. Since the issuance of -1l0105, Rev. 1, B his idntified further con- servatisms which wmount to a reduction in peak cladding tempera- ture of approximately 3W . Therefore, if the evaluation vere The 9Obcok&VWikt Coz-V8V I W4tstt~f 1867

t Rabcock &Wfl=o Mr. Darrell Eisenhut November 20, l979 to Include these further conzervatims, and tie MC ramp rate correla- tions employed, we wioud expect a peak cladbiuS temperature intreate of about 50)F (2123V7 peak) with no difficulty in demonstrating om- pliance to 10 CFR 50.46.

In suzaary, BW bAs examined the effect of the use of the Staff's ramp rate correlation as requested and found the calculated PCT to be either unebanged at lowered as a result except for the one case noted above. If there are any questions concerning this response, please call me or Resry Bailey (Ext. 2678) of my staff.

Very truly your,.

3anes R. Tylor Manager, Lkensing JT/lc zz t Nd Oa 'AQN 6,wv e'.

A~oV I ..

ComhLEDfl nrinrr;sn3. o TI.cl~e 9 229

1W0 PIO5PeCl Hiii RoOi Windsm, connemicul: W095 POWER

SYSTEMS

Roveber I6. 1979 LD-79-067 Mr. Darrell C. Elsenhut -

Assistant Director for Systvs and Projects Division of Operating Reactors U- S. luclear RegulatorY cu'issidO

Wshingtnn, D.. L 20555 Subject: Fuel Cladding Shelling and Rupture Nodels G. Eisenhut.

Reference: Letter LD-79-064, A. E. Scherer to 0.

ste November 2. 1979 Dear Kr. Eisenhut:

(C-E) response The rferenced letter prmi4ded Conbustion Enginfeerig's to seyeral NRC concerns regatrdino rupture strain ard flow boage.. Su- questilons armse concerning sequent to receipt of the letter, additionaltonperatre effects on rupture the impact of heating rate dependent burst an peak cladding tem-per- time and rupture strainfblotkage arnd ultimately LO£A. The following evalua- ature (PCiT) in the analysis of a large break burst taterature tion of the potential impact of heating rate independent support of our oparating on C-E'S licensing calculations is provided plant custineri.

heating rate dependence.

The C-E rupture temperature model does not have 2-10yCfter znd using For the heating rate range of C-E operating plants, the ORL mrodel reccmended by the Staff. heating rate effects would lower predicted rupture temperatures by 25-750C. produce The resulting lower rupture earlier rupture times, tMrperatures due to low heat rate effects

(2-20 se-wnds earlier).

degraded heat tran&-

If rupture occurs after the tine of <1 Wnsec ref loIW.by Appendid K) is invoked fe on the rupture Woe ard above (as required could lead to higher re- at the time of rupture. Earlier rupture times i silipirentatAor. of degraded flood PC in this case because of the earlier experInze c3ad rupture heat transfer. However, all C-E operating plants the initiation of prior to the time of el in/sWm reflWd and thereforeby lower rupture topiatures.

degraded heat transfer would not be affected or refill periods mnay alter local Earlier rupture times during the blawdown or radiaticn enclosure beat transfer *ncmentarily, thrTouh wp conductancereflood its imPact on PIT

effects. Bowever, if the PCT occurs during late woUld not be significant.

r Dc Orrell G. Ekernhut-Z

Lower rupture temperatures due t I

low e3t rate effects may produce higher rvpture strains and blockages. The effect of increased rupture strain and blockage was addressed in the referenced letter tg The results of the previously discussed System 80 sensitivity the Steff.

studies show that PCT calculated witb the revised flow blockage/heat model is slightly lower than PT calculated with the present transfer heat tran5fer model. In addition,, the results of a study showflowblockaagef creasing the that in- degree of flow blockage from 6DO to 8DX only PCT by 400 F. &ised on these results, we conqlude that all Increases C-E

the operating plants continue to coinply with the ZZOCOF peak cladding teperature criterion, Including the effects of increased rupture strain/blo-kage.

The above discussions indicate that the reported PCT for all C4 opeating plants would not be siganficantly affected by a haating- rate rupture dependent temperature model. The magnitude of the effect on PC

no greater than effects observed in the System 80 sensitivity would be using the C-E alternate rodels. In facts it is expected that studies vised flow blockagefheat transfer models with or without a uting- rL-

heating -ate dependent burst tm aperature model for the analysis of C-E operating plants h-Duld produce lower PCT than presently reported values. C-E

therefore believes that our Evaluation %Idelanalysis with thf revised flow blockage/heat transfer miodl meets Appendix K requirements and the ZZaODF peak cladding tepperature criterion.

If I can be of any further assistance on this rmitter, pfiase contact uze or Mls. 3. H. Cicerchia of uy staff at (20)3S1-1911. ExtenSon slo 5.

Very truly )ours, CtiBWST1It BEIE~ThERINr INC.

Licensing tBanager AES:d-ag

lk*CtJ UCLt&Ahb :IYIne:.1m tC),ll L0 lute flt Horn hopids AD.,d

0. V>. &&x fjO, P~h0"ffJ~trl. li'st~t~fo g3t72 PAiinr f5019 94.Y R1.0 rale 32ti3 Novotnbpe 1t, 1q79 Mr. barrell G. tisenhut. Acting tirector bivision of Operating Peaictors Office of Nuclear Rpart.ror Regulation U. S. Nuclear Regulatory Conmnmission Washington, D. C; 2O!55 keferenrce ti: MNt letter frnmi G. r. Owtlpy to h. A.. tist.albut dated Nnvft*p&tt 4, 19I/9.

betit Har. Eisenhut:

A!s iaquested by your staff on November 13, 1979, tMt hp! cnmpleted Ah additional review of the licensing imapct of thr: rsriujed NRC rupture/

blockage model with pattLuler emphasis on the impact of the NRC tempera- ture-ramp-dependent rupture temperature curves. Tthi . Peview SupportS

the conclusion of ENC's earlier anal yte (Aeferehrt! 1) that there is ho adverse impact on licensing lintts for plants AnhAled by ENt modelt ttfo ute of the NRC rupture/blockage model.

the DC Cook analyses reported in tPftPvPnhr. (1) UP'ed the to:mrole.tr: Lt'ruLso'stl tuptuiC

NRC rupture/blorkage model including the temperature-tImp-dependent the temperaturv, rupture strain And flow blockage cbrrelatinF.. Thin.,

temperature ramp rate effects had been included. The teqserature rtip rate dependence in the NRC model is such that the difference between the predicted rupture temperature of the HRC and MNC tmdels Is gredlest for the 6lowest ramp rate. Resultt of this additional review are summarited below:

• All PiR plants licensed with tBC models have temperature set.onds ramp rates in the Slow range (clOGC/sec for & period of more than 10

prior to rupture.)

• For the category of plants where t.he POI oiturs dowtritreim of the ruptured nr,^d (it, st'am cooling) the DC Cook. plant is the mostttmpel-At'WF' sensitive to the NRC rupture/blockage model beculue it hat the tlnwp.t ramp rate prior to rupture.

Application of the NRC rupture/blockage model to the plant ti.withuntberattiVt the 0iowest tamp tate (DC Cook) Shows that the turrent FNC PttM' MnndI

&s discussed in Reference (1). thusr, it.i!, euutlutJvd tt;tconservative. the currtht tNC

tCCS 6nalyses for plants which tall in this cate gory are Thest plants are Palisades, Kewaunee and Prairie Isllnd 1 And 2.

T i rrl . Nknu

• Par one WH Ah~tyt+/-d 0thltt (At dIihhA) thl P4I bc~ti A~ hlth uipturad node Ind early in the 1-fil pod petiod ThA Mf~a h*n I~

ruturf/blockage model tb ft Ginna hat bmm 1akutetdo the tdtrot qodbi§

PC? than 6e"at found to be ltes tha" a 206r IthcIs wit Pv? still more than 20U6F below the f200tU Ilimtit.

uth

1h

& T'he remainingp tiC anhltyild P6" lap.nt. Oh ftabinstrh) db~5 h8t NAVI

&Steam tonling period nor dues thts Fit occur at thP 1-U turod nodi.fo Ir this Plant the' ruIpture Straifl calculdtO~by the NRC tuptiturs htrkift iftoda (considering the tamp rate ueIfpct oh tuptura. ternperhtuefi) I getrthatn the- tupture Itrain calculated by th Ml E Thuj,Sto CUptutcq/

blockage model would yield is low~er P'CT Siir the "biautf hlq ier Ma~d strain on the hotn-ruptured PC? ttudej Would Impruov& Lid tool in.

In guutmary. it is concluded that Application at~the ftRt 1uptut&blotkA06 Model in the ENC tCCS model would hot Affect liet~h~itq limits Oil ENC

Plants because:-

OCT's would be reduced by Using the MrC rtipture/blockagis hodel in all olants in which PCT does not occur an the 1tUPtUred hafde I in thp nite P ant Where Ptl` dofig Occur at' thp 1ruptut~d flod (tp tt the impact of the ftRC rupture/blockwjt' tunndl On OvC isleA than favl with imtre than A 9006F margin to tha 22Q0tF! limit r?-hn&intg.

Utiton kutil#&r ttnim0hy

GENERAL* ELECTRIC NUCLEAR POWER

SYSTEMS DIVISION

GENERAL ELECTRIC COMPANY, 176 CURTNER AVE., SAN JOSE, CALW*FRNA 95125 MyN 278-79 MC 682, (408) 925-5722 November 16, 1979 U. S. Nuclear Regulatory Commission Division of Operating Reactors Office of Nuclear Reactor Regulation Washington. D.C. 20555 Attention: Darrell G. Eisenhut, Acting Director Division of Operating Reactors Gentlemen:

SUBJECT: GE CLADDING HOOP STRESS AT PERFORATION

Reference: (1) Letter, R. H. Buchholz to D. G. Elsenhut (NRC),

ORNL Cladding Swell and Rupture Data - BWR

Evaluation, November 2, 1979.

(2) Draft Report, R. 0. Meyer and D. A. Powers, Cladding Swelling and Rupture Models for LOCA

Analysis, October 31, 1979.

(3) General Electric Company Analytical Model for Loss-of-Coolant Analysis in Accordance with 1OCFR5O

Appendix K, NEBO 20566, January 1976.

(4) Letter, A. J. Levine to D. F. Ross (NRC), GE Loss-of- Coolant Accident Model Revisions - Core Heatup Code CHASTE05, January 27, 1977.

During a telephone conversation between GE and R. Denise of transmittedthe NRC

Staff on November 14, 1979, additional information to that in Reference 1 was requested. Reference 1 outlined the reasons the data contained in Reference 2 did not affect the GE LOCA cladding swell- ing and rupture models (References 3 and 4). It is GE's understanding that the NRC Staff is concerned with the method affects cladding the used to calculate hoop ramp rate (clad heatup rate) during a LOCA as it of this letter is stress versus temperature at perforation. The purpose to address these concerns.

Section I.B of GE Appendix K Topical Report NEDO-20566 discusses fuel based swelling and clad rupture thermal parameters. this analysis was studies on our previously applied CHASTE04 model. Extensive sensitivity

GENERAL1 ELCICt'iC

U. S. Nuclear Regulatory Commission Page 2 were carried out by GE to prepare for NRC review of the currently approved CHASTE05 swelling and rupture model. These studies aro of direct relevance to the current NRC concerns. The sensitivity studies (results of which are included for completeness in Supplement A) indi- cated only a small sensitivity of PCT to variations in cladding strain and hoop stress at perforation. In particular, Figure 4 of Supplement A

depicts the variatton of the hoop stress at perforation with temperature.

The lower bound of the investigation has been re-plotted on Figure 54 of Reference 2 (attached). This figure shows that the lower bound of the CHASTE05 sensitivity analysis produces a more conservative relationship of hoop stress to perforation than the 0C/sec curve for temperatures above approximately 740'C (i.e., perforations are not expected in GE

BWRs below 925 0C). The change in PCT for this lower curve compared to the base case was -50F.

We understand that the Staff is also concerned about the statistical significance of the range of values over which ramp rates are deter- mined. The calculated cladding heatup rate for GE BWRs is between 1°

and 71F/sec. This range of heatup rates is based on an average value over the ballooning portion of the ramp.

The foregoing discussion, together with Supplement A, clearly indicates that the ORNL data for hoop stress at perforation for several heatup rates does not impact the conclusions of References 3 and 4 over the BWR

ranges of application.

I sincerely hope that this resolves any questions you may have regarding this matter as it pertains to the BWR.

Yours truly, R. H. Buchholz, Manager BWR Systoms Licensing Safety and Licensing Operation RHB:cas/4J

Attachments cc: G. G. Sherwood R. Mattson (NRC)

R. Denise (NRC)

L. S. Gifford (GE-Beth)

Supplement A

CHASTE05 SWELLING AND RUPTURE MODEL

Sensitivity Studies To evaluate the effects of the change in the calculation of the grey body factors (GBF) in the CHASTEO5 code, a number of sensitivity studies wore done. The studies show that the more realistic calculation of the cladding temperature GBF's results in a smaller sensitivity of the peak and rupture model.

(PCT) to various parameters of the rod swelling The studies were performed for a plant with hence 7x7 fuel at high exposures, any sensitivity of the to maximize the number of perforations and relatively long reflooding calculated PCT. The plant selected had athen results in a longer period time and a shorter blowdown period which perforated and hence a greater over which the rods are calculated to beand rupture model. The results sensitivity to change; in the swelling of those expected for presented here can be considered representative calculated to occur.

BWRs with fuel where perforations are in detail below:

The following studies were performed and are discussed

1. Variation of cladding strain at perforation

2. Variation of swelling initiation criteria

3. Variation of thermal expansion coefficients

4. Variation of perforation stress versus temperature curve

5. Variation of plenum volume

6. Vpriation of the GBF calculation time The base case for all the calculations was models calculated using the strains, described in NEDW-20566, perforation curve, strain rates and other and 23%as on inner rods for i.e., nominal strains of 16% on outer rods temperature transients for perforation hoop stresses <1500 psi. in The several rods for this case are shown Figure 1. Figure 2 shows the relative positions at the different rods.

results in a smaller In general, the use of CHASTEDS instead of CHASTE04 The two major reasons for sensitivity to changes in various parameters. the in the parameters the smaller sensitivity of the results to changes are:

A more accurate calculation of radiation heatheat transfer in a) transfer CHASTE05 has reduced the impact of radiation degradation when rods are calculated to perforate.

b) Better nodalization of the cladding in CHASTE05 (it has two cladding nodes instead of one in CHASTE04) and better control of the time step has reduced the sensitivity of the temperature response to inside metal water reaction as a result of perforations, takes a i.e., when a rod is calculated to perforate, the code small time step.

NS: cas: at/4T 1

1.0 Variation of Cladding Strain at Perforation The values of strain after perforation used in the based on the FLECHT Zr2 tests described in Section calculation I.B.2.4 ar NEDO-20566. It 15 assumed that for rods with hoop stress of rods next to the channel will have a maximum strain after (1500 psi, of 16% of nominal radius and for the reinianng rods, the perforation strain Is assumed to be 23% of nominal radius. The purpose maximum study was to determine the change in the temperature responseof this individual rods and the peak cladding temperature of the of a result of changing the various assumptions regarding bundle as perforation strain. The base case for this study was the assumed done using the nominal strains (i.e., 23% an Inner and 16% on outer rods).

The study shows that there is a very small (*5F) sensitivity PCT to changes in the perforation strain. This Is because, of the though individual rod temperatures are affected (by as even just after a rod perforates during the transient), the such as 200F

of all the rods in-the bundle tend to equalize as a resulttemperature of redistribution of energy by radiation heat transfer, consequently the overall effect on PCT is small. The studies show strain is increased on an individual rod its temperature that as the because for larger strains there is a larger area for decreases, heat and, hence, lower temperatures. For smaller strains the transfer are higher as the area for heat transfer Is smaller. temperatures The results for the different cases are presented below;

1.The strain on the first rod to perforate (Group was changed to 40%. The calculation showed no change 12)

in PCT but did show a slight decrease (c20F) in the temperature the transient for the first rod to perforate shortly after perforated. the rod Case 2. The strain on the second rod to perforate was changed to 40%. In this case, the PCT decreased(Group 10)

by SF

compared to the base case. The change was larger compared Case 1 because of the closer proximity of the final PCT to the second rod to perforate; but despite the change, rod to be noted that the change Is small. it should Case 3. The perforation strain on all rods was set at represents the maximum strain that adjacent rods can 30% (30%

without touching). The PCT decreased by only 3F even expand to the variation in individual rod temperatures during though were lower by as muchas 25F during the transient, the transient just after the rod perforated.

Case 4. The perforation strain on all rods was reduced

1aTtfie nominal value and the PCT decreased by 3F. to case also, the individual rod temperature transients In this by a larger value (up to 15F at certain times in the changed compared to the PCT. transient)

1S: cas/4T 2

The above studies were supplemented by studies using the strains strains measured in the FLECHT Zr2 test, instead of the nominal used in the above studies.

Case S

Strains measured in the Zr2 test (shown in Figure 2, p 1-175, NEDO-20566) were input into the CHASTE code, instead of the nominal strains of 16% and 25% on outer and inner rods, respectively.

Figure 3 shows the effect on the first rod to perforate (Group 12)

of using the nominal versus measured perforation strains for all the rods. The difference in the temperature transient for individual rods in the two cases is small, and the differences in the calculated PCTs is zero. As discussed earlier, the of reason for the small PCT sensitivity is the redistribution the temperature due to radiation heat transfer and the fact that the PCT rod at the end of the transient is a nonperforated rod. Early in the transient, the PCT rod is often a rod that perforates (as shown In Figure 3).

Case 6 This case was similar to Cases 1 and 2. In this case, the to strain on the first rod group to perforate was set equal

16%, 30%, and 40%. For all other rod groups, the Zr2 measured strains were used. The calculated temperature transient for the first rod to perforate is platted in Figure 3. The results show that for higher strains, the cladding temperature is lower shortly after perforation because of a larger heat transfer area, but there is much smaller sensitivity to strain than was calculated in CHASTE04. The reason for this is that CHASTEOS has finer nodalization of the cladding and hence a more accurate calculation of the surface temperature and the metal water reaction rate. Also, CHASTE05 has better time step control which results in smaller time steps after a calculated perforation. Because of these two reasons, the of cladding temperature does not increase rapidly as a result inside metal reaction as was observed in CHASTEN4.

The calculation for the 40% strain appears to show a larger sensitivity because of a slight delay in the perforation time.

But after a few seconds, the temperatures using the various strains all are about the same. The slight difference in the time of perforation for the various cases is a result of the slight differences in the strain rates before the rod perforated.

(The strains and strain rates before perforation are a function of the final perforation strain.)

The conclusion from this study is that the cladding temperature of perforated rods is relatively insensitive (<lOF, 15 seconds after perforation) and the PCT is almost completely insensitive to the perforation strains and, hence, use of the nominal values is appropriate.

NS: cas/4T 3

2.0 Variation of elling Initiation Criteria CHASTE calculates plastic swelling on rods for all temperatures above a certain temperature. This temperature is nominally set at 200F below the perforation temperature. Calculations were done assuming that plastic swelling starts OF, 20OF and 400F below the perforation temperature. The results show that for the case of OF, the PCT decreased by 3F, and for the 400F case the PCT was unchanged relative to the 200F nominal case. The effect on PCT was small (<SF), and the effect on Individual rod temperatures was also small (c20F), and hence it can be concluded that the use of 200F is still appropriate.

3.0 Variation of Thermal Expansion Coefficients This study was done to determine the sensitivity of PCT to uncertainty in the thermal expansion coefficients of the fuel and cladding material. The changes in the PCT are caused by changes in the gap conductance resulting from changes in the pellet-cladding gap size for different thermal expansion coefficients. For larger thermal expansion coefficients, the gap size is larger early in the transient resulting in lower removal of stored energy during the blowdown phase of the transient and hence higher PCT. Conversely for smaller thermal expansion coefficients, the PCT is lower than the case using the nominal thermal expansion coefficients. But in all cases, the sensitivity is relatively small and is documented below for two extreme cases, I.e..' no thermal expansion or contraction and twice the nominal thermal expansion coefficients. For zero thermal expansion coefficients for both fuel and cladding, the PCT

decreased by about 25F; and for twice the nominal expansion coeffi- cients, the PCT increased by about 15F. The small sensitivity thermal expansion, even for the extreme cases, justifies the useto of nominal thermal expansion coefficients.

4.0 Variation of Perforation Stress Versus Temperature Curve The purpose of this study was to determine the effect of changing the perforation stress versus temperature curve from the standard curve used. Two cases were studied, one for which the curve was below all the data, and another for which the curve was above majority of the data points (see Figure 4). The change in the a calculated PCT compared to the base case was S6F for the lower curve, and +2F for the higher curve. For the lower curve, the perforations occurred earlier and at lower temperatures; hence, the effect of inside metal water reaction was minimized and there was larger surface area for heat transfer on some rods for a longer a period of time. For the higher curve, even though perforations occurred later, they occurred at higher temperatures where inside metal water reaction is higher and hence a higher PCT. The important conclusion from this study is that the sensitivity to time and temperature of perforation has been reduced considerably with the Improved GBF calculation.

NS:cas:at/4T 4

5.0 Variation of Plenum Volume and temperature of The previous study determined the effect of tim the effect of the initial perforation on PCT. This study determines the effect of the initial plenum pressure on the PCT. To determine was varied from the plenum pressure, the Initial plenum volume pressure, the calculated nominal value by +/-40%. For the increased pressure, it decreased PCT increased by 5F; and for the decreased internal pressures have by IF. This study shows that the fuel rod PCT. This is primarily an insignificant effect on the calculated is also small, because the effect on the number of perforations did not change i.e., the number of rods calculated to perforatepressure) but when the volume was decreased (i.e., to increased decreased by only two rods compared the base case when the In this study also, volume was increased (i.e., lower pressures).

different conditions but the time of perforation changed for the the sensitivity was small.

Time

6.0 Variation of Grey Body Factor Calculation at different times Different rods are calculated to perforate the rod swelling occurs a during the heatup transient. As most of perforate, it is appro- few seconds before the rod is calculated to of each rod. A

priate to calculate GBFs at the time of perforation PCT to two bounding study was done to show the sensitivity of theIn one case, for the assumptions about the calculation of G6Bs. that the perforated rods radiation calculation only, It was assumed assumed that at the did not swell. In the second case, it was to 23% strain in the first perforation all the rods had swollenthe procedure that is used calculation of GBFs.* The second case is was assumed, the in CHASTE04. For the case in which no swelling case where GBFs PCT was lower by about 20F compared to thethestandard second case where GBFs are calculated at each perforation. For calculated PCT was were calculated assuming all rods swollen, the Figure 4 shows about 11OF higher compared to the standard case. study shows that the variation of PCT for the three cases. This extremely conserva- the PCT using the old (CHASTE04) procedure was transfer is not very heat tive, and the degradation in radiation a few rods, which is the large as a result of the perforation of calculations.

typical case in BWR loss-of-coolant accident KS: cas/4T 6

. 2500 a500

vD T A~IAe PEWM~tAT~ot4

2000 GROUP 15 GFOUFPIP.

(

5?0

-GPZU?19

50o OIUNG 1RAN4ITON N/4N Oe.eff p-~

(

PA i//C Af A*Ot'e Di1. E , . e

0

2 4 6 40 60 ieCO 300

7-A."t

00.001,. ow 5e, FIGURE 1 CALCULATED ROD ThNPERATUES

USING K014INAL STRAIN VALUES

I

,ea.i A'vs94gA'

AAN~owj" eoLoo 00VAW60A

-,I

Z#**/o

00 00000

2

3 34

00 00000

00

_

v

0000t?

%-

00 - - - - - ---

O

-- -

ot

- - - - ---

Oa

-A

J

' i

,Ru4.t,#wgf ,4's?7e-A

Iv *A VAAr.1 v A 4,VOE 2.-

1eob 4.~p

I, . -. _ .. -. .9.... .. .- . - , .............-

• <

b §@ *

b  ;*  ;;;4. -- *. 4- d I

.

. ,  ;. . ,

n--  ; ... .. ..

I4 -

..

• -

- ,,- ,.

..

.

,

. .

-

. . .

s

, .

-0I

. .

..

. .

A/,

9 ...

-

/,.3s3-

/ .-..

.

4 ..

s - ~~* - .9  ;*o- le s}r* t-w-@i;;

2 . - 7tS g9 - 4 / .. -:

.. /.. _..,, ',_ . _.

. . . 4 6_.z . 9  ;

,*

.* .. ..

.

3 I - -4~ S . .5- 1-:; ---

~~~~.

-

t 1%O * s - /z .  ;. .**- * * **-- ;. ...............-

.-. _2

. ** CX z . S**' ',* - , .,-,:..

• .

gl 4 5~iO . , e . C Sj....  ;.-

.- * .AAC '. -* ' ,F1UR

-62

4 .

5. 9§M..

NEDO21426 bOat UPPER

BOUND

5~

I

I

wwo- V

DESIGN CURVE

LOWER

BOUND

looc0-

'oFQ

l I I I I

aCI-

1*20 13o0 100 2200 Wm 2500

CLAVOING TtMPERATURE 16PI

Figure 4 Cladding Perforation Stress

9-11

2500 emoeplew1^01--el aA 046'* AL

,&OyAwhom -,,O r^ls -ro"

esrhfcv,4. s. -

qtax eooo/

5o00

'400

• 500

TR*41T1014G UNCOVERY)1 SPWRAY START

IRST P MOProI~rON OCCURS

0

0 100 aojof'%2oW

P. 4 to FIGURE 5 EFFECT OF CHANGING GBF

CAL CULATION PROCEDURE

z

I

f I

GE & BURST TEMVIPERATURE CURVES

I

5 la 15 20

ENGINEERING HOOP STRESS (KPSI)

FIGURE 54 OF REFERVNCF 2

NLM3r TechrKcay M.vit'.n Westinghouse Water Reactor Divisions Box 355 Electric Corporation PitTsurgh Penrnii-Wama U

November 16, 1979 NS-TMA-21 63 Mr. Darrell G. Eisenhut Director, Division of Operating Reactors Nuclear Regulatory Commission

7920 Norfolk Avenue Bethesda, Maryland 20014 Dear Mr. Eisenhut:

Letter NS-TMA-2147, dated November 2, 1979, responded to LOCAIECCS NRC concerns related to the fuel rod models used in the Westinghouse of evaluation model and potential non-compliance with the requirements on fuel rod lOCFRPart 50. Table 1 of that letter included information on our initial heatup rate prior to burst. That information was based evaluation of the results of current LOCA analyses for Westinghouse plants with operating licenses. Subsequent to completion and transmittal of that letter,.Westinghouse continued investigation of heatly rote Westinghouse then calculations. As a result of that investigation, to burst. That prior developed a procedure to determine clad heatup rate the LOCA transient to procedure keys on the calculated clad strain during rate-calculation.

establish a starting point, in time, to use in the heatup meeting on November That procedure was presented to NRC personnel during a interim basis, as adequate

13, 1979, in Bethesda, and was accepted on an the revision to the with respect to Appendix K LOCA analyses. Table A shows heatup rates previously given in Table 1 of Letter NS-TMA-2147.

than Inspection of Table A shows heatup rates, in some cases, less

250F/sec.

In the current WECCS Evaluation Model (Feb, '78) used for the above analyses, a fuel0 rod burst curve which represents Table A, since someforcases burst conditions heatup rates of 25 F/sec and larger was used. From have heatup rates less than 250F/sec and burst conditions change for lower that some of those analyses could be heatup rates, Westinghouse recognized non-conservative with respect to the time of rod burst.

Therefore, W performed an evaluation of all operating heatup plants licensed with the WECCS Evaluation Model with respect to use of a rate dependent bursT model. The heatup rate dependent burst model currently used in the W

Small Break Evaluation Model (documented in WCAP-8970-P-A "Westinghouse System Small Break, October 1975 Model" and approved Emergency Core Cooling by the NRC) was used in this evaluation.

-

NS-TMA-21f3 November 16, 1979 Page Two The results of that evaluation, the status of each plant evaluated and justification of conclusions reached are as follows:

PLANT (1) MODEL FEB. '78 FQ 2.31 PCT 2172 A new analysis was performed using the appropriate heatup rate burst curve and water residing in the accumulator lines (not previously accounted for) was considered. The resulting PCT was 21350 F at an FQ of 2.31.

Therefore, lOCFRSO criteria are satisfied.

PLANT (2) MODEL OCT. '75 FQ 2.17 PCT 2199 A LOCTA run was made using the Oct. 75 evaluation model with appropriate heatup rate burst curves for FQ 2.16. PCT - 2127 Use of Feb. '78 evaluation model, in particular the new accumulator discharge model, will compensate for the AFQ, shown above, to maintain 22009F. (This is a burst node limited plant)

PLANTS (3) (4) (5) (6)

Since the heatup rate for the hot rod is greater than 250 F/second and the PCT does not occur during the steam cooling period, the current analysis for these plants remains valid.

PLANT (8) MODEL OCT. '75 F 2.10

PA 2188 F

An Oct. '75 model LOCTA run was made using appropriate heatup rate burst curves. Results were: FQ - 2.10, PCT - 2227.

Application of the "Dynamic Steam Cooling" modification of the Feb. '78

0

evaluation model will result in a 60 F reduction in PCT and the Feb. '78 accumulator discharge model will result in at least a 200 F reduction in PCT. Results of a Feb. '78 model analysis are expected to result in a PCT of approximately 2147 0 F at an FQ of 2.10.

Therefore, lOCFR5O criteria will be satisfied and there is no safety concern.

PLANT (9) MODEL OCT. '7'

FQ 2.25 PCT 2142

Novemni1er 12, 1979 Page Three Based on the results of a calculation for plant 1(14). the use of approximate heatup rate burst curves would result in a maximum PCT

increase nf 680F. Thus, the estimated (maximum) PCT = 2142 + 68 = 22100 F

at an Fq - 2.25.

The benefits associated with the Feb. 78 accumulator discharge model and accounting for paint on containment heat sinks will result in a PCT reduction well in excess of lOF.

Therefore, no safety problem exists.

PLANT (11) MODEL FEB. '78 F 1.90

POT 2124 A LOCTA calculation was performed using appropriate heatup rate burst curves. An F of 1.89 resulted in a PCT of 21610F.

Therefore, a peaking factor reduction of less than 0.01 is required for this plant to remain in compliance with lOCFRSO.

PLANT (12) MODEL OCT. '75 Fn 2.21 P1,T 2198 Based on analyses performed for plant f(15), a 15F/second reduction in clad heatup rate impacts hot rod burst to effect PCT by +42°F. Extrapolating, a 170F/second reduction in heatuD rate results in a 480F PCT increase. Use of the dynamic steam cooling calculation on the accumulator discharge model in the Feb. '78 ECCS evaluation model results in an estimated (600F + 200 F)

80OF reduction in PCT.

Therefore, a Feb. '78 model analysis would result in a PCT of 2198+48-8O=2166 F

at F of 2.21 and no safety problem exists.

PLANT (13) MODEL FEB. '78 Fn 2.05 P12T 2172 A LOCTA calculation was done using appropriate heatup rate burst curves and the results were:

F - 2.05, PCT 2191F

Q

Therefore, no safety problem exists.

PLANT (14) MODEL FEB. '78 Fn 2.32 PET 2124

NS-1MA-2163 November 16, 1979 Page Four A LOCTA calculation was done using appropriate heatup rate burst curves and the results were:

F - 2.32, PCT - 2192OF

Q

Therefore, no safety problem exists.

PLANT (15) MODEL FEB. '78 FT 2.32 PET 2158 A LOCTA analysis was done using.appropriate heatuo rate burst curves and the results were:

FQ - 2.32, PCT = 2200OF

Therefore, no safety problem exists.

PLANTS (16) and (17)

The latest licensing analyses have been verified to use appropriate heatup rate burst curves and therefore remain valid.

PLANTS (18) and (19)

New LOCTA analyses. were performed using aoDrOpriate heatup race burst curves.

The PCT was virtually unchanged. Therefore, no safety problem exists.

Based on the detailed information provided above, the Westinghouse Safety Review Committee concluded that two plants were found to require a reduction of 0.01 in allowable core peaking factor to maintain a PCT of 22000F. Four other plants have current analyses to the October, 1975 version of the Westinghouse model and may require a peaking factor reduction. However, we.

believe that reanalyses with the most current Westinghouse LOCA/ECCS

evaluation model (February, 1978) would show that no changes are necessary.

That is, we believe margins available in this model will more than offset any effect associated with the change in the fuel clad burst curve. A copy of the NRC notification letter (NS-TMA-2158) retarding this iss;;o is attdcheu.

The above information was also presented to the NRC Staff at the November 13,

1979 meeting.

Following the November 1, 1979 meeting, Westinghouse has again reviewed the, ORUL data quoted as a basis for NRC concern regarding adequacy of the WAppendix K blockage model. Comparison of individual rod burst strains from ORNr data to the corresponding Westinghouse data which has used as a basis for our blockage model indicates the ORUL data is in excellent agreement with the W data. Since the axial distribution of the burst strains in the ORNL multi rod '3urst test has

I-

NS-TMA-2163 November 16. 1979 Page Five been shown by ORNL to conform to local temperature distributions in the specific heating rods used in the tests, conclusion as to the applicability of the axial distribution of bursts (which is the Darameter that relates individual burst strain to flow blockage) cannot validly be made. Never- theless, the blockages measured from the ORAL tests are similar to those calculated by the Westinghouse model, which has been approved by NRC, when due consideration is made in translating blockages measured in 4X4 bundles to blockages applicable to 15X15 or 17X17 rod fuel assemblies using accented statistical techniques. Thus, we believe no immediate action is aoDropriatp.

with respect to reanalysis of Diants using the proposed NRC blockage model pending detailed review of the proposed model.

As a result of further investigation and evaluation, the following can be concluded:

1) A modification to the W model to account for the heatup rate dependence is necessary for compliance to Appendix K.

2) The impact of this modification is relatively small, effecting only two ooerating plants in terms of requiring peaking factor adjustments to meet the criteria of lOCFR50.46. The affected utilities ana the NkC ndve been czdeiqadLly inifurr-ied.

3) Comparison of the Westinghouse data and ORNL data shows excellent agree- ment and the current Westinghouse model, in the range of interest, is still appropriate.

It is therefore concluded that no safety problem for Westinghouse plants has been identified and all plants are in conformance with NRC regulations since the burst temperature modifications (1 and 2 above) are accounted for.

Very truly yuurs, T. 14. Anderson, Manager Nuclear Safety Department

NS-TMA-2163 November 16, 1979 Page Six TABLE A

REVISION TO HEATUP RATES TRANSMITTED

IN lETTER NS-TMA-2147 CASE HIEATUP RATE ( 0 F/SEC)

HOT ROD AUG OR AW ROD

1) 8.5 10.9

2) 20.3 13.1

3) 25.6 18.0

4) 25.0 15.4

5) 31.5 19.4

6) 27.4 23.8

7) (Not Westinghouse Fuel)

8) 19.1 7.4

9) 12.3 12.0

10) (Not Westinghouse Fuel)

11) 6.2 11.3

12) 8.0 11.4

13) 18.3 16.1

14) 9.3 14.3

15) 8.2 13.8

16) 39.6 23.7

17) 43.2 26.7

18) 22.7 17.6

19) 26.5 16.7

Westinghouse Waler Reaclor PBtothPxI),AL

Electric Corporation DiviSionS

November 16, 1979 NS-TMA- 2158 Mr. Victor Stello Director, Office of Inspection and Enforcement U.S. Nuclear Regulzatory Coornnission East West Towers Building

4350 East West Highway Bethesda, MtD 20014 Dear Mr. Stello:

Subject: ECCS Evaluation Model This is to confirm our telephone conversation with Mr. Frank Nolan on Friday afternoon, Noverrmer 2, 1979. In that Conversation we reported a non-conserva- tive feature in Westinghouse large break ECCS'evaluation models.

The Nuclear Regulatcry Com-mmission staff met November 1, 19iW, with representa- tives of reactor vendors and nuclear fuel suppliers -- Combustion Engineering Inc., Exxon Corporation, General Electric Company, Westinghouse Electric Corporation and Babcock and Wilcox Company. Utilities which operate nuclear power plants were informed by NRC.

The purpose of the meeting was to discuss the staff's ongoing evaluation of the results of tests on electrically-heated fuel assemblies conducted at the Oak Ridge (Tennessee) National Laboratory, .JRC indicated that emergency core cooling system analytical codes currently used to evaluate the effects of postulated loss-of-coolant accidents (LOCA) might not be in compliance witn NRC regulations. The portion of the codes in question deal with the effects of fuel clad swelling and rupture and blockage of cooling water.

Subsequent to the meeting, Westinghouse performed a detailed evaluation of the most recent analyses for operating plants and on November 2, 1979, Westinghouse confirmed, in writing, that the impact of the information pre- sented by the NRC has negligible impact on the LOCA analysis results of tne plants licensed with the Westinghouse LOCAIECCS evaluation model. The %RC

staff has concurred with this conclusion.

<-

Mr. Victor Stello -2-. NS-TMA-2158 However, as a result of that detailed evaluation, Westinghouse has now recog- nized that . non-conservative feature could exist In-the Appendix K LOCA

analysis with respect to the portion ot the calculation reltited to tue' rod burst. The potential non-conservative feature of Westinghouse larget break ECCS evaluation inodels is as follows. The models use a curve which represents fuel clad burst conditions tor clad heatup rates of 25F/second and yreater The evaluation discussed revealed thdt heatup rates could he less than

25*F/second. During the LOCA tra~nsient, thle tuel cldd burst curve dstablishes the time of clad burst 3nd (since :le clad temperature ana the pressure dit- ferential across the clad 3re.changing throughout the LOCA trenslert) "ne post-burst conditions of the clad. The fuel clad burst curve is deoendeft on the clad heatup rate prior to burst and a reduction in heat.Ap rate ca.stes earlier clad burst. A shift in clad burst time can affect the peak clac tem- perature (PCT) calculated for the LOCA transient.

Therefore, in order to more fully evaluate'this effect, the clad heatup rate prior to burst was determined from the most recent LOCA analyses for ti'osd plants licensed with the Westinghouse LOCA/ECCS evtluation model. Plants having heatUp rates less than 25 /seccnd were reanalysed to ascertain tne effect on peak clad temperature. Two plants (Turkey Point 'Jnits 3 and d) were found to require a reduction of 0.01 in Fg to maintain a peak Clao Tem-pera- ture (PCT) of 2200 F. A third plant, Indian Point Uni' ac. 2, was nct expected to reculir any FQ recuction, considering -he prsoit PC7. Ind avaii- able sensitivity studies. Analyses, underway at the time of cur telephone conversation, have now been completed and confirm this.

Four other plants, currently not operating (Trojan, forth Annd Unit 1, Indian Point Unit 3 -and D. C. Cook Unit 2) have current analyses Lo the October 3975 Westinghouse rrodel and on that basis might require a reduction in Fn. How- ever, we believe that reanalyses dtth the most recently aporoved Westinqho;se LOCA/ICCS evaluation model (February 1973) would show that no changes are necessary. That is, we believe margins available in this model will more than offset any effect associated with the change in the fuel clad burst curve.

We have advised the affected utilities of this unreviewed safety question. As part of this overall evaluation, we are examining plants under construction and will report as dppropriate. Please teel free to contact Dr. Vincent Esposito (412-373-4059) if yuu should have any questions.

Very truly yours, T. M. Anderson, Manager Nuclear Safety Department

/wpc

I.0; ^

T:Vwnr: . IttieS4 -4C

fiG

mSr-3*Q,-07fl PAE -I 3 YANTEE ATOMIC ELECTRIC COMPANY

taSX e >, e 9 niteS Stas Ese- kgitory nsss Attntiat Of f f le tor Xteulatitm I-f. Darell tizezstkt ketin'r-Iraw t~efer- (1) Lit N. P71-3 (Docket No. . 1C-i

(2) Ietter ta lEfl Mt "Tvalmtitf Ce dn sl ai4'4 Ewtv~n Nodels, t dated Nontber 2, IS79.

(DUIf NaG. 0630 dAted 111S179, entitfle, "Claddin trilling cn- tuptrs Pa-Ilc foar L= Anilywit.

(4)'Lnnfif~4, Nule~r o~cy~-kuu aGenric tM ra ti Nd Upt Etv--11, ily 1976a De=r Sir:-

Snh-jfl -Snstia of CleAdin, Swlling - w el This letter is - aAdakn4 to : Fsulitted- toyn an trmlcr 2, l97l, Lterrnne t. U is ftrrizc in reponse to DitiWAf £Vweicmu raises VI your staff cocrnng te haling of cl*ding bc-irp rate depentnCe iu T's li C- dels fw clin 1 afliug mae-rp.. The Ifoi am hopefully respauive to YOE qwstitx

¢ £t tk sltle t te rt cktr orretia lt irri& taperutc regim n ( 10C 000e or grectorY Tfhis Is X t the reltively 1w f11 pa rpmwure is ]thITake twA IU. IC thisr

-t taseratn range, rn'ttr tesw ttwv is oxtresely sauitiuw to oCflcumtJrap twera

- ftr E taui- Zf jt Writ a 1w,,

a a t 543. C/5OCsa.

o At rresat, Tain nes strz Ccrrwitic, Wcdified for Yats RCm

4wintry, -of Wrst taperature vs.. I Taint's Wrst strain n

-ttrufl- aru9titi6Ci averpredicts th- Stais capred t ~ hs cwrrulstzca n%- ct rap~-rat&e£sjtret. With regard to rlnt

KY -

1. 5. Thtlnr Zeplstw Cciac Uo!g 70 19

'.04 -

tnperature, we currently perceive tbe corretctin of burnt ttr-xturw

-v*, %tres to 'be ap1inb1. to afl ,rmp ntn ic high teprattre ruptnre. With regard to bar-it atroiu, tek &rrelations that envelope the slcwrnp ani fast-rasp fLC draft ur-£retatha of bunt train vs. turmr4t tay o - y not be cors tin eit n

1. wtlwr cmt tAlcutste clad ruptua.

This pvscptIcm of rp rate imsensitivity at high tflptrtztms way be modifitd as -Mr Ut-s in th& lrNt ttatup rY TS, hi_% biWrw tmIqxprtttfl ngicn is assessed. There apears to be a ratstepcmec AA2octiAted with Icktr tetperatce burst. no clear Jstifieation existe s f extrspolati of.w rarw teperate rzqtnre data to hit tee(veratutrt rtpture eictiticns., On czia handzlple rod brtdata frIn nxeriats dons ina tnwA r to be limted to high (2t 0 Chia.)

tww"rltre r=pS in this high te ertur ngi. 0. the other ,

limted dnta frcu sh&Is rod tests in wems, for ezaple, indicate tazt ra dspen4uuve at bigb t*tperatnra coul1 be possible.

tan'" has Ammd tM i pct of atilizit tbe fC draft turns

.s"dasted with sio rp rate effects (dfetrca 3) by per ca nw-2 cal~cluticn vrwerk£ tin exposure Jistributica for th reainder ofb te prn"vt cycle. Than calcialatians ban teen pqrtormeA

in the fcflcmTo amr:

CD) Mi- S= 0%C/se-cuen for burnt tepr In. *trus was mified to reflect T ' rgtu of by iterrati ba3yJse.

latwn= thu OtCnc. ad -te 2VC/aa. rrelatitn touaed in refi

0) e ai-ras p burst ain- (7igure & of reference 3) was used in tW~-2 satbouj* n data points grs *usbsin the- hi

(3) thw sir-rnm iocal fiw biocka" c- ofvnftc e 3 was med

.a;g with the I'-WXfl-j (refereoce 4) 11w tt umdttplin tunes associted with 20% rdrteicu La fleot aresX, oeistat with the prutdidn of the time-rap local blockap *cc-.

sed an this aysis, Tn* c ed tt r prt.

the pleat for tes raaindut of tb* pryet cle Is zinuffecced by the Irluzsia of the attn r rate draft corralatis in the TIO liSed

1.d Armeuti in i sd Titr r w d cated mnmy. Speufil fAr frIh tsl a i. present expo l r bea-titw Of 9.60? wIlft meets Appendix 4 critais with the 7CV-Z

ucxifieatita mted shov.; 9.550 kw/ft is t l pner oPeration.. This fuel is CUrrently licese St1d tm7/ft At tm present

• ~Vezm T the MiO pr etposA fuel, Pf bufit Apendix I

nd 9.4 r/ft it neefk for full wr b ft

- CrrZeatly limied. Etd of cycle oO eaft Irm thee i l id the Cmis .-. l to etasize that

jcu: K.- r* 4.

V. S. tla ,ttr~terv ot;*f ic Lw-r e 20 1979 d nr we -CZsZ1scy euppprt tim nlidty Of the c tin rsfccre S cA that ! p1 it od sky krateEt. ke -- s opraE n the bc4nain of Cj1 s~taO! 1nfp hxbe avaafl to Q& tse anlyes If there it little beutuprqate dee e ot rupture t t for hig ttmpta-tv.i- failure, tesenldgl4 q6=f tt=-= posil inpaCt of low te-ea--tare rate tube ruptar fl ee tYak M- cppiia-eaittcas by t all 'tx trh Orto/nc. " IwXIT

jz f- s lbs 0C s data fr ntuTre terrratnx fl10C. A quadrmtic fit thrt@- this &t* Fet yialdz tOE

foli-win Vorelaticn: t T' F a "7.75 - IMQ p. o.ooz e42 (2i DAI; ?I in &6grn, AT inzi)

Substitntf! this c1prrtla1tio i ntt 7VQf-, Iz ith the c 'c strain corz-elsth; of n n 3 c fta tbht th_ peat ct.!

teaperatudre for fresh fiel at £,O PMr= (the most t ti it Lim exposw-c rcgic caawidred abowF) is JJT0 le thE Ycket'r airreomt lfeiee luedstel predicts.

c idtratita it"se Yu tet the fol

(1) Knp depenent c- sr asratlited with ca-dng swelling ski rupture nAela shnold ld t affet plat e tit for t

(2) j-aze 4 rn= rwruaticu of tht dat of reference 3, (and in prart c te paucity of data appropriate to cleddin-4 swellng Amd rutr rrfw at Toalx Rwr) w.e cwwwitvt tbat IUm 's arnt

( ) CoottAilo arqttCiritim ad evAlna-ticn of i LUs, pitulty i th $-c raW ratc1 Lit lPeraturC Ingiotii requirred. During IP, till k* ewroacbing t flZ with a. rcz4 kntupm odal to re-place* U -2As pest of ou Uenr EC- ca"S. This m.odel, wiln MdTe ni ATable data.

(4) ParricuIla attenle stould be placed c' the strain a btzrst data.

amd ft nalaricnbh to rwrtrt ivt cla ;resure zan£-d If yo 1 Ay quenties resadixn this letter, please fuel Ie to concz Dr. LAss!f mac or Dr. Stephen?. tehpitt of rItle tnu iUnri rizg

,R r--tQfw vCvwf-z Sire i7.Mmy Twm &ImC izceIC Wwr U. j4&m tA. 4a

£Cflna

w.

Mr. William J. Cahill, Jr.

Consolidated Edison Company of New York, Inc.

cc: White Plains Public Library Ms. Ellyn Weiss

100 Martine Avenue Sheldon, Harmon and Weiss White Plains, New York 10601 1725 I Street, N.W.

Suite 506 Joseph D. Block, Esquire Washington, D. C. 20006 Executive Vice President Admi nistrative Consolidated Edison Company of New York, Inc.

4 Irving Place New York, New York 10003 Joyce P. Davis, Esquire Law Department Consolidated Edison Company of New York, Inc.

4 Irving Place New York, New York 10003 Richard Remshaw Nuclear Licensing Engineer Consolidated Edison Company of New York, Inc.

4 Irving Place New York, New York 10003 Anthony Z. Roisman Natural Resources Defense Council

917 15th Street, N.W.

Washington, D. C. 20005 Dr. Lawrence R. Quarles Apartment 51 Kendal at Longwood Kennett Square, Pennsylvania 19348 Theodore A. Rebelowski U. S. Nuclear Regulatory Commission P. 0. Box 38 Buchanan, New York 10511 John D. O'Toole Assistant Vice President Consolidated Edison Company of New York, Inc.

4 Irving Place New York, New York 10003