ML20125D695

Responds to NRC 791109 & 27 Ltrs Re Info on New NRC Fuel Cladding Strain & Fuel Assembly Blockage Models & Compliance w/10CFR50.46.Currently Operating Under 1975 Westinghouse ECCS Model Modified in 1978

ML20125D695
Person / Time
Site:	Trojan File:Portland General Electric icon.png
Issue date:	01/09/1980
From:	Goodwin C PORTLAND GENERAL ELECTRIC CO.
To:	Eisenhut D Office of Nuclear Reactor Regulation
References
TAC-07718, TAC-12689, TAC-7718, NUDOCS 8001150423
Download: ML20125D695 (6)

Text

,e

• rd 2crer.i E:ecdc Cccccr.y January 9, 1980 Trojan Nuclear Plant License NPF-1 Docket 50-344 Mr. Darrell G. Eisenhut Acting Director Division of Operating Reactors Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, D. C. 20555

Dear Mr. Eisenhut:

Your letters of November 9 and 27, 1979 requested information concerning the new NRC fuel cladding strain and fuel assembly blockage models and compliance with if- CFR 50.46. This letter responds to your concerns as they relate to t 2 Trojan Nuclear Plant.

The currently applicable Trojan ECCS evaluation was performed using the October 1975 version of the Westinghouse Evaluation Model, with correc-tions to accommodate the zirconium-water heat-of-reaction error identi-fied in 1978. Pending NRC approval of proposed Technical Specification changes that reflect this latest ECCS evaluation (submitted as License Change Application 51; March 15, 1979), Trojan has been operated since the startup of Cycle 2 with an administratively controlled total peaking factor limit (Fq) of 2.25. With this value of Fq, the calculated peak cladding temperature (PCT) was 2142*F.

On the basis of recent ORNL experimental data, draf t NRC report NUREG-0630 presents new correlations to be used in LOCA calculations for predicting cladding burst and strain and flow blockage. At issue is the extent to which incorporating these new NRC models would impact the Westinghouse ECCS evaluation model. An independent bu t related issue has been identi-fied by Westinghouse. Namely, in the February 1978 version of the Westinghouse ECCS Evaluation Model the heatup rate dependence of fuel rod burst was not properly considered.

90023150 f 'p.yfes 4

< - a

D D d 3 i ~l 5 2 '~ ' : [ .' ?~ 2 2,' ' ,

Mr. Darrell G. Eisenhut January 9, 1980 Page Two The implications of the current fuel rod model and heatup rate concerns have been addressed in the following four letters to you from Westinghouse:

NS-TMA-2147, 11/2/79; NS-TMA-2158, 11/16/79; NS-TMA-2163, 11/16/79; and NS-TMA-2174, 12/7/79. Of particular concern was the conclusion in letter NS-TMA-2163 that, for Trojan (plant 9), accounting for a clad heatup rate prior to burst <25'F/sec could result in a calculated PCT as high as 2210*F. However, it was also noted that using the February 1978 version of the Westinghouse Evaluation Model instead of the October 1975 version would be expected to result in a reduction in the calculated Trojan PCT well in excess of 10*F. Therefore, it was concluded that no safety problems exist for Trojan.

Further details of the combined effect of using NRC versus Westinghouse fuel rod models and properly accounting for heatup rate prior to burst were provided in letter NS-TMA-2174. It was concluded on a generic basis that the maximum impact on an ECCS evaluation performed with the February 1978 version could be a required Fq reduction of up to 0.045 or an increase in the calculated PCT of up to 81*F.

Af ter reviewing this situation with Westinghouse, we have elected to have a complete ECCS evaluation performed for Trojan using the February 1978 evaluation model. This will provide a suitable long-term basis against which to assess the impact specifically on Trojan of the use of the NRC fuel rod models. These calculations are currently underway at Westinghouse, and are scheduled to be completed by the end of January  !

1980. We anticipate forwarding the results to the NRC, together with l proposed Technical Specification changes, if appropriate, in February  ;

1980. Attachment 1 provides a preliminary Westinghouse evaluation of the impact of the fuel rod modeling concerns from the results available to date from the February 1978 version evaluation of Trojan. Based on i these results, we expect that when the February 1978 version evaluation is completed, it will be shown that no changes in the Trojan Technical Specification limits will result from fuel rod modeling concerns.

]

As outlined above, for the short period until the February 1978 version j evaluation has been completed for Trojan, an extrapolation based on j the October 1975 version Trojan ECCS evaluation, with the present peaking J l

factor, has the potential of yielding a PCT of 2210*F. In order to assure  ;

l during this interim period that the 10 CFR 50.46 PCT limit of 2200*F is  :

not exceeded, we are administrative 1y limiting the maximum Fq to 2.23. l Westinghouse evaluations performed for Portland General Electric Company have indicated that this Fq reduction of 0.02 from the present value will decrease the PCT by more than 10*F.

90023151

b T'": :::, ". [ .. : ::,~ r '

Mr. Darrell G. Eisenhut January 9, 1980 Page Three l

To summarize, a revised ECCS evaluation is being performerd for Trojan using the latest h7C-approved Westinghouse ECCS Evaluation Model. This evaluation and a specific discussion of the impact. of the current fuel rod modeling concerns on Trojan will be submitted to the NRC in February 1980. Preliminary results indicate that following this submittal no Technical Specification penalties will have to be imposed on Trojan pending completion of NRC review of the benefits to be associated with 1 improved Westinghouse analytical and modeling techniques. In the interim until the February 1980 evaluation is completed, an administrative 1y 1 imposed limiting Fq reduction of 0.02 will assure that the Trojan ECCS meets the requirements of 10 CFR 50.46.

Sincerely, 9~ D(W C. Goodwin, Jr.

Assistant Vice President Thermal Plant Operation and Maintenance CC/DIH/CJP/gah/4mg3A17 Attachment c: Mr. Lynn Frank Subscribed and sworn to before me this day of dl40/d/M 1980.

)

Onl/ AA

1. Nota 6:f Publi(6(Oregon My Commission Expires: ~[ f" h

/

ATTACINENT 1 A. Eva[uation of the potential T@act of using fuel rod models pre.

sented in draf t NUREG-0630 on the Loss of Coolant Accident (LOCA) analysis for T ro tan h et t .

! This evaluatien is based on the limiting break LOCA analysis identi-

{ fied as follows: ,

BREAK TYPE - DOUBLE EXDED COLD LES GUILLOTINE BREAK DISCHARGE COEFFICIENT O.6 L'ESTINGWUSE ECCS EVALUATION H0 DEL YERSION kben 4 \%

I

. CORE PEAKING FACTOR 2 . '57_

WT ROD FrulHUN TEMPERATURE CALCULATED FOR THE BURST REGION OF THE~

CLAD - M 90. DF = PCTg _ y

, ELEYATIom - kO Feet.

HOT FDD F/dIntA TEMPERATURE CALCULATED FOR A fCN-RUPTURED REGIOR OF i THE CLAD - \ c69. OF = PC4x r. . . .. , 1 ELEVATION - 1.7 Feet [

CLAD STRAIN DJRING BLOWDOWN AT THIS ELEVATION O Percent Tru1K'd 4 CLAD STRAIX AT THIS ELEVATION - 3/gs _

Percent Maxisum te=perature for this node occurs when the core reflood rate is (GREATELGii2id than 1.0 inch per second and reflood heat transfer is based on the (FLECHT/"#*NJ calculation.

AYERAGE 10T ASSEMBLY R00 BURST ELEVATION - _ Ia. OFeet 4 T.9 Percent '

HDT ASSEMBLY BLOCKAGE CALCULATED -

2. B'JRST ft0DE .

Tne utxt=ss potential impact on the ruptured clad node is egressed in letter 55-TMA-2174 in terms of t.te change in the peaking f actor limit (FQ) required to maintain a peak clad tzz-

, perature (PCT) of 2200cF and in terms of a change in PCT at a constant FQ. Since the clad-water reaction rate increa.se: sig- -

. nificacitly at temperatures above 2200.0F, individual effects (such as APCT 'due to changes in several fuel md codels) '

indicated here may not accurately apply over irge rarrges 90023153

L .f but a simultaneous change in FQ which causes the PCT to remain in the neighborhood of 2200,of justifies use of this evalua- h.

- ' tion procedure.

.. From NS-TMA-2174:

O for, the Burst Node of the clad: j

. y 0.01 AFQ + .c 150'F BURST N0DE APCT .

I

- Use of the NRC burst model could require an FQ reduction of 0.015  ; .

i The $N#$ estimated imact of using the NRC strain f -

model is a required FQ reduction of 0.03. j k

Therefore, the maximum penalty for the Hot Rod burst node is:

APCT1 = (.015 + .03) (1500F/.01) = 6750F Margin to the 22000F limit is:

APCT2 = 2200.0F - PCT 8= 2 "bO. OF ,

jI The FQ reduction required to maintain the 22000F clad tempera- -

ture limit is- .

l AFQ g = (APCT1 - LPCT2 ) I~ W) '

150'F  ; j

= (I *2.5Q (h)

~

• • O.05 (butnotlessthanzero). ,

2. NON-8URST NODE The maximum teraperature calculated for a non-burst section of h

~

clad typically occurs at an elevation above the core mid-plane  !

. during the core reflood phase of the LOCA transient. The poten-tial impact on that maximum clad temperature of using the NRC fuel rod models can be estimated by examining two aspects of the-analyses. The first aspect is the change in peitet-clad gap

' conductance resulting from a difference in clad strain at the non-burst maximum clad tc.:perature node elevation. Note that clad strain all along the fuel red stops after clad burst cccurs and use of a different clad burst model can change the tf.se at

- which burst is calculated. Three sets of LOCA analysis results

< were studied to establish an acceptable sensitivity to apply 1 1 generically in this evaluation. The possible PCT increase resulting from a change in strain (in the Hot Rod) is +20.0F per percent decrease in strain at the maximum clad temperature .

e 9002 W . I s

- 7, t #

Tocations. Since the clad strain calculated during the reactor coolant system blowdown phase of the' accident is not changed by~

the use of tRC fuel rod models, the maximum decrease in clad '

strain that must be considered here is the difference between I

the " maximum clad strain

• and the " clad strain at thc cd cT 20S

' blowdown

• indicated above,

, dan *x) 2 Therefore: .

i

. 6 PCT 3 = (3%.01 strain ) (

• N # - O N # N N)

=([)(.0I.%-A)

= Tt. . 'F The second aspect of the analysis that can increase PCT is the flow blockage calculated. Since the greatest value of blockage  ;,

indicated by the NRC blockage medel is 75 pe cent, the maximum .

PCT increase can be estimated by assuming that the current level I

of bloctage in the analysis (indicated above) is raised to 75 percent and then applying an appropriate sensitivity formula l shown in NS-TMA-2174. -

Therefore, l

APCT4 = 1.25cF (50 - PERCENT CURRENT BLOCKAGE)

+ 2.360F (75-50)  !

= 1.25 (50 iS S ) + 2.36 (75-50) l

= (A. 8F .l If PCTg occurs when the core reflood rate is greater than 1.0  !'

. inch per second APCT4 = 0. The total potential PCT increase '

for the ncn-burst ncce is then  ;'

6PCTS*#+#4 3 Karpin to the 22000F limit is APCT6 = 22000F - PCTs  : Ml' f ,

The F0 reduction recuired to maintain this 22000F clad tem- I perature limit is (from NS-TMA-2174) I 1

AFQ g = (APCT5~#)( 6 0 )

10 T APCT s

AFQg = - AtA but not less than zero.  !

l O . .

9002M% . ll .

I L -_

g , , .

,4 g

b The pcating factor reduction required to 'raintain the 2200 'F u

- clad temperature limit is therefore the greater of AFQg andMQN ' k, e

or; 0.03 4 FQP N TY = ,

p i The effect on LOCA analysis results of using improved analytical and b-S. N stodeling techniques (which are currently approved for use in the Upper Head Injection plant LOCA analyses) in the reactor coolant systr.m blowd:wn calculation (SATAN computer code) has been quanti-fled yta an analysis which has recently been submitted to the NRC for review. Recognizing that review of that analysis is not yet  !

co=plete and that the benefits associated with those model ic: prove. '

cents can change for other plant designs, the NRC has established a '

credit that is acceptable for this interim period to help offtat

. penalties resulting from application of the NRC fuel rod models. i That credit for two, three and four loop plants is an increase in the LOCA peaking factor limit of 0.12, 0.15 and 0.20 respectively. l C. The peakirs factor limit adjustment required to justify plant il optration for this interia period is determined as the appropriate 'l ArQ credit identified in section (B) above, minus the A TQ l calculated in section (A) above (but not greater than zer$f." !l ,

s To Acm Txtni - o,7 o - o,c3 i j

-9 O

(,

I

=

1;;

l l1 l

i

_.....__...._.gm.....,m._.,

~ *

• a.- -e.+. . . , . . ~ . - . . . , , , , , , _ _ _ _ , ,,

l 9

+

. s 90023156 li I ,