ML20058Q367

Requests Withholding of WCAP-11397-P-A, Revised Thermal Design Procedure from Public Disclosure Per 10CFR2.790

ML20058Q367
Person / Time
Site:	Seabrook
Issue date:	12/09/1993
From:	Liparulo N WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:	Murley T NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
Shared Package
ML20058Q365	List: ML20058Q362 ML20058Q367
References
CAW-93-553, NUDOCS 9312290047
Download: ML20058Q367 (8)

Text

.

Westinghouse Energy Systems 82 355 Md@ Penn::ytvania 15230 0355 Electric Corporation December 9,1993 CAW-93-553 i

Document Control Desk US Nuclear Regulatory Commission Washington, DC 20555 Attention:

Dr. Thomas Murley, Director APPLICATION FOR WITHHOLDING PROPRIETARY INFORMATION FROM PUBLIC DISCLOSUBE

Subject:

North Atlantic Energy Service Corporation Letter and Application for Withholding Proprietary Information from Public Disclosure to Document Control Desk.

Dear Dr. Murley:

The proprietary information for which withholding is being requested in the above-referenced letter is further identified in Affidavit CAW-93-553 signed by the owner of the proprietary information, Westinghouse Electric Corporation. The affidavit, which accompanies this letter, sets forth the basis on which the information may be withheld from public disclosure by the Commission and addresses with specificity the considerations listed in paragraph (b)(4) of 10 CFR Section 2.790 of the Commission's regulations.

l Accordingly, this letter authorized the utilization of the accompanying Affidavit by the North Atlantic l

Energy Service Corporation and Yankee Atomic Electric Company.

' Correspondence with respect to the proprietary aspects of the application for withholding or the Westinghouse affidavit should reference this letter, CAW-93-553, and should be addressed to the undersigned.

3 Very truly yours, Nicholas J. L paru 3, h.anager Nuclear Safety and Replatory Activities Enclosures

~~

'~

cc:

Kes in Bohrer / NRC (12115) l 9312290047 931217

'8 l

PDR ADOCK-05000443 $I P

PDR h c

Proprietary Infor nation Notice Transmitted herewith are proprietary and/or non-proprietary versions of documents furnished to the NRC in connection with requests for generic and/or plant-specific review and approval.

i in order to conform to the requirements of 10 CFR 2.790 of the Commission's regulations concerning the protection of proprietary information so submitted to the NRC, the information which is proprietary in the proprietary versions is contained within brackets, and where the proprietary information has been j

deleted in the non-proprietary versions, only the brackets remain (the information that was contained l

within the brackets in the proprietary versions having been deleted). The justification for claiming the information so designated as proprietary is indicated in both versions by means of lower case letters (a) through (f) located as a superscript immediately following the brackets enclosing each item of information being identified as picprietary or in the margin opposite such information. These lower case letters refer j

to the types of information Westinghouse customarily holds in confidence identified in Sections (4)(ii)(a) through (4)(ii)(f) of the affidavit accompanying this transmittal pursuant to 10 CFR 2,790(b)(1).

-1 i

i i

Copyright Notice The reports transmitted herewith each bear a Westinghouse copyright notice. The NRC is permitted to make the number of copies for the information contained in these reports which are necessary for its internal use in connection with generic and plant-speci6c reviews and approvals as well as the issuance, denial, amendment, transfer, renewal, modification, suspension, revocation, or violation of a license, permit, order, or regulation subject to the requirements of 10 CFR 2.790 regarding restrictions on public disclosure to the extent such information has been identi0ed as proprietary by Westinghouse, copyright protection not withstanding. With respect to the non-proprietary versions of these reports, the NRC is permitted to make the number of copies beyond these necessary for its internal use which are necessary in order to have one copy available for public viewing in the appropriate docket files in the public document room in Washington, DC and in local public document rooms as may be required by NRC regulations if the number of copies submitted is insufficient for this purpose. The NRC is not authorized to make copies for their personal use or for members of the public who make use of the NRC public document rooms Copies made by the NRC must include the copyright notice in all instances and the proprietary notice if the original was identiGed as proprietary.

q l

. CAW-93-553 t

I AFFIDAVIT COMMONW EALTH OF PENNSYLVANIA:

ss t

COUNTY OF ALLEGHENY:

l t

i

.t Before me, the undersigned authority, personally appeared Henry A. Sepp, who, being by me i

duly sworn according to law, deposes and says that he is authorized to execute this Affidavit on behalf of Westinghouse Electric Corporation (" Westinghouse") and that the averments of fact set forth in this i

Affidavit are true and correct to the best of his knowledge, information, and belief:

j i

i W

Henry A. Sepp,(Manager L

/ '

Strategic Licensing Issues Sworn to and subscribed j

before me this /Yf day of b

,1993.

[

1. 1&u Y a

p Notary Public tetria! Seal 1.onaba M. Polec. Notary Pubic MaravneBac Avonycon/

My Commion Ex;ies Dec 14.1995 Membw,Pennspane Asenwn at tc,ures

!. CAW-93-553 -

l (1)

I am Manager, Strategic Licensing Issues, in the Nuclear and Advanced Technology Division, of the Westinghouse Electric Corporation and as such, I have been specifically delegated the f

function of reviewing the proprietary information sought to be withheld from public disclosure in connection with nuclear power plant licensing and rulemaking proceedings, and am authorized to apply for its withholding on behalf of the Westinghouse Energy Systems Business Units, i

(2)

I am making this Affidavit in conformance with the provisions of 10 CFR Section 2.790 of the

]

Commission's regulations and in conjunction with the Westinghouse application for withholding accompanying this Af6 davit.

l m

(3)

I have personal knowledge of the criteria and procedures utilized by the Westinghouse Energy l

Systems Business Units in designating information as a trade secret, privileged or as conndential commercial or Gnancial information.

(

i (4)

Pursuant to the provisions of paragraph (b)(4) of Section 2.790 of the Commission's regulations, l

the following is furnished for consideration by the Comalission in determining whether the information sought to be withheld from public disclosure should be withheld.

f 1

(i)

The information sought to be withheld from public disclosure is owned and has been held t

in confidence by 1 stinghouse.

j i

(ii)

The information is of a type customarily held in confidence by Westinghouse and not customarily disclosed to the public. Westinghouse has a rational basis for determining l

t the types of information customarily held in confidence by it and, in that connection,._

utilizes a system to determine when and whether to hold certain types of information in j

confidence. The application of that system and the substance of that system constitutes

.l Westinghouse policy and provides the rational basis required.

)

j Under that system, information is held in confidence if it falls in one or more of several l

types, the release of which might result in the loss of an existing or potential competitive advantage, as follows:

1 l

I I

i

. CAW-93-553 l

i

~

(a)

'lle information reveals the distinguishing aspects of a process (or component, structure, tool, method, ' etc.) where prevention of its use by any of Westinghouse's competitors without license from Westinghouse constitutes a l

~

competitive economic advantage over other companies.

l 1

i (b)

It consists of supporting data, including test data, relative to a process (or i

component, structure, tool, method, etc.), the application of which data secures a competitive economic advantage, e.g., by optimization or improved marketability.

(c)

Its use by a competitor would reduce his expenditure of resources or improve i

his competitive position in the design, manufacture, shipment, installation, assurance of quality, or licensing a similar product.

i (d)

It reveals cost or price information, production capacities, budget levels, or commercial strategies of Westinghouse, its customers or suppliers, i

(e)

It reveals aspects of past, present, or future Westinghoase or customer funded i

development plans and programs of potential commercial value to Westinghouse.

i r

(f)

It contains patentable ideas, for which patent protection may be desirable.

j i

There are sound policy reasons behind the Westinghouse system which include the l

following:

t (a)

The use of such information by Westinghouse gives Westinghouse a

_j competitive advantage over its competitors. It is, therefore, withheld from disclosure to protect the Westinghouse ' competitive position.

i b)

It is information which is marketable in many ways. The extent to.which such information is available to competitors diminishes the Westinghouse ability to sell products and services involving the use of the information.

+

.i 9

+

t

-- CAW-93-553 (c)

Use by our competitor would put Westinghouse at a competitive disadvantage by reducing his expenditure of resources at our expense.

(d)

Each component of proprietary information pertinent to a particular f

competitive advantage is potentially as valuable as the total competitive l

advantage. If competitors acquire components of proprietary information, any

{

one component may be the key to the entire puzzle, thereby depriving i

Westinghouse of a competitive advantage.

l i

1 (e)

Unrestricted disclosure would jeopardize the position of prominence of Westinghouse in the world market, and thereby give a market advantage to the i

competition of those countries.

l i

(f)

The Westinghouse capacity to invest corporate assets in research and development depends upon the success in obtaining and maintaining a i

competitive advantage.

l l

(iii)

The information is being transmitted to the Commission in confidence and,' under the provisions of 10 CFR Section 2.790, it is to be received in confidence by the Commission.

i (iv)

The information sought to be protected is not available in public sources or available information has not been previously employed in the same original manner or method to I

the best of our knowledge and belief.

!i (v)

The proprietary information sought to be withheld in this submittal is that which is

[

appropriately marked in " Revised Thermal _ Design Procedure," ' WCAP-11397-P-A

(

i (Proprietary), April 1989, for information in support of North Atlantic Energy Service f

Corporation's submittal to the Commission, transmitted via letter, North Atlantic, NYN-

![

93020, dated February 2,1993 " Request for NRC Review and Approval of Analysis

-l Methodologies to be Applied to Seabrook Station", and Application for Withholding i

Proprietary Information from' Public Disclosure, Nicholas J. Liparulo, W.', Manap,er

[

Nuclear Safety and Regulatory Activities to the attention of Dr. T. Murley, Director, f

Office of NRR. The proprietary information as submined for use by the Yankee Atomic -

Electric Company for the Westinghouse reload cores is expected to be applicable in other.

j licensee submittals in response to certain NRC requirements for justification of statistical thermal design procedures.

I i

.. ~

4

.u.

c

, CAW-93-553 l

1

)

]

This information is part of that which will enable Westinghouse to:

i (a)

Justify the statistical methodology associated with the Revised Thermal Design l

Procedures.

l 1

fb)

Assist its customers to obtain licenses.

j (c)

Optimize reactor design and performance while maintaining a high level of fuel integrity.

-l t

Further this information has substantial commercial value as follows:

i (a)

Westinghouse plans to sell the use of similar information to its customers for purposes of future fuel upgrades.

-l (b)

Westinghouse can sell support and defense of the product to its customers in the licensing process.

l i

Public disclosure of this proprietary information is likely to cause substantial harm to the i

t competitive position of Westinghouse because it would enhance the ability of competitors j

r to provide similar improved core thermal performance methodology and licensing defense.

l services for commercial power reactors without commensurate expenses. Also, public

{

disclosure of the information would enable others to use the information to meet NRC

.l requirements for licensing _ documentation without purchasing the right to use the l

information.

De development of the technology described in part by the information is the result of applying the results of many years of experience in an intensive Westinghouse effort and the expenditure of a considerable sum of money.

.j l

L!

In order for competitors r,f Westinghouse to duplicate this information, similar technical programs ~

would have to be performed and a significant manpower effort, having the requisite talent and i

experience,.would have to be expen: led for developing the enclosed improved core thermal i

performance methodology.

]

Further the deponent sayeth not.

n

, d

, j ;f y }$.}

,l.

". * ' : [

~,

f 1

+

by 4;

't-

c y

..;g Wt :q

[

3

j; T,e

- }

t-l

,y; ~ 1

- Qs.

i;f t -

L--

y-ss

...e, 3

3 y~

a M

v r,

n.-

L.,;

' 4.a

' E.;

j lj.:

I.'

- 4 e

't;..

e m: c.,

M..,

y

'r{.,'\\ -

a l:

'i.

af

._3 n j,,

.,f n

s,

ib.h liS.Ndade:N$b45* >Ec.hd,-+

M^ 5'~R h ', _.

g cm<q

j{,' '.

ll e( p m r

m 4.

< s,> +

' si y l ~., w

=

w

\\.

.c.

jf Glll; < ' L

• L;.;lZE Q

'l0

<1 g

j j!(#"'

y;3.5.{yl :?f L )';;}.

w g * ;,ib gwag y;

4 i

y 3

m w.

.c 3

"7 k '

.\\..

l f

ug ip p, e,

4 u

i

.%}_ '

.-3\\ s{

_gJ y e/

M-w t

\\,

w_

p,.

mg

'k++

emm mvefMs jdM L

%r n% a#a-s - im nn v

4 *.

a

,Ps

h. H 41

)a

^

~.

>, a

- w3.,l,p

,; r

] ^-

= Q ' N-

j.,

a wn

+

g k

N nh.

'A t

Q:

,e

+

[w(cs:c3 ;i M e "

i s rt,_

'yA y im s

y, e

ww 3-e a;as

,~ alY

%. w@?? W.

W 1 J.

,,gMahn j_.?

,.:y(A 1&y%

~

' ; ; @' 'l%g IM,p.>'dA w Mi ll yg. '

4

/

4gg gy g 4.;;

m x

~m

~

'y# Y t.,. :{ -

q.

TC.[i Q y.: ^[n;gy%v-$, h.

c4

'J n

n.

mA

.wO wa "I.M@Q,

-E-J$$ -

'E-g

.g 7

I" g

97]lWestindiousel &

ijh[h,)

phf I

g Commercial NuclearnFusi Olvisiott J4; 4

i M

i

~

$0 I

m n.

m..

f.D

~

g y

"-\\.

'~

.6<...~..a i.

... I i 1

. Nh g.

Westinghouse Class 3 l

l WCAP-11397-A i

b I

t i

=,

REVISED THERMAL DESIGN PP.OCEDURE l

A. J. Friedland S. Ray Original Version: February 1987

[

Approved Version: April 1989 Approved:

E. H. Novendstern, NFD Manager, Thermal-Hydraulic Design and Fuel Licensing i

WESTINGHOUSE ELECTRIC CORPORATION l

Commercial Nuclear Fuel Division l

P.O. Box 3912 Pittsburgh, Pennsylvania 15320 t

I 545tLir 880328 h.

~.-

WCAP-ll397-A TABLE OF CONTENTS Section Description A

NRC Acceptance Letter for Referencing of l

Licensing Topical Report WCAP-ll397, i

" Revised Thermal Design Procedure" B

NRC Safety Evaluation (SER) j C

Westinghous'e submittal letter to NRC for-UCAP-ll397-P i

t D

WCAP-ll397-A Text l

E Response to NRC Questions on UCAP-11397

" Revised Thermal Design Procedure

{Non-Proprietary)

^!

i e

f t

-l

.I I

t i

I t

~

t i

SECTION A I

1 9

1 t

r

)

i l

1 k

J f

'5 b

+

+

I I

t i

i 1

f.

n i

1 e

b e

i k

t l.

9 a

-I e

4 s

iI

.f.

I 5

t b

')

?

b F

.I i

c

[:w:u %

UNITED STATES 8

NUCLEAR REGULATORY COMMISSION n

WASHWGTON. D. C. 20$$5

\\..... p'l M I 71963 2.'lU []

Mr. W. J. Johnson, Manager Nuclear Safety Department Westinghouse El.ectric Corporation Box 355 Pittsburgh, PA 15230-0355

SUBJECT:

ACCEPTANCE FOR REFERENCING OF LICENSING TOPICAL REPORT WCAP-11397, " REVISED THERMAL DESIGN PROCEDURE" We have completed our review of the subject topical report submitted by Westinghouse by letter dated March 16, 1987. We find the report to be acceptable for referencing in license applications to the extent specified and under the limitations delineated in the report and the associated NRC evaluation, which is enclosed. The evaluation defines the basis for acceptance of the report.

We do not intend to repeat our review of the matters described in the report and found acceptable when the report appears as a reference in license applications, except to assure that the material presented is applicable to the specific plant involved. Our acceptance applies only to the matters described in the report.

In accordance with procedures established in NUREG-0390, it is requested that Westinghouse publish accepted versions of the report, proprietary and non-proprietary, within 3 months of receipt of this letter. The accepted versions shall incorporate this letter and the enclosed evaluation between the title page and the abstract. The accepted versions shall include an -A (designating accepted) following the report identification symbol.

Should our criteria or regulations change such that our conclusions as to the acceptability of the report are invalidated, Westinghouse and/or the applicants referencing the topical report will be expected to revise and resubmit their respective documentation, or submit justification for the continued effective applicability of the topical report without revision of their respective documentation.

Sir.cer y R

~

Ashok C. Thadant Assistant Director for Systems Division of Dgineering & Systems Technology Office of Nuclear Reactor Regulation

Enclosure:

Topical Report Enclosure i

. 4

.a-..-.

._x

. =..

---

.xn..

x -

w. a

,.a-.

.a,--.

a a

a

_e..

a a.

h i

SECTION B i

I i

f

?

6 1

' I t

4

' I I

i

. i h

+

i s

1 n

f 6

6 a

m

[

s f

t I

t 1

- I5 i

6 s

i

-l I

5 1

t

' t!

?

l

f SAFETY EVALUATION BY THE OFFICE OF NRR RELATING TO TOPICAL REPORT WCAp-11397 REVISED THERMAL DESIGN PROCEDURE WESTINGHOUSE ELECTRIC CORPORATION f

1.

INTRODUCTION The subject topical report WCAP-1397, describes a revised thermal design procedure (RTDP) for predicting the departure from nucleate boiling ratio (DNBR) design limit in Westinghouse pressurized water reactors (PWRs). This is a modification of the existing improved thermal design procedure (ITDP) methodology (Ref.1) which has been reviewed and approved by the NRC (Ref. ?).

As with the ITDP, the RTDP is also based on consideration of system uncertainties in plant operating parameters, fabrication parameters, nuclear and thermal parameters, and the use of the appropriate departure from nucleate boiling (DNB) correlation for the plant.

The DNB design basis remains the same, namely, that there must be at least a 95 percent probability at a 95 percent confidence level that the limiting power fuel rod will not experience DNB during normal operation and anticipated operational occurrences.

Parameter uncertainties or variances obtained from the evaluation of data are detennined at a 95 percent confidence level.

With the ITDP methodology, these system uncertainties were statistically combined separately from the DNB correlation uncertainty. The two were then combined directly rather than statistically to determine the DNBR limit. The proposed RTDP methodology would combine the system and correlation uncertainties statistically rather than deterministically. This is similar to the statistical DNBR evaluation methodologies developed by the other PWR i

vendors and approved by the NRC.

j The staff review encompassed the original submittal as well as responses to j

staff requests for additional information (Ref. 3). The staff was assisted in this review by our consultants at Pacific Northwest Laboratories.

2.0

SUMMARY

OF TOPICAL REPORT The topical report describes the mathematical relationships used in both the approved ITDP and the proposed RTDP and presents a sample calculation of a representative plant using both methods to illustrate the difference between them.

In addition, a description of how fuel rod bow is accounted for in both methodologies is given and the effect on rod bow penalty is compared.

Finally, a discussion as to how sensitivity factors are determined for various statepoints over an appropriate range of conditions is given.

3.0 EVALUATION The existing ITDP method of protecting against DNB in Westinghouse pressurized cater reactors was reviewed extensively by the NRC and a staff evaluation was issued in 1978 (Ref. 2). Because the ITDP resulted in a large reduction in DNBR margin as compared to the traditional method, referred to by Westinghouse as the fixed value method, the staff examined all the parameters to assure that all uncertainties had been appropriately considered. The general methodology for the RTDP will be implemented in the same manner as for the 1TDP. Sensitivity factors will be determined for each new correlation, set of design parameters, or range of applicability. The design parameter variances will be determined on a plant specific basis by the identical procedure currently used for the ITDP.

However, since the RTDP extends the ITDP methodology in that the DNB correlation will be statistically combined with the system uncertainties, the adequacy of the relationship of the DNBR uncertainty factor to changes in the values of the design parameters given by S dx /x, dy/y =

g q

was evaluated. The procedure used to check the validity is the same as that previously used and evaluated by the NRC and was based on test calculations performed by Westinghouse using the THINC-IV computer program. These test results indicate that, with the sensitivity factors used, the above relationship i

provides a conservative model for changes in DNBR, as calculated by THINC-IV, for small changes in parameter values. Therefore, no uncertainty allowance is required for this equation with the sensitivity parameters used in the topical report. However, if sensitivity factors change as a result of correlation

+

changes or changes in the application or use of the THINC-IV code, the uncertainty allowance must be re-evaluated.

A linearity approximation was made in the ITDP to obtain Ecuation (2-9) in the topical report. This equation is used to determine the statistical parameters for the DNBR uncertainty factor from the design parameters. There is a corresponding linearity assumption in the derivation of Equation (2-17) in the report for the RTDP which must be validated.

In investigating the adequacy of l

the linear approximations, Westinghouse compared results both with and without f

the assumption of linearity. The cases tested were those suggested by the NRC I

in the evaluation of the ITDP (Ref. 2).

In every case, the linearity assumption gave more conservative results than the calculation which did not assume linearity.

Based on this, the staff feels it is reasonable to expect the linear approximations to be conservative and, therefore, acceptable.

In the existing ITDP, fuel rod bow is accounted for by a correlation which relates the upper 95 percent tolerance limit for the standard deviation of channel closure for the worst span and burnup. This is combined with a relation between DNBR penalty and channel closure. The DNBR penalty is then combined statistically with the CHF correlation uncertainty to calculate a i

limit DNBR. The rod bow penalty is the percent difference between this limit DNBR and the limit DNBR with no rod bow. The analysis in the proposed RTDP methodology is performed in the same way except that the DNBR correlation f

uncertainty is statistically combined with the plant parameter and other uncertainties for both the unbowed and bowed cases. This results in a slight l

decrease in the rod bow penalty compared to the ITDP methodology. For example, for 17 x 17 standard fuel at the rod limiting burnup of 24 GWD/MTU, the ITDP resulted in a rod bow penalty of 1.1 percent whereas the RTDP resulted in a rod bow penalty of 1.0 percent using the WRB-1 correlation. The staff considers the effect of the small difference to be negligible and, therefore, finds the rod bow treatment described in the topical report acceptable.

4.0 STAFF POSITION i

The RTDP procedure for calculating DNB limits, as presented in WCAP-11397, is acceptable for use in licensing applications.

It provides a reasonable approximation to the proposed statistical basis. As with the existing ITDP, however, certain restrictions must be imposed on the implementation of the method because of the sensitivity of the method to changes in the correlations and codes used. These restrictions are:

1.

Sensitivity factors used for a particular plant and their ranges of 1

applicability should be included in the Safety Analysis Report or reload submittal.

2.

Any changes in DNB correlation THINC-IV correlations, or parameter values listed in Table 3-1 of WCAP-11397 outside of previously demonstrated acceptable ranges require re-evaluation of the sensitivity factors and of the use of Equation (2-3) of the topical report.

3.

If the sensitivity factors are changed as a result of correlation changes or changes in the application or use of the THINC code, then the use of an

)

uncertainty allowance for application of Equation (2-3) must be re-evaluated and the linearity assumption made to obtain Equation (2-17) of the topical report must be validated.

4 Variances and distributions for input parameters must be justified on a plant-by-plant basis until generic approval is obtained.

5.

Nominal initial condition assumptions apply only to DNBR analyses using RTDP. Other analyses, such as overpressure calculations, require the appropriate conservative initial condition assumptions.

)

l

1 l 6.

Nominal conditions chosen for use in analyses should bound all permitted methods of plant operation.

7.

The code uncertainties specified in Table 3-1 ( 4 percent for THINC-IV and : 1 percent for transients) sest be included in the DNBR analyses using RTDP.

The statistical method as presented includes no explicit design margin to accommodate unknowns. Such a margin could reduce or eliminate the impact of core relat'ed problems which are discovered after a core is designed and after a plant is operating. Although no particular margin is quantified, margin is inherent in the overall procedure used with the revised thermal _ design procedure. This margin is available to offset the effects of yet-to-be-discovered design problems. However, if newer procedures are proposed which substantially reduce thermal nargin, then a design margin to accommodate unknowns should be explicitly identified.

The parameter ranges do not cover the range required for part loop operation.

If the method is to be used for analysis of part loop operation, the topical report must be amended to cover this wider range.

5.0 REFERENCES

1.

Chelemer, H., Bowman, L. H., and Sharp, D.R., " Improved Thermal Design Procedure, "WCAP-8567-P, July 1975.

2.

Letter from D. F. Ross, Jr. (NRC) to C. Eiche1dinger (W), " Staff Evaluation of WCAP-7956. WCAP-8054, WCAP-8567, and WCAP-8762," April 19, 1978.

3.

Friedland, A.

J., and Ray, S., " Revised Thermal Design Procedure,"

WCAP-11397, Addendum 1, June 1988, t

m

I I

SECTION C k

l a

s e

i A

i t

5 l

r h

i I

t t

t i

L f

s s

i s

l I

f e

r I

f L

1 3

k l,

1 l

3 t

h n

Ia I

B i

1 h

5 h

t t

r e

t i

h

~h i

i t

s e

t

+

.. ~.....

O i

. Westinghouse _,

Power Systems pQreema m3003n i

Electric Corporation NS-NRC-87-3209 March 16, 1987 t

1

~

Mr. James Lyons, Chief Technical & Operations Support Branch Office of Nuclear Reactor Regulations

][

U.S. Nuclear Regulatory Commission Washington, D.C.

20555 t

ATTENTION: Document Control Desk ATTENTION: Carl H. Berlinger, Chief

\\

-i Reactor Systems Branch _

Division of PWR Licensing-A Submittal of Westinghouse Topical, WCAP-11397, " Revised' SUBJECT :

Thermal Design Procedure", for Review and Approval i

Reference:

1.

Chelemer, H., Boman, L.H., and Sharp, D.R., " Improved

~

Thermal Design Procedure," WCAP-8567-P (Proprietary) and VCAP-8568 (Non-proprietary), July 1975.

.l 1

Dear Mr. Lyons:

Enclosed are twenty-five (25) copies of the topical report, " Revised Thermal Design Procedure", WCAP-11397 (Proprietary).

The enclosed topical has been submitted to revise the Improved Thermal i

Design Procedure, Reference 1.

Our objective is to provide a more l

realistic prediction of the DNBR limit which satisfies the design f

criterion, by removing some of the conservatism in the Improved Thermal j

Design Procedure methodology. The procedure described in this report will be applied in our' standard reactor design methoaology and i

I referenced in future licensing' applications.

1 It is requested that the review of this topical be completed in the third quarter of 1987 so that Westinghouse can extend their analytical l

capabilities as the need arises, 1

~

'l l

1 l

i Mr. James Lyons Paga Two his submittal contains Westinghouse proprietary information of trade secrets, comercial, or financial infonnation which we consider priviledged or confidential pursuant to 10CFR9.5 (4). herefore, it is requested that the W:stinghouse proprietary information attached hereto be handled on a confidential basis and be withheld frcra public disclosure.

This material is for your internal use only and may be used for the purpose for which it is submitted.

It should not be otherwise used, disclosed, duplicated, or disseminated, in whole or in part, to any other person or organization outside the Office of Nuclear Reactor Regulation without the t.. press written approval of Westinghouse. Correspondence with respect to the Application for UithholdinE should reference AW-87-022, and should be addressed to R. A.

Wiesemann, Manager of Regulatory and Legislative Affairs, Westinghouse Electric Corporation, P. O. Box 355, Pittsburgh, Pennsylvania 15230-0355 Very truly yours, f

l i

W. J hnson, Manager Nu e Safety Department

/pj Enclosure (s) i t

t

Westinghouse Power Systems 8Minp,,,,s,,,33230c333 g

Electric Corporation March 16, 1987 AW 87-022 Mr. Herbert M. Berkov Standardization 6 Special Projects Branch D* vision of PVR Licensing-B U. S. Nuclear Regulatory Commission Washington, D.C.

20555

Reference:

LETTER JOHNSON TO LYONS, NS-NRC-87-3209, DATED MARCH, 1987

SUBJECT:

VCAP-11397, " Revised Thermal Design Procedure" The subject report transmitted by the ret 3renced letter contains information proprietary to the Westinghouse Electric Corporation.

The material will not be employed as a part of a license application or other action identified in 10CFR2.790(a) at this time.

It will be separately submitted with an Application for Withholding accompanied by an Affidavit meeting the requirements of 10CFR2.790(b) prior to such use.

Accordingly, we request that the material be treated as proprietary information within the provisions of 10CFR9.5(4), " Freedom of Information Act Regulations".

If there is a need to make public i

disclosure of the material prior to a separate Vestinghouse submittal for docket in accordance with the provisions of 10CFR2.790(a), please notify Westinghouse prior to making a disclosure determination.

Correspondence with respect to the proprietary aspects of this submittal should reference AW-87-022 and should be addressed to the undersigned.

VeryN.truly yours, fI ) '.

si

• l!lGlualuu Ro ert A.

iesemann, Manager Regulatory 6 Iegislative Affairs cc:

E. C. Shomaker, Esq.

Office of the General Council, NRC

\\

i i

5 4 m

A 5 4

4 4

m.

e_4aJ-a

.al,-=

e.--~---

4+-

Me 5-.

p 4

i Al%4e

-.4-.

e of b

F l

l 1

SECTION D 3

i i

i i

t 4

i k

i o

4 i

i 1

i l

l l

b i

1 4

1 t

I t

6 s

N i

5 i

i I

i i

I s

r

)

i I

i a

p

-9

=t w

e m-s&-

i 1

ABSTRACT

)

A Revised Thermal Design Procedure (RTDP)is developed which satisfies the design i

criterion that protects against Departure from Nucleate Boiling (DNB) in a PWR core.

Variations in plant operating parameters, nuclear and thermal parameters, fuel fabrication I

I parameters, and DNB correlation predictions are considered statistically to obtain a DNB i

uncertainty factor. Applying this factor leads to a limiting DNBR value to be used for accident analysis. Since the uncertainties are all included in the uncertainty factor, the l

accident analysis is done with input parameters at their nominal or best estimate values.

l RTDP revises the previous procedure, called the improved Thermal Design Procedure (ITDP),

in that DNB correlation uncertainties are combined statistically with the ITDP uncertainties instead of being treated separately. This provides a more realistic prediction of the DNBR limit which satisfies the design criterion.

The mathematical relationships are derived and a sample calculation is presented using numerical values for a 3 loop plant with 17x17 standard rod array fuel assemblies, and the i

WRB-1 DNB correlation. This Revised Thermal Design Procedure retains the capability of f

readily and realistically accommodating additional parameters which affect DNB or changes in the values or uncertainties of parameters.

B 5

?

)

i f

2 h

I sasst e-escus i

a

-~u a

__a r

t i

ACKNOWLEDGEMENTS i

The authors would like to thank E. H. Novendstern for guiding this project H. Chelemer for his helpful comments on the statistical methods, and J. R. Reid who performed many of the calculations. The discussions with R. C. Anderson and K. L Basehore of Virginia Power are also greatly appreciated.

l l

I I

l

~ !

- i l

l li 54m e-eso2n

TABLE OF CONTENTS SECTION TITLE PAGE 1.

INTRODUCTION 1

2.

MATHEMATICAL RELATIONSHIPS 2

3.

SAMPLE CALCULATION 9

4.

FUEL ROD BOVV 13 5.

SENSITIVITY FACTORS 14 r

6.

CONCLUSIONS 17 7.

NOMENCLATURE 18 8.

REFERENCES 19 LIST OF TABLES TABLE TITLE PAGE 3-1 Design Limit DNBR Using ITDP 10 3-2 Comparison of Design Limit DNBR's 12 5-1 Evaluation of Linearity Assumption 16 r

LIST OF FIGURES FIGURE TITLE PAGE 2-1 Illustrative Comparison of RTDP with ITDP 8

.i i

u m.-.. m.

iii

SECTION 1 i

INTRODUCTION A Revised Thermal Design Procedure (RTDP) is presented in this topical report for purposes of realistically predicting the Departure from Nucleate Boiling Ratio (DNBR) design limit in l

Westinghouse PWR's. This procedure removes some of the conservatism in the improved Thermal Design Procedure (ITDP) methodology [1]* while satisfying the design criterion that protects against DNB in a PWR core. The DNB thermal design criterion is that the probability that DNB will not occur on the most limiting fuel rod is at least 95% (at a 95%

confidence level) for any Condition i or 11 event.

i With ITDP methodology, system uncertainties are statistically combined separately from the DNB correlation uncertainty. The two are then combined directly, rather than statistically, to determine the DNBR limit.

The system and correlation uncertainties are independent variables, and in the revised l

procedure they are statistically combined to more realistically predict the DNBR limit by

{

removing some of the unnecessary conservatism in ITDP. The RTDP is a natural extension of the ITDP and is similar to the approach developed by Virginia Power [2]

1 i

The development of the mathematical relationships for the revised procedure is given in Section 2 and a sample plant analysis is shown in Section 3. Section 4 describes the application to fuel rod bow, and the justification for the use of sensitivity factors is presented in Section 5.

I 1

J t

I t

i

+

i Superscripts in brackets refer to list of References l

i 4

um e-mus 1

1 y

v py

-ig wei-

,+--

-a,2 44

SECTION 2 MATHEMATICAL RELATIONSHIPS RTDP is essentially an extension of ITDP. The n.athematical relationships used in ITDP are l

derived in the first part of this section. The second part shows the extension of these relationships for RTDP.

2.1 ITDP Methodology The following is a summary of the methodology used in ITDP.UI For a DNB correlation such as WRB-1 the statistical parameters of the data base are obtained: mean (mM/P), and standard deviation (sM/P), where M/P is the ratio of measured-to-predicted heat fluxes at the point of minimum DNBR. When ITDP is used, the DNBR Correlation Limit (CL) is set so that with 95% confidence there is at least a 95%

probability that DNB will not occur for a statepoint with DNBR 2 CL CL is given by I

I CL =

(2-1)

• M/P - Ks /P H

where K is obtained from tables prepared by Owen #I and is a function of the confidence I

level, the probability, and the number of data points in the DNB data set.

in order to relate the variations in design parameters to DNBR variations, an uncertainty factor, y, defined by the following equation, is used:

y = DNBR(variable)/DNBR(nominal)

(2-2)

The value of DNBR(nominal) is determined by considering the values of all the design parameters to be at their nominal or best estimate values. The value of DNBR(variable) is I

based on values of the design parameters including their uncertainties and deviations from

')

nominal values. Consequently, in any particular application, DNBR(nominal) will have a single determinable value while DNBR(variable) will be a random variable.

The DNBR uncertainty factor is considered to be affected by changes in the values of the design parameters according to a relation of the form m

(M d1 2 3

+3

+,,, 3*

y 1 x 2

x 1

2 m

m:.a-mus 2

i

I i

where th x;

is the value of the i design parameter, dx is the differential change in the value of xy, g

dy is the differential change in y resulting from the differential changes dx.

)

g The factor s, represents the sensitivity factor associated with the i parameter. If all the

'f th parameters in Equation (2-3) are held constant except for one, then it is clear from Equation j

(2-3) that if the x; are independent i

f 4

j

= B_Z i

a(in y) 2-0

=

3 i

i y

x$

a(in x$)

i i

Thus the value of s; can be interpreted as representing the percentage change in DNBR l

resulting from a one percent change in xy, all other parameters being held constant.

Integrating Equation (2-3), considering the s values fixed, and taking antilogarithms gives i

2 i

s s1 5

y=Cx1 x2

• • • *m (2-5) where C is obtained from the constant of integration.

i 1

in order to evaluate the uncertainty factor to be used in the design value of the DNBR, it is-necessary to obtain a relationship between it and the uncertainties in the design parameters used to determine DNBR.

Consider each of the independent design parameters x; as being distributed about a mean the following expression is obtained value u;. If y is expanded in a Taylor's series about the pg 1

(xy uy) + gf (x u2) + ***

~

(*m~"m}

(

}

i.

yu p

x y

1 2

m i

+ higher order terms 1

The partial derivatives in Equation (2-6) are evaluated at the point where all the x; are at their mean values u. The value of y at this point is represented by p.

g y

i i

9058LS-33C3:a 3

l i

t From Equation (2-5) i 5 2

... v,\\

(24) p = C ul 1 y

"2 If the pertubations from the mean values are small, the higher order terms in Equation (2-6) will be considerably smaller in magnitude than the first order terms and can be ignored.

Under these conditions, the variance of y determined using Equation (2-6) results in the f

following expression 2, (ay )2,2 (ay )2,2

,,,, (ay )2,m 2

(2-8) oy ax ax 2

y 2

ax, Using Equation (2-5) and (2-7) in Equation (2-8) leads to the equation I

+

...s

(

)

(2-9)

s (

)

+ s

(

)

(

)

y 1

2 m

i The ratio c/u is called the coefficient of variation. Equation (2-9) enables the coefficient of variation of the DNBR uncertainty factor y to be determined in terms of the sensitivity I

factors s, defined by Equation (2-4) and coefficients of variation c /ug of the design 3

parameters x used in evaluating DNBR.

j The central limit theorem of statistics indicates that the probability distribution function for j

y will approach a normal distribution with mean p and standard deviation e even if the l

y y

individual distributions of the x are not normal. It should be noted that Equation (2-9) is -

y subject to the restrictions that the x; are independently distributed and that the variations in the x can be considered small. In addition the sensitivity factors s are considered to be y

g constant, thus independent of the x;.

I In order to satisfy the DNB thermal design criterion, an ITDP DNBR design limit value DL, is' determined such that the probability that CL, the Correlation Limit DNBR (given by Equation (2-1) ) is exceeded is 95% with 95% confidence. The governing variables are considered to be at such levels that with each at its mean value the DNBR value on the peak power rod is DL. This results in the following relation for tho' design limit DNBR:

g DL;=I-If645o (2-10) y sasu e-nom 4

I

where the values 1 and 1.645 represent the mean value of y and the standardized normal variate corresponding to a 95% probability, respectively.

&2 RTDP Methodology RTDP utilizes the DNB correlation statirtical characteristics, m and sM/P, and the gjp uncertainty factor statistical parametsrs, v and c, calculated in the same manner as in the y

y ITDP.

The statistically combined system and correlation design limit DNBR for RTDP (DL ) is R

selected such that for a statepoint with mean DNBR at the DL, there is a 95% probability R

that the DNBR(variable) for the limiting fuel rod exceeds the correlation P/M(variable) with 95% confidence. DNB will not occur if

(

) (a,c)

(2-11)

Using Equation (2-2) gives i

[

] (a,c)

(2-12)

Rearranging Equation (2-12) results in

[

] (a,c) -

(2-13)

RTDP uses a parameter z defined by z =[

](a,c)

(2-14)

L

]Ifzis (a,c) enpanded in a Taylor's series about the mean value, the following expression is obtained (a,c)(2-15)

The partial derivatives in Equation (2-15) are evaluated at the point where each variable is at its mean value. The value of z at this point is represented by u.g l

i mn e-emn 5

From Equation (2-14)

[

] (a,c)

(2-16)

If the perturbations are small, the higher order terms in Equation (2-15) will be considerably smaller in magnitude than the first order terms and as a result can be ignored This being 5]

the case, the variance of z determined using Equation (2-15) results in the following expression (2-17)

(a,c)

[

l

)

(a,c)

(2-18) l (a,c) 1

[

l (a,c)

Using Equation (2-14) and (2-16) in Equation (2-17) leads to the equation (2-19)

(a,c)

E

~3 (a,c)

From Equations (2-13) and (2-14), with[

] If this is the case (a,c)

[

] (a,c)

(2-20)

(2-21)

. (a,c) 6 u sat 6-ee:22s

Substituting Equation (2-16) in (2-21) and noting that

[

](a,c)

(2-22) results in DL =

(2-23) p t

- (a,c)

Figure 2-1 schematically illustrates the calculation of the design limit DNBR using RTDP compared with that using IT0D lt should be noted that Equation (2-19) is subject to the restrictions that the x;[

Jare (a,c) independently distributed and that the variations in the x;[

]can be considered (a,c) small. In addition, the sensitivity factors s are considered to be constant and independent of the x;.

l Sant s-esens 7

Figure 2-1 ILLUSTRATIVE COMPARISON OF RTDP WITH ITDP FOR A DNB CORRELATION WITH vM/P = 1.0 AND LIMIT DNBR = 1.17 ITDP ITDP Desien Correlation Distribution Parameter Distribution bi /

1.0 1.17 Design Limit DNBR DNBR ITDP=[

] (a,c)

RTDP Combined Correlation and ITDP Design Parameter Distribution 1.0 Design Limit DNBR DNBR RTDP=[

]

l l

i l

Sasu e-seesis 8

SECTION 3 SAMPLE CALCULATION A representative plant will be analyzed using both ITDP and RTDP to illustrate the dif5rence between them. The selected plant is a three-loop plant with 17x17 standard fuel. Nominal plant operating conditions are listed in Table 3-1.

A sensitivity study was performed using the THINC-IV computer program.IO' l Sensitivities l

of DNBR to changes in plant parameters were determined for a range of statepoints l

covering various operating conditions with minimum DNBR values near the expected DL '

R The most limiting statepoint was selected as the one for which the sensitivities resulted in the highest DL. The resulting sensitivities are shown in Table 3-1.

R Plant parameter uncertainties are also listed in Table 3-1. These are determined on a plant-specific basis.

For the WRB-1 correlation, the statistical analysis of the data base results in [3]

(a,c)

When iTDP is used, the Correlation Limit obtained from Equation (2-1) is 1.17 CL

=

The ITDP analysis for the representative plant using the sensitivities calculated above is given ir Table 3-1. The resulting DNBR design I:mit values from Equation (2-10) are:

[

] (typical cell)

(a,c)

DL

=

g

[

] (thimble cell)

(a,c)

DL;

=

sam e-asens 9

TABLE 3-1 DESIGN LIMIT DNBR USING iTDP Typical Cell Thimble Cell Nominal or 2

Parameter o

c /tj S

S( )

S 5( )

7 F

Power 100%

T 553.6*F in Pressure 2200 psia Flow 100%

Bypass 0.965 F

1.49 2;

rin.i THINC 4 1.0 Transient Code 1.0

{

0 0 2 2=

E 5

( p, )

=

p I

y DNOR Design Limit

=

1 - 1.645 (O /p )

Y Y 1.17

=

=

L j

1 - 1.645 (0 /p )

(a,c) y y 5458F $-990328

-i 4

When RTDP is used, the analysis in Table 3-1 gives for a typical cell, j

[

]

(a,c)

Rom Equations (2-18) and (2-19):

i I

i.;

(a,c) i From Equation (2-23),

l DLR*

- (a,c)

Similarly, for a thimble cell, t

[

]

(a,c) i (a,c)

I DLR=

~

(a,c)

The DNBR design limits using RTDP are compared to those using ITDP in Table 3-2.

I

,I i

i

-f i

[

i

)

TABLE 3-2 COMPARISON OF DESIGN LIMIT DNBR'S Design Limit DNBR ITDP RTDP

~

~

Typical Cell Thimble Cell (a,c) i h

f 64 set e-asc32s g

i i

SECTION 4 FUEL ROD BOW I0l Rod bow is accounted for with current methodology by a correlation based on reactor data which relates S, (the upper 95% tolerance limit for the standard deviation of channel closure for the worst span) and burnup. This is combined with a relation between DNBR penalty and channel closure, a Monte Carlo type calculation being performed to generate the DNBR penalty statistics. The DNBR penalty is combined statistically with the correlation uncertainty and the results are used to calculate a limit DNBR. The percent difference between this limit DNBR and the limit DNBR with no rod bow is the rod bow penalty.

With the new methodology, the analysis is performed in the same way except that for both the unbowed and bowed cases, the DNBR correlation uncertainty is statistically combined with the plant parameter and other uncertainties. This results in a slight decrease in the rod bow penalty compared to the previous methodology.

A sample calculation was performed for 17x17 standard fuel, high flow conditions, at the rod bow limiting burnup of 24,000 MWD /MTU,Ibl using the WRB-1 correlation. With the previous methodology, the rod bow penalty was [

](a,c). With the new methodology, for the plant conditions evaluated in Section 3, the rod bow penalty was [

](a,c),

I l

s

}

l 3

i I

m n e-enc 2:s 13

SECTION 5 1

SENSITIVITY FACTORS Sensitivity factors were deterrnined over a wide range of statepoints covering various operating conditions with minimum DNBR values near the expected DL. The most limiting R

statepoint was selected as the one for which the sensitivities resulted in the highest DL '

R This leads to conservative results in terms of setting the limits on the core operating conditions determined by the DNB limitation.

This procedure differs from that described in Reference 1, in which a range of DNBR values was covered and the largest numerical value of each sensitivity factor over the DNBR range of interest was chosen for use in the DNB analysis.

Since RTDP extends the ITDP methodology in that the DNB correlation uncertainties are statistically combined with the ITDP uncertainties, the following evaluations were necessary with regard to this new methodology:

(a) Verification of the adequacy of Equation (2-3) in predicting changes in DNBR.

(b) Verification of the adegoacy of the linearity approximation made to obtain Equation (2-17).

The above evaluations were performed by using procedures previously evaluated by the NRC in Reference 10 and are discussed below.

As a check on the form of Equation (2-3), a pair of test cases was executed using the 1HINC-IV program. One case used values of the design parameters at nominal conditions while the second case used values at extreme conditions. For the latter case, significant changes were made in all of the parameters in which uncertainties were being considered in the DNB design procedure.

The ratio of the minimum DNBR's for the two cases was 0.80 for both typical and thimble cells as determined by the THINC program. Alternatively, the minimum DNBR for the second case can be estimated from the value for the first case by a relationship obtained by integrating Equation 2-3 assuming that the sensitivity factors, s, are constant, and taking g

the ratio of the results as applied to the two cases:

• m s (y1*1)s(y22)s x

(5-1) y 2

(v ) m y_

y m

y met e-sec22:

34

i i

where the y, and x; represent the parameter values from the first and second cases, f

and s gave values for y/p of respectively. Substituting the numerical values for the x, pg g

y g

0.76 (typical cell) and 0.77 (thimble cell). These represent the calculated ratio of the minimum DNBR's assuming the parameters are related to the DNBR by Equation (2-3) with constant s values. Comparing these values (0.76 and 0.77) with the values obtained from

]

g comparing the two THINC cases (0.80 and 0.80) indicates that the assumptions used in j

developing Equation (2-3) are conservative. Therefore, no uncertainty allowance is required for Equation (2-3) with the sensitivity factors given in Table 3-1 for the WRB-1 correlation i

f with RTDP.

i A linearity approximation was made to obtain Equation (2-9), the equation used to determine the statistical parameters for the DNBR uncertainty factor from the variation in the design parameters. There is a corresponding linearity assumption in the derivation of f

Equation (2-17).

i e

This approximation was tested by choosing the value of each of the parameters to be one i

standard deviation from its nominal value (in the direction leading to a decrease in DNBR)

{

and using Equation (5-2) to determine the combined edect on the DNBR.

?

(5-2)

(a.c) i I

Note that no linearity relationship is used in this calculation. The calculation results in a z value which is 0.74 times the nominal value for a thimble cell.

l 1

F The same combined effect is also determined using Equation (2-15) neglecting terms of i

4 second order and higher, which considers the relationship between the DNBR uncertainty factor and the design phrameters to be linear. This leads to l

f i

(5-3)

{

(a,c)

The value of 0.70 is obtained for the ratio of the resulting z to the nominal value.

The comparisons were repeated at several other values of the parameters (i.e.,2 c/2,2 o, f 2 c) and the results are given in Table 5-1. In every case, the linearity assumption gave more I

conservative results than the calculation which did not assume linearity. This confirms that i

the linearity assumption is somewhat conservative.

i I

um s-asen 15

TABLE 5-1 EVALUATION OF LINEARITY ASSUMPTION r

z/p g

Assumed Deviation Equation (5-2) of each x; Value Equation (5-3)

(Does not assume From Mean (Assumes Linearity)

Linearity)

Typical:

1/2 o adverse 0.85 0.86 beneficial 1.15 1.16 1e adverse 0.70 0.74 beneficial 1.30 1.34 2o adverse 0.41 0.54 beneficial 1.59 1.79 Thimble:

1/2 e adverse 0.86 0.87 beneficial 1.14 1.15

~

1o adverse 0.72 0.75 beneficial 1.28 1.32 2e adverse 0.43 0.56 beneficial 1.57 1.74

• z is calculated using values of parameters which deviate from the nominal value by the amount specified in the first column.

u, is calculated using nominal values of parameters.

sasn e-eee22 16

SECTION 6 CONCLUSIONS Based on the results of this study, the following conclusions were reached; o

The Revised Thermal Design Procedure given here satisfies the design criterion that protects against Departure from Nucleate Boiling in a PWR core.

o The procedure described in this report will be applied in the standard reactor design process and referenced in future licensing applications.

l l

l 1

4 l

i 1

uuve-eeane

(

l

.a F

SECTION 7 NOMENCLATURE C

Constant of integration CL Correlation limit DNBR DL; Design limit DNBR using ITDP Di R Design limit DNBR using RTDP DNB Departure from nucleate boiling DNBR Ratio of the expected DNB heat flux to the actual local heat flux K

Owen's Factor for 95% probability with 95% confidence M/P Ratio of measured-to-predicted heat fluxes in DNB test at the point of minimum DNBR m

Mean of measured-to-predicted heat flux ratio in DNB data set M/P m

Number of design parameters affecting DNBR uncertainty factor n

Sample size s

Standard deviation of measured-to-predicted heat flux ratio in DNB data set M/P th s;

Sensitivity factor of i parameter th i

parameter x;

y DNRR uncertainty factor = DNBR(variable)/DNBR(nominal) g

[

](a,c) u Mean

]

o Standard deviation i

usate-eeen 18

k SECTION 8 i

REFERENCES 1.

Chelemer, H., Boman, LH., and Sharp, D.R., *lmproved Thermal Design Procedure,"

WCAP-8567-P (Proprietary), July 1975 and WCAP-8568 (Non-proprietary), July 1975.

t 2.

Anderson, R.C., " Statistical DNBR Evaluation Methodology," VEP-NE-2, July 1985.

3.

Motley, F. E., et al., "New Westinghouse Correlation WRB-1 for Predicting Critical Heat Flux in Rod Bundles with Mixing Vane Grids,* WCAP-8762-P-A (Proprietary), July 1984 and WCAP-8763 (Non-proprietary), Ju!y 1976.

4.

Owen, D.B., " Factors for One-Sided Tolerance Limits and for Variable Sampling Plans,"

SCR-607, March 1963.

5.

Jaech, J.L., " Statistical Methods in Nuclear Material Control," TID-26298,1973.

6.

Chelemer, H., Chu, P.T., and Hochreiter, LE., "THINC-IV - An improved Program for Thermal Hydraulic Analysis of Rod Bundle Cores," WCAP-7956, (Non-proprietary), June 1973.

7.

Hochreiter, LE., and Chelemer, H., " Application of the THINC-IV Program to PWR Design," WCAP-8054 (Proprietary) and WCAP-8195 (Non-proprietary), September 1973.

8.

Skaritka, J. (Ed.),

• Fuel Rod Bow Evaluation," WCAP-8691, Rev.1 (Proprietary), and WCAP-8692, Rev.1 (Non-Proprietary), July 1979.

9.

Letter dated 6/18/86, C.H. Berlinger (NRC) to E. P. Rahe, Jr. (Westinghouse), " Request for Reduction in Fuel Assembly Burnup Limit for Calculation of Maximum Rod Bow Penalty" t

10.

Letter dated 4/19/78, D. F. Ross, Jr. (NRC) to C. Eicheidinger (Westinghouse),* Staff Evaluation of WCAP-7956, WCAP-8054, WCAP-8567, and WCAP-8762".

f 1

i se.at e-ssene 3g

a-

... ~

4 i

, - 1

. I SECTION E' P

l i

=

f f

k e

)

i I

1 1

1 I

1 I

[

l p

b r

.i I

1 P

f I

i i

f 6

4 1

t l

. I

' l 1

1 1

1

.- i I

a l

I 4

F

-m

Bm 355 Westinghouse Power Systems e,,333u, rennsri,,,,,5230 0355 Electrk: Coqxnation June 13, 1986 NS-NRC-88-3346 U. S. Nuclear Regulatory Commission ATTN:

Document Control Desk Washington, D.C.

20555 ATTN:

Mr. Marvin W. Hodges, Reactor Systems Branch Chief Division of Engineering & System Technology

SUBJECT:

Responses to NRC Questions on WCAP-11397, " Revised Thermal Design Procedure" [Non-Proprietary]

REFERENCE:

(1) Letter from M. W. Hodges (NRC) to W. J. Johnson (h'),

(Questions on WCAP-11397), " Revised Thermal Design Procedure" dated April 4, 1988.

Dear Mr. Hodges:

Enclosed are:

Twelve (12) copies of WCAP-11397, Addendum 1, " Responses to Additional Questions on Revised Thermal Design Procedure" [Non-Proprietary).

The enclosed information is being submitted in response to additional NRC questions, Reference (1), as a result of the Staff's and their consultants review of the subject Topical.

Very truly yours, V

J hnson, Manager Nc

. Safety Department

/mit Enclosure

r f

NCAP-11397 Addendum 1 o

t REVISED THERMAL DESIGN PROCEDURE A. J. Friedland S. Ray June 1988 Approved:

NM E. H. Novendstern, Manager Thermal-Hydraulic Design And Fuel Licensing I

HESTINGHOUSE ELECTRIC CORPORATION Commercial Nuclear Fuel Division P. O. Box 3912 Pittsburgh, Pennsylvania 15230 5458L:6-880614

RESPON3E TO NRC QUESTIONS ON UCAP-ll397,

' REVISED THERMAL DESIGN PROCEDURE" 1.

Dplain sint is cxxxt by DM6w1&le) arti P/H(swi&1e) in de Ekv1sai1ha:ral Lbsig1 P rrrd rre (FAP) daelqrtnt en pge 5 4pttutly they are hxh nrxin unubles, hr it is ur?nr which scutces d uxxrtaitty are 2rrh dai in axh. Ibes de DM(1w151e) ally aztam de urettaimy htzn i

de dsigt.mmer~s fcr a given txt: ital se d exxxiitias? Ibes PA!(tw1&1e) cnly attain uratury han de an:v?xten for a ginn inmral set d exxx!iticns? Whr are de cxms a11 ~

st:xtinti cLvirirrs (wh m ccircticn) fcr axh d these crxizn swibles? Is de "(wrible)"

desadptar asstrixi en de P$1's axi&P's in nynHens (2-12) thtug1 (2-14)?

RESPONSE TO OUESTION 1:

In the development on page 5:

DNBR(variable) only contains the uncertainties from the design parameters t

for a given nominal set of conditions.

P/M(variable) only contains uncertainty from the correlation for a given nominal set of conditions.

l y is defined as DNBR(variable)/DNBR(nominal).

y has a mean of 1.0 (p - 1.0) and a standard deviation designated a.

y y

The reciprocal of P/M(variable) is M/P(variable), with mean and gpp standard deviation ogjp.

The "(variable)" descriptor is assumed on the P/M's and M/P's in equation (2-12) through (2-14).

l 1

4 j

i 2.

Itu is z in m eim (2-14) irrapntai? It dxs rrt gpur to be " min *1 av11q1, n to y, Judi sas 7

aistmtairty fantr Is z Jsc a 11ER nextra variable cmtainity bod 1 cztreieim an design pararzter trnrtainties?

RESPONSE TO OUESTION 2:

Z is defined by equation (2-14) and is a DNBR random variable containing both correlation and design parameter uncertainties, as inferred in the question.

3.

De agressim in m e fm (2-18) is ckscribal as a1 tger 95% cmfi&rre 1 Lit a1 de startuti deviefm nm + red wid1 de trmtninty in de axreir1m 1 beer, T/1.E45 is in gar:ral la:yr dm (dC/d11-aput)%.S, de epprpriat.e mir1 plier to ctodn as tgtr 958 cmfi&rre limit. De c.o cpttitles am ay11 dm de tite mm of P/P is kxm (Judi is rxt de cace). Still, de tse af T/1.f>45 ;1utks Ecr de nvisai drig1 lictit Q apaing de amirim licit (C2.) utm dnm is to trartairty in de dsigt pmrtecs is das de rmarz for cbiig dxc was cire in qwim (2-18)7 Or ucs it to alicu fcr timtire de tne wriax:e cf PS as trinw? In a1y case, plase cylain de ratianle fcr mesm (2-18) azi de umdity dxcriptim RESPON'SE TO OUESTION 3:

i The term K/1.645 is used in equation (2-18) to account for the uncertainty in the mean of M/P as well as the uncertainty in the standard deviation of M/P.

The resulting o is that for a normal distribution which has the same gpp upper 95t tolerance limit as is given by Owen's table.

~

For known p and a, upper 95% tolerance limit (UTL) - p + 1.645o For sample E and s, UTL - Y + ks For equal UTL's y + 1.645a - E 4 ks

,, E-

+ ks 1.645 Y s - o if p - E 1.645

i 4.

As trtal in de pwias qtu;tkn, a2 "qpr a:nfidwe limit" is phrwt at sigm(t'yP). Pmn:ibly ttas ms ire dan to sigm(y) heare ic is beirg tzread as &. C1mtly, it is trc lerw1 awcly Neem > c::tr,wHm crthxh (Wkh av ggtrimte arri rmie artain asstrptias) wrv usa! to derive it. Also, t!xcv a:e sf neims than the hxiividn1,mx vmiax:es cay mc be justifi&1y t1stal as icxuz. De Guiwi ci sigm(y) as latu2 laxis to Q but sardurg diffemt v1dd be cbaalini if sigm(y) mre tmtal as trictm Planse discuss azi styret yur hwui af sigm(y),

RESPONSE TO OUESTION 4:

is treated identically to the way it is treated in ITDP (WCAP-8567).

The o

NRC review (Ref.10 of WCAP-11397) of UCAP-8567 states:

"The proposed design basis is that there must be at least a 95%

probability that the minimum DNBR of the limiting power rod is greater than or equal to the DNBR limit of the correlation being used; parameter uncertainties or variances obtained from the evaluation of data are determined at a 95% confidence level.

Implementation of the errors in correlations or procedure involves assumptions that postulated functions are random variables where repeated use of a correlation with identical conditions gives identical results. Also, distributions of uncertainties for variables such as power, flow or r

temperature are not well known so that the functional forms of the uncertainty distributions must be assumed.

Therefore, a rigorous statistical statement of the type implied in the proposed DNBR design basis cannot be obtained.

i "However, Westinghouse has either chosen distribution functions which are typical of observed distributions or which give conservative variances.

Also, biases in correlations have been reduced insofar as practical.

Therefore, although a rigorous statistical statement cannot be made, the Vestinghouse method provides a reasonable approximation to such a statistical statement."

For RTDP, the design parameter variances will be determined on a plant specific basis by the identical procedure currently used for ITDP.

l

'I k

B i

5.

He last sataxe cf de thst pa&1 cf Smciaa l stem du dsign critir1at dat is de basis fcr tte RIP dimmrIin tte ah im1, De "95% pthbility at 95% cmfidstd' ;rrtiat af de cdteria, is t}T m 117 icpla:mtni via tse cf a au-si&d tolante incenal h1gd a1 tte rnmn1 distrih rim A l

d.x:n;tia1 cf a 95%B5% tolenrre intenal a111xu it diffets f1an a 95% cmfidra utmul is S rm i

i in de hyrdix.

t b

A tolaum krmal is a stata:xit ant a sirgie diswhrIm. Aldog1 de MIP figxe ad a;tia1 at de toc:m cipp B agtst dnt de MIP nxird is cmskkrity a siigle distrihtia, dur arbines de awir!m ad dsign pu:rucer trm: curries, tte develcp:mt cn pp 6 sans to irdicte cdnruise. SpcilIrn11y, Eq. (2-18) arprec a tole:ure htmal for de "awikim turaui:ty" distrih rim. J1ile Eq, (2-31) is de nsult cf a 95% antidra ittcnul at de exrbitxd distrih tim wtxte a "toletare htenal k:flataf' mrizm is usai for de contbrjat tsmtzirry. 1his CP l

zyndt is simil.x to, br rxt cpite de sxo as utir swid be cbcaitxd if a 95%B5% tolante brenul uns ctrytni on de ctrbimi distrihria, tsxt:r de aarpcia, ttat sign (y) is kxut vidar eaar.

Plaase aqda:n tte raciaule for tte MIP develgrrit stqs nfamsi to dxxe, azi c9 ain Irv it l

I urIcfles de "958 pthbility at 95% cxrfidstd' pxtlas ef de asign critadat fia:t Sa:tia,1.

RESPONSE TO OUESTION 5:

The approach used in RTDP is the satte as that in ITDP with M/P included among the independent variables.

The response to question 4 includes the NRC comments on the statistical approach used in ITDP.

6.

Dim mias dragar de a h*rn1 iq11y the MKT) axi s%T) are en1~1*al crer de ser of.%T m1us auwyniiry to t!e dra tsai to cheky de crzzelatim 21.o anctnts are curle :elathe to d1is.

i i

(a) It is higptpdate to cxrpte a uma axi scxxhid &virim in etnis fastua1 tr11 css it ca1 be I

d.uuauani dut de t$9 whxs are a zuxin sxple Exon a amn,rrwineIm. Das vill ton!ly tre be de ase far a gism am1 rim axi its auwyniirg d1tn base axi ngia, cf qphr+111ty. If tte ftf vahrs d> rrt are f:o:t a ctrr:m ;rwhHm, de nun ad statind i

dwirim im m:y sig2Lfiancly ener mdas stingias cf de nyial cd im?lthilfry. (Dds vill nipixe ctrptity diffenre safety limits in diffamt sixtgias, cr drrsi:g tte latyst safety limit if de Jule ngfas is to be ctswd tsirg cnly a sirgie m1ue.)

,ee

(b) Ic in iray,rtpriate to emhmm arf rarne themLw.w d a cw7eim taftg tte sre dra tad to dewl:p it, P ally if spiriod fittirg tatrifg = su:n as last squarts a:e used.

f To. fauna marts (e.g., anus ad staczked dwiatlas d Pft) dadd be at:puted fitzn dra tot tad to dtw.kp the amfefm If azh " extra" dca are tre asafftle, chta-splittfrg or crtss-ualf+ rim tainfges dudd be tad to emhnm arzf carne puf-w, with tte resultirg mus axf starkrd dwiatian tad in ~1m1#frg the desig211mic.

Alth;qh this ah ier#1 amrs psm:a1 cedrxinqy, tte due scirts ast be rury11zaf dm qplyiry tte utdtxtdqy to specific n-r'1fmeims. P1mse the m hv ytur qptah ftr applyirg tte R1TP xm res far tte o.o Issues raisaf due.

RESPONSE TO OUESTION 6:

he DNB correlation statistics are the same as those which were used to develop the DNBR correlation limits in the previous ITDP methodology which have been accepted by the NRC.

For the URB-2 correlation, for example (Ref.

6-1), the NRC SER presented a statistical analysis of the test data and concluded that while the data do not all represent a single population, the error introduced into variance estimates by assuming that the data are all from the same population is negligible.

Similar results are obtained for the WRB 1 correlation (Ref. 3 of UCAP-11397) applied to "R" grid data.

The evaluation of DNB correlation is handled in separate topicals, as in the above examples.

The RTDP derivation treats the data as coming from a common population, and assumes that the error introduced by this assumption is negligible as shown in the appropriate correlation topical report.

(Ref. 6-1, UCAP-10444-P-A, " Reference Core Report - Vantage 5 Fuel Assembly," September 1985)

1 7.

1he first puyst ci onIm 5 (pse 14) is rit "wiely cimr :q;udity fru yu prpre to dxahs de smsitivity fxtors Ib yu prpze daziry as de msitivity Crms dise rulas aswquxibg to de litdtiry sv'1" int (de av vid2 de hfgust Q. as sMial at pse 14)7 1his sm:s to te a numlaie agnuds (S nn det cryp swqnut:s are ecriral), hr anEhets with de 62scri; tim cf i

sanitivity fants ghm in %ctim 2.1. De pee 3, de msitivity fax:xs w.tv da::ribai as beby pa> Op dxrys in DG zwirfry fzon 18 days in tte x,. Sirre to rarias is td cf de msitivity fanns dqxndkg en de s, at tte exx1rirfm, de W14e#Im is dat dey are iniprrirr cf de x aid tfe cxx2rlefm Nere, de 4prar cxrfilce vida de pqrsal cn gge 14 j

Plate clarify yur ytynal far Aminity de msitivity Erms irr a ghm W fcrim, ircludity hv yu vill cin-re de nr.ber ax! 1errim cf state;oints to be cx1shk::ui.

E 1ain dy das 9

apprai2 is a1 iq1tve:xst cur de qpuzh mtlani in de sauxi pg41 cn pge 14 RESPONSE TO OUESTION 7:

We will choose as the sensitivity factors those values corresponding to the limiting statepoint (the one with the highest DL. 85 SPeCified n Page 14).

R The statement on page 3 that the sensitivity factors are percentage changes in DNBR resulting from 1% changes in the x, applies at the limiting g

statepoint.

The sensitivity factors do depend somewhat on the xy and the correlation, so some of them may be higher at some other statepoints.

However, as noted in Refs. 1 and 10 of UCAP-11397, the observed variations in sensitivity factors are generally small over a wide range of conditions.

Selecting the sensitivities at the limiting statepoint is sufficient to assure a conservative DL '

R As is currently done for ITDP, sensitivity factors are determined for various statepoints over a range of conditions for which the RTDP methodology will be applicable.

With RTDP, the sensitivity factors at each value.

The highest of these is used

,statepoint are used to calculate a DLR as the design limit DNER. This approach removes some unnecessary conservatism from the approach mentioned in the second paragraph on page 14, which combines worst sensitivities from different statepoints.

y b

B.

7he disamskn hicre ard after eqtxrin (5-1) dnIs with a sirs e innstipcial cf de valifity cf l

unim apticn Q-3). in Sunal, "p:uf by ecrple" is dmrtus, vfally Jun m1y me ccuple is umd. Pnsu:ubly mults anId differ far diHaur aatre?ebr, diffmrt whs of de anv72im hpr ws: Wies, diffaut dsigi para: rte:s, ad diffe:ur valas cf de drig2 pz:nmte s.,Astify nu:hn trxier cchar yssible ciranstmoes?

dat (2-3) is a anse:vrive w RESPONSE TO OUESTION 8:

As noted on page 14, the procedure used to ' check the form of Eq. (2-3) was the same as that previously evaluated by the NRC in Ref. 10.

The form of the equation is based on observation and experience with the THINC computer program and the Westinghouse DNB correlations, which generally show small variation in the S over a wide range of conditions.

The numerical example which was made at nominal plant conditions is considered to be a check rather than a proof of the methodology.

A similar check was made with one test case at the limiting statepoint conditions where the sensitivities resulted in the highest DL and the second case with all design parameters R

one standard deviation from the limiting statepoint conditions in the direction leading to a decrease in DNBR.

The :ombined effect on DNBR of these deviations is considerably greater than 1.645 times the RMS of the individual DNBR standard deviations.

If Eq. (5-1) were exact with constant S then it would be expected to agree exactly with the THINC analyses, since g

the S were evaluated near the limiting statepoints.

This was found to be g

the case with both the THINC analyses and Eq. (5-1) showing the limiting statepoint ratio of perturbed to nominal DNBR to be 0.834 for the typical cell and 0.847 for the thimble cell.

Similar evaluations will be performed if there are any changes to the DNBR correlation or in the limiting i

r statepoint.

i 9.

In nynHm (5-3), drul6t't anh cf de s(x-nu) tet::s be dividai ty de aua.,udirg nu?

RESPONSE TO OUESTION 9:

Yes.

The equation was misprinted and should read:

+ S I# ~#

+

... + Sm(xm m) + (M/P)

(5-3) y 2 2 2)

_Z - S (x -pg) 1 2

1. m M/P

1. z

P P

h The ceiculated results in Table 5-1 are correct.

The misprint will be corrected in the approved version of the topical.

10. De dim mias axi arpriscns en pys 15-if> cf ICAP-1D97 am wmui with investptiig tk akspry cf de Dnar egysu las in wy eims Q-6) axi Q-15). As in quasdat 8, w h:re I

a:rxxens dx1r de sificiary cf ytw in&m'im. In this case ycu did cmsidr sural saltus ed i

(cr etxrgs in) de x, hr all are fx de saw sec cf'desig2 ;-

w (with a find set cf ams 3

axi sortimi dv!*1ms). Platse Jsdfy yew cxxvincfm dat de liint sqzrtnincia2s are akunw azi cxramadve fx all pzsible diffetut arbuntias cf dsigs,m' rr tm.

)

El ONSE TO OUESTION 10:

The cases tested are those suggested in the NRC evaluation of WCAP-8567 (Ref. 10 of WCAP-11397). They provide a confirmatory check that the linear approximation is reasonable and conservative.

1 It can be shown that, in general, if all the deviations are in the direction leading to a decrease in DNBR, and all S are numerically greater than or g

equal to unity as is generally true, then the linear approximation is i

conservative.

This is shown by expanding 2 in a Taylor's series about the p and including second order terms.

g l

The values of the variables which are major contributors to. defining the lower 95% limit of Z/p will be mostly in the direction of negative deviations of Z from p. Therefore, it is reasonable to expect the linear i

approximations to be conservative.

D. De stinittal ctxs rre di.nns hu de gsnial adrrh1qy vill be L.ple:ntal in specific elim*ims. CIarly de sesitivity f:rrrrs rad to be ckrmmird fer axf2 rxv armlefm, set cf dsigt,m'-vrems, cr nqp ci c1fckI1Ity. Ic is roc clar dedxr it vill be rarswry to dxxk the varias am qrims axi sgprtxnr:eims cf de F2P ftr an32 9 mtIm. Firully, de almitral 1I ctxs tre d.m m hu de asigt pmrxte v.rlans vill le &tmmrud Plaxse drm m dxse Im v e nptdiry de 9 mrIm cf de & F2P adixtaqy to spelfic c1Imrims. De reviens 1I runi to kw Judxr it is Jsc de grvral cxdub1qy dxt is Lt1kr xxviiv, cr if de L"planzreim ad wrificeim y ftr.gxcific c11mesms a-e also to be uviamVegzzwed.

)

i

4 RESPONSE TO OUESTION 11:

The general methodology for RTDP will be implemented in the same manner as for ITDP.

Sensitivity factors will be determined for each new correlation, set of design parameters, or range of applicability. The design parameter variances will be determined on a plant specific basis by the identical i

procedure currently used for ITDP.

12. He first-crder e:2tr ptpgrJai frm17a (2-8) assums dat de nrd:m vadahles x,.., x are 1

m j

7 r&&z. M1Lle das =~"r_Im uns nxstJavd in de a h-irt=1, its rmstrabianss

[

sr=~I etm ?1 i

uns rcc dim &, P1mse pt@kr axh a disenssia1, cakfig idenme to spcific desig: proriters l

ati givirg imms W tie j~1 --,ws am artnfly briprzire i

13. P1mse clar1fy yt1r s n*mnt a1 pge 4 dat "De cerral limit dxxim d _e#seIm in4 form dat i

de ptiddlity distr,hrIm itsrcia2 fcr y vill mittxxix a trr-n1 disti h eim vida cxm cu(y) ad stzrxicd 6v1stiat sig:o(y) em if de irdivhini distriheims d de s, am tre rocal." Clearly de cerral lhdt tturra is roc 4p7Imh7e to ayn~fm (2 5). Are yt12 hashg d11s stataurt at de

=m vim d (2-5) beiry #%cerly ww unki by (2-6)7 Alm, shre x, are rrt 16srsmily distritucal, a axe restrictiw itx1o d tte cerral licit ttutzm ast te c'Idini.

Fkn11, tte 7

crnegrre to r-7 f ry d a 1sure crzbirxclai d nrrizn wzridles is &parirt a2 de nz6er d variah'n axi tivir hx!ivhini distr,hr1ms P1mse exrsidr all d de a'xne purts in cladyirs ad l

anxxt:rg yc1r smtemts extranirg de cetral Licit dutzm axi a1 ww:..uw racm1 distrIh efm ftr y.

RESPONSE TO OUESTIONS 12 AND 13:

Questions 12 and 13 deal with the ITDP development, which has been covered in the NRC review and evaluation of WCAP-8567 (Ref. 10 of WCAP-11397).

The RTDP development is identical to the ITDP development in these areas.

i

14. He ckw7gmr d de reef.saf dsig) life g assu:rs dxe ewry dsig przr:rter is diserthraf l

vith a cara vahe egal to its racral w21ue. Will d1is abeys be de ase, cr is it pzsible the l

biases cigt scist? If m, hw d:es de R:IP hrrile de hias az1 hw u2dd g darge?

l

1 RESPONSE TO OUESTION 14:

4 Biases might exist in design parameters.

In such cases, the bias is included separately in the DNBR analysis.

The bias is not included indevelopmentofDg.

)

l l

I L

I I

i 4

l

~!

t i

i I

l 1

1 b

i

L.

A k

.7 r

(3.h.

1 e..

l

."A.

2 f" 4

+

.'.,!['

e 9

\\

\\

en.

3gi z

y n-t d

i,L

.k,/ \\

I l';j,.

Tg

[: :9s?f' w

r_,

x

. x,g,

,7, [', L

I ;bf' g$

.p

~. y-^ '; '

p i

.- )

- ).' f,'

'u.

f a

i. ;; o

,,y.,

i a,

f.

m

+

g 3

['i

1

';;,.( '['

~

j

_ ]

?.I0i.<'

,$d?)

y

'"?]f,I'

>y.

4 a.,,

(

6

.,j yb.9

,,LJ

.'.6 ; -

s.,

l-.'

w s,

t V kf

[jf' _

L A.

A o

n i

1 A

\\

..y

(

. w

'\\

.-v

~

)(y Ipi,

. :1 j

1 m

v' t

I g

.<t.*,,

wl$,'

','4--

r Westinghouse d

Commercial Nuclear Fuel Division-7'V*').

J &

P.O. Box 355 1 Wg.. ~

.:r,.

l Pittsburgh, PA 15230 0355-SC q

4 1

Wi:STINGHOUSE PROPRIETARY CLASS 2 WCAP-11397-P-A l

REVISED THERMAL DESIGN PROCEDURE

^

i A. J. Friedland S. Ray Original Version: February 1987 Approved Version: April 1989 Approved:

E. H. Novendstern, NFD Manager, Thermal-Hydraulic Design and Fuel Licensing WESTINGHOUSE PROPRIETARY DATA This document contains information proprietary to Westinghouse Electric Corporation; it is submitted in confidence and is to be used solely for the purpose for which it is furnished and returned upon request. This document and such information is not to be reproduced, transmitted, disclosed or used otherwise !a whole or in part without authorization of Westinghouse Electric Corporation, Commercial Nuclear Fuel Division.

WESTINGHOUSE ELECTRIC CORPORATION Commercial Nuclear Fuel Division P.O. Box 3912 Pittsburgh, Pennsylvania 15320 O/9M/,-I.??OC:;.,bC:Tp.

$456L6-690328 p.

.. - =. -

WI5iJGH0!;3~ FRf4RE!aRY Ci>% 2

?

WCAP-11397-P-A TABLE OF CONTENTS

-i

)

i Section Description A

NRC Acceptance Letter for Referencing-of l

Licensing Topical Report WCAP.-11397, j

" Revised Thermal Design Procedure" l

B NRC Safety Evaluation (SER)

{

C Westinghouse submittal letter to NRC for_

,j WCAP-11397-P

'l D

WCAP-ll397-P-A Text E

Response to NRC Questions on WCAP-11397

" Revised Thermal Design Procedure

[Non-Proprietary]

i i

6 i

k 4

f t

7 I

i I

h

.I i

l l

.i

WESTINGHOUSE PROPRIETARY CLASS 3 3

SECTION A i

I r

3 1

I i

i 5

l t

u I

i i

'k l

f i

l l

t 9

5 t

WESITCH0 TEE Mt0PR!ETART CLASS 2 f cecoq'o, UNITED STATES NUCLEAR REGULATORY COMMISSION

~

g( g( )g WAsmNGTON, D. C. 20C55 gg..v{f JAN1 7 1933 g, ;7 g Mr. W. J. Johnson, Manager Nuclear Safety Department Westinghouse Electric Corporation Box 355 Pittsburgh, PA 15230-0355

SUBJECT:

ACCEPTANCE FOR REFERENCING OF LICENSING TOPICAL REPORT WCAP-11397, REVISED THERMAL DESIGN PROCEDURE" We have completed our review of the subject topical report submitted by Westinghouse by letter dated March 16, 1987. We find the report to be acceptable for referencing in license applications to the extent specified and under the limitations delineated in the report and the associated NRC evaluation, which is enclosed. The evaluation defines the basis for acceptance of the report.

We do not intend to repeat our review of the matters described in the report and found acceptable when the report appears as a reference in license applications, except to assure that the material presented is applicable to the specific plant involved..Our acceptance applies only to the matters described in the report.

In accordance with procedures established in NUREG-0390, it is requested that Westinghouse publish accepted versions of the report, proprietary and non-proprietary, within 3 months of receipt of this letter. The accepted versions shall incorporate this letter and the enclosed evaluation between the title page and the abstract. The accepted versions shall include an -A (designating accepted) following the report identification symbol.

Should our criteria or regulations change such that our conclusions as to the acceptability of the report are invalidated, Westinghouse and/or the applicants referencing the topical report will be expected to revise and resubmit their respective documentation, or submit justification for the continued effective applicability of the topical report without revision of their respective documentation.

Sincer y,

M

./

Ashok 'C. Thadani, Assistant Director for Systems Division of Engineering & Systems Technology Office of Nuclear Reactor Regulation

Enclosure:

Topical Report Enclosure i

WESTINGHOUSE PROPRIETARY CLASS 2 SECTION B i

i 1

s t

i t

F t

?

I f

l l

i

i v;ESTNMUSE PROPMETARY CLASS 2 SAFETY EVALUATION BY THE OFFICE OF NRR RELATING TO TOPICAL REPORT WCAP-11397 REVISED THERMAL DESIGN PROCEDURE WESTINGHOUSE ELECTRIC CORPORATION 1.

INTRODUCTION i

The subject topical report WCAP-1397, describes a revised themal design procedure (RTDP) for predicting the departure from nucleate boiling ratio (DNBR) design limit in Westingnouse pressurized water reactors (PWRs). This is a modification of the existing improved themal design procedure (ITDP) methodology (Ref.1) which has been reviewed and approved by the NRC (Re' ?).

As with the ITDP, the RTDP is also based on consideration of system uncertainties in plant operating parameters, fabrication paramet_ers, nuclear and thermal parameters, and the use of the appropriate departure from nucleate boiling (DNB) correlation for the plant.

The DNB design basis remains the same, namely, that there must be at least a 1

95 percent probability at a 95 percent confidence level that the limiting power fuel rod will not experience DNB during normal operation and anticipated operational occurrences.

Parameter uncertainties or variances obtained from the evaluation of data are detemined at a 95 percent confidence level.

With the ITDP methodology, these system uncertainties were statistically combined separately from the DNB correlation uncertainty. The two were then combined directly rather than statistically to determine the DNBR limit. The proposed RTDP methodology would combine the system and correlation uncertainties statistically rather than deterministically. This is similar to the statistical DNBR evaluation methodologies developed by the other PWR vendors and approved by the NRC.

The staff review encompassed the original submittal as well as responses to staff requests for additional information (Ref. 3). The staff was assisted in this review by our con ultants at Pacific Northwest Laboratories.

MsivE00SE POPmETARY Cl32 2.0 SUft!1ARY OF TOPICAL REPORT The topical report describes the mathematical relationships used in both the approved ITDP and the proposed RTDP and presents a sample calculation of a representative plant using both methods to illustrate the difference between-them.

In addition, a description of how fuel rod bow is accounted for in both methodologies is given and the effect on rod bow penalty is compared.

Finally, a discussion as to how sensitivity factors are determined for various statepoints over an appropriate range of conditions is given.

3.0 EVALUATION The existing ITDP method of protecting against DNB in Westinghouse pressurized water reactors was reviewed extensively by the NRC and a staff evaluation was issued in 1978 (Ref. 2).

Because the ITDP resulted in a large reduction in DNBR margin as compared to the traditional method, referred to by Westinghouse as the fixed value method, the staff examined all the parameters to assure that all uncertainties had been appropriately considered. The general methodology for the RTDP will be implemented in the same manner as for the ITDP. Sensitivity factors will be determined for each new correlation, set of design parameters, or range of applicability. The design parameter variances will be determined on a plant specific basis by the identical procedure currently used for the ITDP.

However, since the RTDP extends the ITDP methodology in that the DNB correlation will be statistically combined with the system uncertainties, the adequacy of the relationship of the DNBR uncertainty factor to changes in the values of the design parameters given by dy/y =

S dx /x g

g g

was evaluated. The procedure used to check the validity is the same as that previously used and evaluated by the NRC and was based on test calculations performed by Westinghouse using the THINC-IV computer program. These test results indicate that, with the sensitivity factors used, the above relationship

MSTm90 TEE PROPR!ETARY Ct4SS 2 '

provides a conservative model for changes in DNBR, as calculated by THINC-IV, for small changes in parameter values. Therefore, no uncertainty allowance is required for this equation with the sensitivity parameters used in the topical report. However, if sensitivity factors change as a result of correlation changes or changes in the application or use of the THINC-IV code, the uncertainty allowance must be re-evaluated.

A linearity approximation was made in the ITDP to obtain Ecuation (2-9) in the topical report. This equation is used to determine the statistical parameters for the DNBR uncertainty factor from the design parameters. There is a corresponding linearity assumption in the derivation of Equation (2-17) in the report for the RTDP which must be validated.

In investigating the adequacy of the linear approximations, Westinghouse compared results both with and without the assumption of linearity. The cases tested were those suggested by the NRC in the evaluation of the ITDP (Ref. 2).

In every case, the linearity assumption gave more conservative results than the calculation which did not assume linearity.

Based on this, the staff feels it is reasonable to expect the linear approximations to be conservative and, therefore, acceptable.

In the existing ITDP, fuel rod bow is accounted for by a correlation which relates the upper 95 percent tolerance limit for the standard deviation of channel closure for the worst span and burnup. This is combined with a relation between DNBR penalty and channel closure. The DNBR penalty is then combined statistically with the CHF correlation urcertainty to calculate a limit DNBR. The rod bow penalty is the percent difference between this limit DNBR and the limit DNBR with no rod bow. The analysis in the proposed RTDP methodology is performed in the same way except that the DNBR correlation t

uncertainty is statistically contined with the plant parameter and other uncertainties for both the unbowed and bowed cases. This results in a slight decrease in the rod bow penalty compared to the ITDP methodology. For example, for 17 x 17 standard fuel at the rod limiting burnup of 24 GWD/MTU, the ITDP resulted in a rod bow penalty of 1.1 percent whereas the RTDP resulted in a rod bow penalty of 1.0 percent using the WRB-1 correlation. The staff considers

o WESW3 HOUSE PROMt!EIAN CUS32 e the effect of the stell difference to be negligible and, therefore, finds the rod bow treatment described in the topical report acceptable.

4.0 STAFF POSITION The RTDP procedure for calculating DNB limits, as presented in WCAP-11397, is acceptable for use in licensing applications.

It provides a reasonable approximation to the proposed statistical basis. As with the existing ITDP, however, certain restrictions must be imposed on the implementation of the method because of the sensitivity of the method to changes in the correlations and codes used. These restrictions are:

1.

Sensitivity factors used for a particular plant and their ranges of applicability should be included in the Safety Analysis Report or reload submittal.

2.

Any changes in DNB correlation, THINC-IV correlations, or parameter

[

values listed in Table 3-1 of WCAP-11397 outside of previously demonstrated acceptable ranges require re-evaluation of the sensitivity factors and of the use of Equation (2-3) of the topical report.

3.

If the sensitivity factors are changed as a result of correlation changes or changes in the application or use of the THINC code, then the use of an uncertainty allowance for application of Equation (2-3) must be re-evaluated and the linearity assumption made to obtain Equation (2-17) of the topical report must be validated.

4.

Variances and distributions for input parameters must be iustified on a plant-by-plant basis until generic approval is obtained.

5.

Nominal initial condition assumptions apply only to DNBR analyses using RTDP. Other analyses, such as overpressure calculations, require the appropriate conservative initial condition assumptions.

^

BMWMUX PROPMEDlp C1 ASS 2

~5-6.

Nominal conditions chosen for use in analyses should bound all permitted methods of plant operation.

7.

The code uncertainties specified in Table 3-1 ( 4 percent for THINC-IV and 1 percent for transients) must be included in the DNBR analyses using RTDP.

The statistical method as presented includes no explicit design margin to accommodate unknowns. Such a margin could reduce or eliminate the impact of core related problems which are discovered after a core is designed and after a plant is operating. Although no particular margin is quantified, margin is inherent in the overall procedure used with the revised thermal design procedure. This margin is available to offset the effects of yet-to-be-discovered design problems. However, if newer procedures are proposed which substantially reduce thermal margin, then a design margin to accommodate unknowns should be explicitly identified.

The parameter ranges do not cover the range required for part loop operation.

If the method is to be used for analysis of part loop operation, the topical report must be amended to cover this wider range.

5.0 REFERENCES

1.

Chelemer, H., Bowman, L. H., and Sharp, D.R., " Improved Thermal Design Procedure, "WCAP-8567-P, July 1975.

2.

Letter from D. F. Ross, Jr. (NRC) to C. Eiche1dinger (W), " Staff Evaluation of WCAP-7956, WCAP-8054, WCAP-8567, and WCAP-8762," April 19, 1978.

j 3.

Friedland, A.

J., and Ray,.S., " Revised Thermal Design Procedure,"

WCAP-11397, Addendum 1, June 1988.

I

WESTINGHOUSE PROPRIETARY CLASS 2 SECTION C t

i 1

f T

I b

I-I q

i n

L i

i t

I

o Westinghouse Power Systems p%3 Gen rennsrivaniais230-0355 Electric Corporation NS-NRC-87-3209 March 16, 1987 Mr. James Lyons, Chief Technical & Operations Support Branch Office of Nuclear Reactor Regulations U.S. Nuclear Regulatory Commission Washington, D.C.

20555 ATTENTION:

Document Control Desk ATTDITION:

Carl H. Berlinger, Chief Reactor Systems Branch Division of PWR Licensing-A SUBJECT Submittal of Westinghouse Topical, WCAP-il397, " Revised Thermal Design Procedure", for Review and Approval

Reference:

1.

Chelemer, H.,

Boman, L.H., and Sharp, D.R.,

" Improved Thermal Design Procedure," WCAP-8567-P (Proprietary) and WCAP-8568 (Non-proprietary), July 1975.

Dear Mr. Lyons:

Enclosed are twenty-five (25) copies of the topical report, " Revised Thermal Design Procedure", WCAP-ll397 (Proprietary).

The enclosed topical has been submitted to revise the Improved Thermal Design Procedure, Reference 1.

Our objective is to provide a more realistic prediction of the DNER limit which satisfies the design criterion, by removing some of the conservatism in the Improved Thermal Design Procedure methodology. The procedure described in this report will be applied in our standard reactor design methoBology and referenced in future licensing applications.

It is requested that the review of this topical be completed in the third quarter of 1987 so that Westinghouse can extend their analytical capabilities as the need arises.

• ?ttrJH0t!SE PRCf9EDU Ga358 Pr. James Lyons Pago Two 7his submittal contains Westinghouse proprietary infonnation of trade secrets, commercial, or financial infonnation which we consider priviledged or confidential pursuant to 10CFE9.5 (4). Therefore, it is requested that the Westinghouse proprietary information attached hereto be handled on a confidential basis and be withheld frcrn public disclosure.

This material is for your internal use only and may be used for the purpose for which it is submitted. It should not be otherwise used, disclosed, duplicated, or disseminated, in whole or in part, to any other person or organization outside the Office of Nuclear Beactor Regulation without the express written approval of Westinghouse. Correspondence with respect to the Application for Uithholding should reference AW-87-022, and should be addressed to R. A.

Ulesemann, Manager of Regulatory and Legislative Affairs, Westinghouse Electric Corporation, P. O. Box 355, Pittsburgh, Pennsylvania 15230-0355 Very truly yours, W. J hnson, Manager Nu. e Safety Department

/pj Enclosure (s) i

M M M m mtauCtz 2 Westinghouse Power Systems

$l$7 p,,,,,, m3g g33 Dectric Corporat. ion March 16, 1987 AU-87-022 Mr. Herbert M. Berkow Standardization & Special Projects Branch Division of PWR Licensing-B U. S. Nuclear Regulatory Commission Washington, D.C.

20555

Reference:

LETTER JOHNSON TO LYONS, NS-NRC-87-3209, DATED MARCH, 1987

SUBJECT:

WCAP-ll397, " Revised Thermal Design Procedure" The subject report transmitted by the referenced letter contains information proprietary to the Westinghouse Electric Corporation.

The material will not be employed as a part of a license application or other action identified in 10CFR2.790(a) at this time.

It will be separately submitted with an Application for Withholding accompanied by an Affidavit meeting the requirements of 10CFR2.790(b) prior to such use.

Accordingly, we request that the material be treateu as proprietary information within the provisions of 10CTR9.5(4), " Freedom of Information Act Regulations" If there is a need to make public disclosure of the material prior to a separate Westinghouse submittal for docket in accordance with the provisions of 10CFR2.790(a), please notify Westinghouse prior to making a disclosure determination.

Correspondence with respect to the proprietary aspects of this submittal should reference AW-87-022 and should be addressed to the undersigned.

Very truly yours, A

l}

i 3 i L.

.f, b& &ldlO U Ro ert A. kiesemann, Manager Regulatory & Legislative Affairs cc:

E. C. Shomaker, Esq.

Office of the General Council, NRC i

.a 6

WESTINGHOUSE PROPRIETARY CLASS 3 i

SECTION D r

i

. i.

l 1

WESTINGHOUSE PROPRIETARY C2. ass 2 ABSTRACT A Revised Thermal Design Procedure (RTDP) is developed which satisfies the design criterion that protects against Departure from Nucleate Boiling (DNB) in a PWR core.

Variations in plant operating parameters, nuclear and thermal parameters, fuel fabrication parameters, and DNB correlation predictions are considered statistically to obtain a DNB uncertainty factor. Applying this factor leads to a limiting DNBR value to be used for accident analysis. Since the uncertainties are all included in the uncertainty factor, the accident analysis is done with input parameters at their nominal or best estimate values.

RTDP revises the previous procedure, called the improved Thermal Design Procedure (ITDP),

in that DNB correlation uncertainties are combined statistically with the ITDP uncertainties instead of being treated separately. This provides a more realistic prediction of the DNBR limit which satisfies the design criterion.

The mathematical relationships are derived and a sample calculation is r

,ented using numerical values for a 3 loop plant with 17x17 standard rod array fuel cssemblies, and the WRB-1 DNu correlation. This Revised Thermal Design Procedure retains the capability of readily and realistically accommodating additional parameters which affect DNB or changes in the values or uncedainties of parameters.

t 5458L 6-890328 i

WESTINGHOUSE PROPRIETARY CLASS 2 ACKNOWLEDGEMENTS The authors would like to thank E. H. Novendstern for guiding this project, H. Chelemer for his helpful comments on the statistical methods, and J. R. Reid who performed many of the calculations. The discussions with R. C. Anderson and K. L Basehore of Virginia Power are also greatly appreciated.

L i

i 5458L'6-890328 Ii

WESTINGHOUSE PROPRIETARY CLASS 2 TABLE OF CONTENTS SECTION TITLE PAGE 1.

INTRODUCTION 1

2.

MATHEMATICAL RELATIONSHIPS 2

3, SAMPLE CALCULATION 9

4.

FUEL ROD BOVV 13 5.

SENSITIVITY FACTORS 14 6.

CONCLUSIONS 17 7.

NOMENCLATURE 18 8.

REFERENCES 19 LIST OF TABLES TABLE TITLE PAGE e

3-1 Design Limit DNBR Using ITDP 10 3-2 Comparison of Design Limit DNBR's 12 5-1 Evaluation of Linearity Assumption 16 LIST OF FIGURES FIGURE TITLE PAGE i

2-1 Illustrative Comparison of RTDP with ITDP 8

sen e-escias lii

i WESTINGHOUSE PROPRIETAQV C5. ASS 3 i

I SECTION 1 INTRODUCTION

' A Revised Thermal Design Procedure (RTDP) is presented in this topical report for purposes of realistically predicting the Departure from Nucleate Boiling Ratio (DNBR) design limit in Westinghouse PWR's. This procedure removes some of the conservatism in the improved Thermal Design Procedure (ITDP) methodology while satisfying the design criterion that protects against DNB in a PWR core. The DNB thermal design criterion is that the probability that DNB will not occur on the most limiting fuel rod is at least 95% (at a 95%

confidence level) for any Condition I or ll event.

With ITDP methodology, system uncertainties are statistically combined separately from the DNB correlation uncertainty. The two are then combined directly, rather than statistically, to determine the DNBR limit.

The system and correlation uncertainties are independent variables, and in the revised procedure they are statistically combined to more realistically predict the DNBR limit by removing some of the unneces:.ary conservatism in ITDP. The RTDP is a natural extension of the ITDP and is similar to the approach developed by Virginia Power [2]

The development of the mathematical relationships for the revised procedure is given in Section 2 and a sample plant analysis is shown in Section 3. Section 4 describes the application to fuel rod bow, and the justification for the use of sensitivity factors is presented in Section 5.

4 Superscripts in brackets refer to list of References 4

maa-mm 1

WESTIMGHOUSE PROPRIETARY CLASS 3 SECTION 2 MATHEMATICAL RELATIONSHIPS RTDP is essentially an extension of ITDP. The mathematical relationships used in ITDP are derived in the first part of this section. The second part shows the extension of these relationships for RTDP.

2.1 ITDP Methodology The following is a summary of the methodology used in ITDP.IN For a DNB correlation such as WRB-1[3] the statistical parameters of the data base are obtained: mean (mM/P), and standard deviation (sM/P), where M/P is the ratio of measured-to-predicted heat fluxes at the point of minimum DNBR. When ITDP is used, the DNBR Correlation Limit (CL) is set so that with 95% confidence there is at least a 95%

probability that DNB will not occur for a statepoint with DNBR 2: CL CL is given by 1

CL =

(2-1)

• M/F - Ksgjp where K is obtained from tables prepared by Owen and is a function of the confidence level, the probability, and the number of data points in the DNB data set.

In order to relate the variations in design parameters to DNBR variations, an uncertainty factor, y, defined by the following equation, is used:

y = DNBR(variable)/DNBR(nominal)

(2-2)

The value of DNBR(nominal) is determined by considering the values of all the design parameters to be at their nominal or best estimate values. The value of DNBR(variable) is based on values of the design parameters including their uncertainties and deviations from nominal values. Consequently, in any particular application, DNBR(nominal) will have a single determinable value while DNBR(variable) will be a random variable.

The DNBR uncertainty factor is considered to be affected by changes in the values of the design parameters according to a relation of the form dx dx dx d_.y = s y

2 m

2-3)

+s*

y 1 x 2 x 1

2 m

m ;. a -an 22.

2

WESTINGHOUSE PR'oPRIETARY CLASS 2 where th x;

is the value of the i design parameter, dx; is the differential change in the value of x;,

dy is the differential change in y resulting from the differential changes dx,.

th The factor s represents the sensitivity factor associated with the i parameter. If all the g

parameters in Equation (2-3) are held constant except for one, then it is clear from Equation (2-3) that if the x, are independent i

i a(1" Y) i = 82 (2-0

=

s 3

a(ln x$)

y x

Thus the value of s can be interpreted as representing the percentage change in DNBR j

resulting from a one percent change in x, all other parameters being held constant.

y Integrating Equation (2-3), considering the s; values fixed, and taking antilogarithms gives sy s2 s

y=Cx1 x2

'"*m (2-5) where C is obtained from the constant of integration.

j in order to evaluate the uncertainty factor to be used in the design value of the DNBR, it is necessary to obtain a relationship between it and the uncertainties in the design parameters r

used to determine DNBR, Consider each of the indepandent design parameters x; as being distributed about a mean i

value ug. If y is expanded in a Taylor's series about the ug the following expression is obtained V

(x ~"2) * ***

(*m'*m)

(2-6)

1) +

yU (X

V y

1 2

1 2

m

+ higher order terms The partial derivatives in Equation (2-6) are evaluated at the point where all the x; are at their mean values ug. The value of y at this point is represented by u.y seu e-nosa 3

WESTirJGHousE PROPRIETARY CLASS 2 l

From Equation (2-5) 5 s

u = C v1 1 y

"2 2... v,m (2-7)

If the pertubations from the mean values are small, the higher order terms in Equation (2-6) will be considerably smaller in magnitude than the first order terms and can be ignored.

Under these conditions, the variance of y determined using Equation (2-6) results in the following expression Y

+ (0Y )

ax o2

+...(3Y)

= (ax )

o (2-8) o o

y 2

ax, y

m 1

Using Equation (2-5) and (2-7) in Equation (2-8) leads to the equation

(

)

=s

(

)

+ s

(

)

+

...s

(

)

(2-9) y 1

2 m

The ratio o/u is called the coefficient of variation. Equation (2-9) enables the coefficient of variation of the DNBR uncertainty factor y to be determined in terms of the sensitivity factors s; defined by Equation (2-4) and coefficients of variation o /p; of the design i

g parameters x; used in evaluating DNBR.

The central limit theorem of statistics indicates that the probability distribution function for y will approach a normal distribution with mean u and standard deviation o even if the y

individual distributions of the x are not normal. It should be noted that Equation (2-9) is g

subject to the restrictions that the x; are independently distributed and that the variations in the x; can be considered small. In addition the sensitivity factors s; are considered to be constant, thus independent of the x;.

in order to satisfy the DNB thermai design criterion, an ITDP DNBR design limit value DL; is determined such that the probability that CL, the Correlation Limit DNBR (given by Equation (2-1) ) is exceeded is 95% with 95% confidence. The governing variables are considered to be at such levels that with each at its mean value the DNBR value on the peak power rod is DL;. This results in the following relation for the design limit DNBR:

1* 1 - 1 645 o (2-10) 54 sate-eeena 4

WESTINGHOUSE PROPRIETARY CLASS 2 where the values 1 and 1.645 represen'. the mean value of y and the standardized normal variate corresponding to a 95% probability, respectively.

2.2 RTDP Methodology RTDP utilizes the DNB correlation statistical characteristics, m and sM/P, and the M/P uncertainty factor statistical parameters, u and c, calculated in the same manner as in the y

y ITDP.

The statistically combined system and correlation design limit DNBR for RTDP (DL ) is R

selected such that for a statepoint with mean DNBR at the DL, there is a 95% probability R

that the DNBR(variable) for the limiting fuel rod exceeds the correlation P/M(variable) with 95% confidence. DNB will not occur if

[ DNBR (variable) 2P/M(variable)] (a,c)

(2-11)

Using Equation (2-2) gives

[(DL )

• y 2 (P/M)] (a,c)

(2-12) g Rearranging Equation (2-12) results in

[ (DL )

• V * (M/P) 2 1.0 ] (a,c)

(2-13)

R RTDP uses a parameter z defined by 2 =[(DL )

• y * (M/P)] (a,c)

(2-14)

R

[The analysis of z in the RTDP is comparable to the analysis of y in the ITDP. Consider each of the independent parameters y and M/P as being distributed about a mean value.] If 2 is (a,c) expanded in a Taylor's series about the mean value, the following expression is obtained (y u ) +

jg) (WP upp) + higher order terms (a,c)(2-15) zu = 7 y

The partial derivatives in Equation (2-15) are evaluated at the point where each variable is at its mean value. The value of z at this point is represented by u. z seat e-sems 5

WESTINGHOUSE PROPRIETARY CLASS 2 From Equation (2-14)

[v

• V /P] (a,c)

(2-16)

=(M)*U R

y M

7 If the perturbations are small, the higher order terms in Equation (2-15) will be considerably smaller in magnitude than the first order terms and as a result can be ignored. This being the case, the variance of z determined using Equation (2-15) results in the following expression [5],

(2-17)

M/P))

M/P_

(a,c)

I

)

+

z

)

(oM/P corresponds to the standard deviation associated with the DNB correlation being used (at an upper 95% confidence level) and is obtained as follows:]

(3,c)

(2-18)

K o /P

= s /P M

M N_

(a,c)

[where s is the standad deWadon associated wM the data set and K is W Owen's M/P factor from Reference 4. The mean associated with the correlation, pqfp, is taken as equal l

to the mean of the data set, mM/P. since K includes the effect of ti.w < ncertainties in both s

and m M/P M/P' (a,c)

Using Equation (2-14) and (2-16) in Equation (2-17) leads to the equation

( 2)2 = ([e) + ("M/P)

(2-19) 2 e

2-o z

y (a,c)

[where (c /p )2 is calculated from Equation (2-9).]

y

)

From Equations (2-13) and (2-14), with[mean DNBR at the DL, there is 95% probability that R

z 21 with 95% confidence.] If this is the case (a,c)

[v - 1.645 oz3 1 ] (a,c)

(2-20) 7 i

_ "z 1 1 - 1.645 a /u 7 z -

(a,c) m n -. m s 6

WESTINGHOUSE PROPRIETARY CLASS 2 Substituting Equation (2-16) in (2-21) and noting that

[p = 1.0 ] (a,c)

(2-22) y results in y

DL =

(2-23)

R

, M/P(1-1.645 o /p2) - (a,c) p g

Figure 2-1 schematically illustrates the calculation of the design limit DNBR using RTDP l

compared with that using ITDP.

It should be noted that Equation (2-19) is subject to the restrictions that the x;[and M/P]are (a,c)

Independently distributed and that the variations in the x [and in M/P]can be considered (a,c) g small. in addition, the sensitivity factors s; are considered to be constant and independent of the x;.

ses:t s-se:22:

7

WESTINGHous! PROPRIETARY CLASS 2 Figure 2-1 ILLUSTRATIVE COMPARISON OF RTDP WITH ITDP FOR A DNB CORRELATION WITH vM/P = 1.0 AND LIMIT DNBR = 1.17 ITDP ITDP Design Correlation Distribution Parameter Distribution 22 1.0 1.17 Design Limit DNBR DNBR ITDP =[1.34 ]

RTDP Combined Correlation and ITDP Design Parameter Distribution 1.0 Design Limit DNBR DNBR RTDP =[1.23] (a,c)

{

54sst e-soens 8

WESTINGHOUSE PQoPRIETARY class 2 l

SECTION 3 SAMPLE CALCULATION A representative plant will be analyzed using both ITDP and RTDP to illustrate the difference between them. The selected plant is a three-loop plant with 17x17 standard fuel. Nominal plant operating conditions are listed in Table 3-1.

A sensitivity study was performed using the THINC-IV computer program.IO'7I Sensitivities of DNBR to changes in plant parameters were determined for a range of statepoints covering various operating conditions with minimum DNBR values near the expected DL 'R The most limiting statepoint was selected as the one for which the sensitivities resulted in

)

the highest DL. The resulting sensitivities are shown in Table 3-1.

R Plant parameter uncertainties are also listed in Table 3-1. These are determined on a plant-specific basis.

For the WRB-1 correlation, the statistical analysis of the data base results in [3]

1.0079 m

=

M/P 0.0859 s

=

M/P 1108 (misprint in n

=

Reference 3 corrected) 1.724 from Reference 4.

(a.c)

K

=

When ITDP is used, the C,orrelation Limit obtained from Equation (2-1) is CL 1.17

=

The ITDP analysis for the representative plant using the sensitivities calculated above is given in Table 3-1. The resulting DNBR design limit values from Equation (2-10) are:

[1.352] (typical cell)

(a,c)

D L,

=

[1.339) (thimble cell)

(a,c)

D L,

=

5458L6-89C328 9

I l

llilllilI llll ll:l4 ll jl1 ll J

gC NOzOc n

hg< nQ$ N

)

c.

a

(

5 5 0 5

9 4 2 9

)

1 0-4 6 0 6 6 0 9 0 2

l

(

4 5 1

8 0 9

4 4 0

l 2

0 0 0 0 0 2

0 0 0

e S

0 0 0 0 0 0 0 0 0

C 0 0 0 0 0 0 0 0 0

e l

bm 6

6 9

i 0 2 7

3 h

S 2 0 3

1 1

6 5 3 3 2 2 0 O

0 3

T 1

0 2

2 0 1

1 1

1 1

i 1

)

2 7 3 9 8 3 8 0 5

l 0-0 2 4

1 0

2 5 4 0

0 7 0 2 0 2

5 7 1

(

0 3

0 0 0

2 0 0 0

1 le 5

0 0 0 0 0 0 0 0 0

C 0 0 0 0 0 0 0 0 0

la P

D ic 9

2 5

1 6

8 3

p 4 9 1

8 8 1

T y

2 4 8 4 4 3 2 O O

0 1

I T

2 1

1 1

0 5

2 7 1

1 1

G N

I S

U 1

=

R 3

)

)

B t

0 6 6 5 0 5 2 0 0

i Y

E N

0 3 6 2 6 4

8 0 5

m p

H 0

0 2 0 2

1 2

0 p

i

/

/

1 L

D 0 0 0 0 0 0 0 0 0

L Y

/

B 0

0 0

A T

0 O O 0 0 0 0 0 0

n

(

(

o 7

I T

5 5

1 i

M t

4 4

a 6

1 6

IL le 1

r N

r a

o G

5 %

C 1

I i

I F

s % 7 6

4 5

S p 5 7 3 2 2 0

=

=

5

. 5 0

0 O.

2 8

0 0

E 0

t 2 0 0

O 0{

D 1

2 1

1 2

i m

)

i L

i 0

(

nig s

e S

e r

t D

oa a

m E

R l

I

~

i F

s 3

a t

p h

6 5

9 D

n s 0% 6 4

0 O

=

i E m

0 3 8 0 9 O

ot 0 5 2 0 1

1 1

y N s 1

5 2

1 0

0 1

e B

e do C

r t

82 e

e n

3 t

r 4

le 09 e

u s

8 m

r s

s 1C s

6 a

e s w a

, N n

r w n e lo p HI a

F 8

a o

i r

y N EbH r

5 4

P P T P F B F

F T T

5

• o

-+

l l

lll

WESTINGHOUSE PROPRIETARY CLASS 2 When RTDP is used, the analysis in Table 3-1 gives for a typical cell,

[o/u = 0.0819 ]

(a,c) y y From Equations (2-18) and (2-19):

t

- = N'(.0819)2 + (1.0079

• I".T45)2 2

.0859 1.724 z

= 0.1212 (a,c)

From Equation (2-23),

-r DLR=

1

- 1.239'

,(1.0079) (1-1.645 x 0.1212)

, (a,c)

Similarly, for a thimble cell, i

[o/u =0.0766]

(a,c) y y t

g

=M(0.0766)2 (1.0079 x N )2

.0859 1.724 2

u g (a,c)

0.1177 DLp

1

= 1.230' (1.0079) (1-1.645 x 0.1177)

(a,c)

The DNBR design limits using RTDP are compared to those using ITDP in Table 3-2.

l f

ut, ate-asen 33

WESTINGHOUSE PROPRIETARY CLASS 2 P

TABLE 3-2 COMPARISON OF DESIGN LIMIT DNBR'S Design Limit DNBR' ITDP

RTDP, Typical Cell 1.352 1.239 -

Thimble Cell

. 1.339 1.230 _

(a.c)

I l

54S8L 6-89c328

)

r

tRfEsTlalGHouSE PROPRIETARY CLASS 2 SECTION 4 i

FUEL ROD BOW IOI Rod bow is accounted for with current methodology by a correlation based on reactor date which relates S, (the upper 95% tolerance limit for the standard deviation of channel

<.iosure for the worst span) and burnup. This is combined with a relation between DNBR penalty and channel closure, a Monte Carlo type calculation being performed to generate the DNBR penalty statistics The DNBR penalty is combined statistically with the correlation uncertainty and the results are used to calculate a limit DNBR. The percent difference between this limit DNBR and the limit DNBR with no rod bow is the rod bow penalty, i

With the new methodology, the analysis is performed in the same way except that for both l

the unbowed and bowed cases, the DNBR correlation uncertainty is statistically combined with the plant parameter and other uncertainties. This results in a slight decrease in the rod bow penalty compared to the previous methodology.

I A sample calculation was performed for 17x17 standard fuel, high flow conditions, at the rod bow limiting burnup of 24,000 MWD /MTU,IN using the WRB-1 correlation. With the previous methodology, the rod bow penalty was [1.1%)(a,c). With the new methodology, for the plant conditions evaluated in Section 3, the rod bow penalty was [1.0%)(a,c),

1 b

snea-es:ns 13 I

WESTitoGHoUSE PROPRIETARY CLASS 8 SECTION 5 i

SENSITIVITY FACTORS Sensitivity factors were determined over a wide range of statepoints covering various operating conditions with minimum DNBR values near the expected DL. The most limiting R

statepoint was selected as the one for which the sensitivities resulted in the highest DL

• R This leads to conservative results in terms of setting the limits on the core operating conditions determined by the DNB limitation.

This procedure differs from that described in Reference 1, in which a range of DNBR values was covered and the largest numerical value of each sensitivity factor over the DNBR range of interest was chosen for use in the DNB analysis.

Since RTDP extends the ITDP methodology in that the DNB correlation uncertainties are statistically combined with the ITDP uncertainties, the following evaluations were necessary l

with regard to this new methodology:

l (a) Verification of the adequacy of Equation (2-3) in predicting changes in DNBR.

i (b) Verification of the adequacy of the linearity approximation made to obtain Equation (2-17).

The above evaluations were performed by using procedures previously ' evaluated by the NRC in Reference 10 and are discussed below.

I As a check on the form of Equation (2-3), a pair of test cases was executed using the THINC-IV program. One case used values of the design parameters at nominal conditions while the second case used values at extreme conditions. For the latter case, significant changes were made in all of the parameters in which uncertainties were being considered in the DNB design procedure.

The ratio of the minimum DNBR's for the two cases was 0.80 for both typical and thimble l

cells as determined by the THINC program. Alternatively, the minimum DNBR for the second case can be estimated from the value for the first case by a relationship obtained by integrating Equation 2-3 assuming that the sensitivity factors, s;, are constant, and taking the ratio of the results as applied to the two cases:

L=

(b)$

(y22)3 h)s 1

2 m

(5-1) u y1 y

y m

mat e-mus 34

WEsTiftSGHouSE PROPRIETARY CLASS 3 where the u; and x; represent the parameter values from the first and second cases, respectively. Substituting the numerical values for the x, vy and s gave values for y/u of g

i y

0.76 (typical cell) and 0.77 (thimble cell). These represent the calculated ratio of the minimum DNBR's assuming the parameters are related to the DNBR by Equation (2-3) with constant s values. Comparing these values (0.76 and 0.77) with the values obtained from g

comparing the two THINC cases (0.80 and 0.80) indicates that the assumptions used in developing Equation (2-3) are conservative. Therefore, no uncertainty allowance is required for Equation (2-3) with the sensitivity factors given in Table 3-1 for the WRB-1 correlation with RTDP.

A linearity approximation was made to obtain Equation (2-9), the equation used to determine the statistical parameters for the DNBR uncertainty factor from the variation in the design parameters. There is a corresponding linearity assumption in the derivation of Equation (2-17).

This approximation was tested by choosing the value of each of the parameters to be one standard deviation from its nominal value (in the direction leading to a decrease in DNBR) and using Equation (5-2) to determine the combined effect on the DNBR.

s

~

- = (h) 1 (

)2

(

)m

)

(5-2) l

("M/P _(a,c)

. 2 "1

2 m

Note that no linearity relationship is used in this calculation. The calculation results in a z value which is 0.74 times the nominal value for a thimble cell.

The same combined effect is also determined using Equation (2-15) neglecting terms of second order and higher, which considers the relationship between the DNBR uncertainty factor and the design parameters to be linear. This leads to

~

(5-3)

~

s (x uy) + s (*2 "2) +

s (*

"m) + (M/P)/pgjp

=

1 y 2

m m y*

"1 "2

"m (a,c)

The value of 0.70 is obtained for the ratio of the resulting z to the nominal value.

The comparisons were repeated at several other values of the parameters (i.e., + c/2, + o, + 2 c) and the results are given in Table 5-1. In every case, the linearity assumption gave more conservative results than the calculation which did not assume linearity. This confirms that the linearity assumption is somewhat conservative.

um e-m2:s 15

WESTINGHOUSE PROPRIETARY Ct. ASS 2 i

TABLE 5-1 EVALUATION OF LINEARITY ASSUMPTION Z/VZ Assumed Deviation Equation (5-2) of each x; Value Equation (5-3)

(Does not assume From Mean (Assumes Linearity)

Linearity)

Typical:

1/2 o adverse 0.85 0 86 beneficial 1.15 1.16 5

1o adverse 0.70 0.74 beneficial 1.30 1.34 2o adverse 0.41 0.54 beneficial 1.59 1.79 Thimble:

1/2 o adverse 0.86 0.87 beneficial 1.14 1.15 1o adverse 0.72 0.75 beneficial 1.28 1.32 2o adverse 0.43 0.56 beneficial 1.57 1.74

• Z is calculated using values of parameters which deviate from the nominal value by the amount specified in the first column.

9, is calculated using nominal values of parameters.

I l

t,atst e-ascus 16 I

WESTINGHOUSE PROPRIETARY CLASS 2 l

SECTION 6 i

CONCLUSIONS Based on the results of this study, the following conclusions were reached; o

The Revised Thermal Design Procedure given here satisfies the design criterion that protects against Departure from Nucleate Bolling in a PWR core, o

The procedure described in this report will be applied in the standard reactor design process and referenced in future licensing applications.

}

7 W

t i

k I

1 i

i 54sn e-enc 22:

37 i

WESTINGHOUSE PROPRIETARY CLASS 2 SECTION 7 NOMENCLATURE C

Constant of integration CL Correlation limit DNBR D L, Design limit DNBR using ITDP p

t DL Design limit DNBR using RTDP l

R DNB Departure from nucleate bolling DNBR Ratio of the expected DNB heat flux to the actual local heat flux K

Owen's Factor for 95% probability with 95% confidence M/P Ratio of measured-to-predicted heat fluxes in DNB test at the point of minimum DNBR m

Mean of measured-to-predicted heat flux ratio in DNB data set M/P m

Number of design parameters affecting DNBR uncertainty factor n

Sample size s

Standard deviation of measured-to-predicted heat flux ratio in DNB data set M/P th s;

Sensitivity f actor of I paramster th x;

i arameter i

y DNBR uncertainty factor = DNBR(variable)/DNBR(nominal) z

[DNBR uncertainty parameter = DL

• y

• M/P](a,c) g r

y Mean t

o Standard deviation m ace-nesa 3g i

WESTINGHOUSE PROPRIETARY CLASS 3 SECTION 8 f

REFERENCES 1.

Chelemer, H., Boman, LH., and Sharp, D.R., " Improved Thermal Design Procedure,"

WCAP-8567-P (Proprietary), July 1975 and WCAP-8568 (Non-proprietary), July 1975.

2.

Anderson, R.C., " Statistical DNBR Evaluation Methodology," VEP-NE-2, July 1985.

3.

Motley, F. E., et al., "New Westinghouse Correlation WRB-1 for Predicting Critical Heat Flux in Rod Bundles with Mixing Vane Grids," WCAP-8762-P-A (Proprietary), July 1984 and WCAP-8763 (Non-proprietary), July 1976.

4.

Owen, D.B., " Factors for One-Sided Tolerance Limits and for Variable Samplirig Plans,"

SCR-607, March 1963.

5.

Jaech, J.L " Statistical Methods in Nuclear Material Control," TID-26298,1973.

6.

Chelemer, H., Chu, P.T., and Hochreiter, L.E., "THINC-IV - An Improved Program for Thermal Hydraulic Analysis of Rod Bundle Cores," WCAP-7956, (Non-proprietary), June 1973.

7.

Hochreiter, L.E., and Chelemer, H., " Application of the THINC-IV Program to PWR Design," WCAP-8054 (Proprietary) and WCAP-8195 (Non-proprietary), September 1973.

8.

Skaritka, J. (Ed.), " Fuel Rod Bow Evaluation," WCAP-8691, Rev.1 (Proprietary), and WCAP-8692, Rev.1 (Non-Proprietary), July 1979.

9.

Letter dated 6/18/86, C.H. Berlinger (NRC) to E. P. Rahe, Jr. (Westinghouse), " Request for Reduction in Fuel Assembly Burnup Limit for Calculation of Maximum Rod Bow Penalty" T

10.

Letter dated 4/19/78, D. F. Ross. Jr. (NRC) to C. Eicheldinger (Westinghouse)," Staff Evaluation of WCAP-7956, WCAP-8054, WCAP-8567, and WCAP-8762".

54SBL 6-690328

)g l

a 3

m L

9 WESTINGHOUSE PROPRIETARY CLASS 2 i

SECTION E.

i r

9 I

-i

[

[

I i

i f

S k

0 b

P f

5 e

b 4

f

pV=,

v) mama Westinghouse PowerSystems

$35]p penn,,n,, m3a g333 Sectnc Corporat. ion June 13, 1986 NS-NRC-88-3346 U. S. Nuclear Regulatory Commission ATTN:

Document Control Desk Washington, D.C.

20555 ATTN:

Mr. Marvin W. Hodges, Reactor Systems Branch Chief Division of Engineering & System Technology

SUBJECT:

Responses to NRC Questions on WCAP-ll397, " Revised Thermal Design Procedure" [Non-Proprietary)

REFERENCE:

(1) Letter from M. W. Hodges (NRC) to W. J. Johnson (V),

(Questions on WCAP-11397), " Revised Thermal Design Procedure" dated April 4, 1988.

Dear Mr. Ilodges:

Enclosed are:

Twelve (12) copies of WCAP-11397, Addendum 1, " Responses to Additional Questions on Revised The.rmal Design Procedure" [Non-Proprietary).

The enclosed information is being submitted in response to additional NRC questions, Reference (1), as a result of the Staff's and their consultants review of the subject Topical.

Very truly yours, J hnson, Manager W

N c1 Safety Department

/mit Enclosure

MMY4UE @>KlE52 UM 2 MCAP-11397 Addendum 1 REVISED THERMAL DESIGN PROCEDURE A. J. Friedland S. Ray June 1988 Approved:

W E. H. Novendstern, Manager Thermal-Hydraulic Design And Fuel Licensing HESTINGHOUSE ELECTRIC CORPORATION Commercial Nuclear Fuel Division P. O. Box 3912 Pittsburgh, Pennsylvania 15230 545BL:6-880614

mpocEE PROPR1EDXf G1.SS ?

RESPONSE TO NRC QUEc'" IONS ON WCAP-11397,

" REVISED THERMAL DESIGN PROCEDURE" 1.

Dplain tint is curt by DBt(sudhle) and P/M(swJble) in the Povised Ihca:n1 Desigt Prmeri rm (P11P) cLvelcy:mt as pqge 5 4pinrtly dry am toch rartin wdhles, he it is urr'wr uhich sz:ces d treettairty are Lx1tried in ath. '1bs the 05t(swi&1e) a11y ctreau1 the trze:tatrty Ezon the dsigt.mmrm Ecr a gism tr1: ural set d exxxiitias? Lhs Pf1(varible) a11 ctncaL, 7

trrem [2an de azreirim fcr a gism ra: uni sec cf azzilcias? Mut am de rurs azi scritmi cLviefms (rnh m cerotada1) Ecr ad1 ef these nrzi1n vad&1es? Is de "(radhle)"

dearipctr acarni cn de Ppt's a11 t*S's in egefas (2-12) dicqf1 (2-14)?

RESPONSE TO OUESTION 1:

In the development on page 5:

1 DNBR(variable) only contains the uncertainties from the design parameters for a given nominal set of conditions.

P/M(variable) only contains uncertainty from the correlation for a given nominal set of conditions.

y is defined as DNBR(variable)/DNBR(nominal).

y has a mean of 1.0 (p - 1.0) and a standard deviation designated cr.

The reciprocal of P/M(variable) is M/P(variable), with mean p and gpp standard deviation ogjp.

The "(variable)" descriptor is assumed on the P/M's and M/P's in equation (2-12) through (2-14).

WrsIE47LT ihnat omi CBS 2 2.

Itu is z in meim (2-14) irtapind? It 6es tot agpeer to be "~%1 aniqps to y, ddch we 7

at tirettainty fm. Is z Jst a 11ER nrxiza variable an;aiturg bcd1 cw1eim ad cksig, pra rnr trnstzunties?

RESPONSE TO OUESTION 2:

Z is defined by equation (2-14) and is a DNBR random variable containing both correlation and design parameter uncertainties, as inferred in the question.

3.

De eqzusia, in mesm (2-18) is drculxd as a2 tgxt 958 axfidere limit a, de staxitd devirim nm"~fruf with de urettainty in de cwTrim it1.wr, T/1.6'S is in genul la:ger dm (d'/dd-speeM.5, de qpqxiate mlHplier to cbtain a, tgper 95% a:rfidare li=it. De n.o cptrities are apal Jm de tue can cf tyP is htwt (Jadt is irt de case). Still, de tse cf K/1.t%S p:vvids fcr de m'ind dsig, limit Q apalirg de cwTrtm li=it (CL) utm due is to ace:tairty in de dsip2 puwetas is this de nun 1 fcr birg dxc ws dxe in njwim (2-18)? Or sus it to allcw fcr tncti:g de txte va:hrte cf PS as tukve2? In azy case, pla2se eplain de ratianle fcr apria, (2-18) ad de aw=.yuiirg descripcia2 RESPONSE TO OUESTION 3:

The term K/1.645 is used in equation (2-18) to account for the uncertainty in the mean of M/P as well as the uncertainty in the standard deviation of M/P.

The resulting o is that for a normal distribution which has the same gpp upper 95% tolerance limit as is given by Owen's table.

For known y and o, upper 95% tolerance limit (UTL) - p + 1.645a For sample E and s, UTL - E + ks For equal UTL's y + 1.645a - E + ks

, _ i-p + ks 1.645 K s - o if p - i 1.645

gAC15 N;mEThM CLASS 2 l

i 4.

As rrted in de pnwias qtustfat, a1 "tgrr arfi&me 1.fo:f.t" is pinri cn sigm(WP). Pasu:nbly dds ms roc dw to sigm(y) hrm it is belig twM as krut. C1arly, it is roc kru1 ccx:tly ba:ase entr,mHm crdzris 6tddi are grt>:i::ste aziede cmain em?!ms) sere used ca derive it. Also, dxte are sinntlas dun de irx!Lvid & ;- : a-- n Lux:es coy rcc be justifiably t w M as kxu1. The Omix cf sig:a(y) as kxw 1"n* tv Q but scuhlig diffennc sadd be obtairxd if sigm(y) n=e ttural as trkruz Please dIo m azi stentt yar tru1:::mc cf sigm(y).

RESPONSE TO OUESTION 4:

is treated identically to the way it is treated in ITDP (UCAP-8567).

The c

NRC review (Ref. 10 of UCAP-ll397) of WCAP-8567 states:

"The proposed design basis is that there must be at least a 95%

probability that the minimum DNBR of the limiting power rod is greater than or equal to the DNBR limit of the correlation being used; parameter uncertainties or variances obtained from the evaluation of data are. determined at a 95% confidence level.

Implementation of the procedure involves assumptions that errors in correlations or postulated functions are random variables where repeated use of a correlation with identical conditions gives identical results.

Also, distributions of uncertainties for variables such as power, flow or temperature are not well known so that the functional forms of the uncertainty distributions must be assumed.

Therefore, a rigorous statistical statement of the type implied in the proposed DNBR design basis cannot be obtained.

"However, Westinghouse has either chosen distribution functions which are typical of observed distributions or which give conservative variances.

Also, biases in correlations have been reduced insofar.as practical.

Therefore, although a rigorous statistical statement cannot be made, the Vestinghouse method provides a reasonable approximation to such a statistical statement."

j For RTDP, the design parameter variances will be determined on a plant specific basis by the identical procedure currently used for ITDP.

1 i

I

i v57mt,ity igy b.t.gf QK S.

The last smcare cf de fizsc my cf Sa:tia11 sentm de asign critaria1 dat is t*e basis Ecr de RilP dim & in de _a hern1 ~de "958 perimbility ac 95% a:rficktre" pxtlas cf de criteia, is typfmily L.piwrrted via tse cf a au-sidai tolmste ittenal based at de rural AMhrim. A dexcipcian cf a 95% 95% toletture inwnul arxi fru it differs fran a 95% a12fidste intanul is S rm i

i in ete 4pnhx A tolante ittenal is a si-L duc a sirgle 6L&he!m. Alduft the PilP flgn ad nytim en de httun cd pqge B sq; gest duc de FilP cxdtd is ansidrirg a shyle di&h rim dn: arbines de cxnelaticn ard desigt ;- -. -,- uruttaincies, de daelqnst cn pge 6 sats to Ldcate S,w!Fim!1, Eq. (2-13) cceptes a toleure itte: val for de "axw7efm urettaL-ty" cdurvise.

7 distrihr1m, skille F. (2-20) is de nmit cf a 95% caf1&rre incarval en de cabirni discrih t!m 4

ufx e a "toleur:e interrul irf7en warlar:e is usal far de cocv7stm ure:rmity 1his T."P agrarh is shnilt to, h: roc quite de sa o as sin: sc2ld be cbcairni if a 958 958 tole:ure

/

brenal uns arpeni cn the cobimi distrih *im trrier de xa smeim dm: sigm(y) is ktwn viduc N.

Pla2se e9 a.n de in 2aule far de PliP develqvinc stqs afemd to drve, axi eplain Irv it 1

t c7'ldies de "95% ptdubility ac 95% cafidsre" p11.ial ci de cksigt critadal fran %*im 1.

RESPONSE TO OUESTION 5:

The approach used in RTDP is the same as that in ITDP with M/P included among the independent variables.

The response to question 4 includes the NRC comments on the statistical approach used in ITDP.

6.

Dimmias dungnt de subnittal irply dat cC$T) axi s(WP) are enlmiral cnw de sec cf.%T vnh-c1rnsprxhq; to de dra tsai to dw1cp de corw7eim 1Lo conres are crde w7eise ec this.

(a) It is ingxquiate to arpre a run arxi sM' deviarial in this fashial tniess it carn be dmrrsccoud dnt de tVP ralues are a nut:n arple Stan a a1mm ;m arIm. This will usually roc te de case fcr a ginn axxv7 elm axi its ww=,,udirg chta base ad ng:a1 cf qplichility. If de WP ralus cb trc are f:cn a ctrum pwin~fm, de can ad s'M civiatias nny tu:7 siglifiantly cne radas alnpas cE de nya1 cf wH44 H7 CD11s vill nxpize arpriry differat stay litits in diEmrt stittgzas, cr dxxshg de latyst safety 1h::i: !! de 1Jcle nyal is to be exxe:xd tsirg enly a sirgle rulue.)

~

i nTJUmrEE PR0f*dEEY CU32 4

L (b) It in iwylate to amhea arri remare thembw d a cerzelrdn usfrg the sste data imd to hekp it, vf=777 ff ifH=7 fittirg tadniquer azh ar lasse apsres are used Pmb-- nessuns (e.g., murs arf standed dsvisimr d P/f) druid be crrp raf fztrn &ta nr i

used to heicp the crrrefefm If such "se2a" duca are tre -f 7A7a, chta-7 refrg cr crtss-l 7f

=7WHm tainigass druid be used to amhmr= arzf rnenerre ;

n-. -w, with tk resuIcirg mrs arf serrhtf civiefnw tatiin ~7a'fe4 the chs.fg2 lide.

Aidag2 this -sf ~7 anrs general nedvvh7w, de arxe pires must be recry21raf enn gplydrg tfr omdrvh7qy to Svr ffic v7imeivs. Plase A=,a frwyur gpund2 fcr 4plyirg the FLIP arnrts fer the tw Imm raised abcne.

l

-i RESPONSE TO OUESTION 6:

l

-The DNB correlation statistics are the same as those which were used to develop the DNBR correlation limits in the previous ITDP methodology which have been accepted by the NRC.

For the WRB-2 correlation, for example (Ref.

6-1), the hTC SER presented a statistical analysis of the test data and concluded that while the data do'not all represent a single population, the error introduced into variance w rimates by assuming that the data are all from the same population is negligible.

Similar results are obtained for the URB-1 correlation (Ref. 3 of WCAP-11397) applied to "R". grid data.

The evaluation of DNB correlation is handled in separate topicals, as in the.

11 above examples.

The RTDP derivation treats the data as coming from a common-population, and assumes that the error introduced by this assumption is negligible as shown in the appropriate correlation topical report.

(Ref. 6-1, WCAP-10444-P-A, " Reference Core Report - Vantage 5-Fuel Assembly," September 1985) i p

--e n,,

qGittavmW I4ifF6ETGI O!Ils 2 lie firs ja yy d c tim 5 (pge 14) is nr "wlely clar zmudirg har p ;19tse to cbcabt 7.

a tte sesitivity fxttes. Ib p 3xqrse dusiry as the sursitivity fxtrus drce vahxs aamqudug to de 11mittig stwqvint (de cre with de higtst Q as.qucifial en pge 14)7 1his sans to be a xmscnble extah (givm ttat augh statqvirts are ect:zud), but exrflicts ith de ckrnpcia2 cf sesitivity fxtats given in krim 2.1. De pqge 3, de smsitivity fxtrrs vem ckscribal as beirg jawu,+ dxtys in LtBR nw1Hrg fitm 18 chtges in de x,. Sime to tuntun is nde cf de s

sesitivity fraxs ckpntry at de s cr de cwirim, de im11w!m is the dey am 3

2rripexirr cf de x ard de amirim Har:e, de 4pmrr czrfbet with de p:qmi cn ;rge 14 j

Plex;e clanfy pr;rqmi fer an~minirg de sesitivity fxtrrs fer a ginn glie'im, irritdirg has p will An-ine de nrler ad in~ rim cf swqvints to be ensidmd. Dplain sJrj this a;pitah is a2 iqxwourt cur tte a;pinxh anturni in de sand yagy cn pge 14.

RESPONSE TO OUESTION 7:

We will choose as the sensitivity factors those values corresponding to the limiting statepoint (the one with the highest DL. as sPecified on page 14).

R The statement on page 3 that the sensitivity factors are percentage changes in DNER resulting from 1% changes in the xg, applies at the limiting statepoint. The sensitivity factors do depend somewhat on the x and the l

g correlation, so some of them may be higher at some other statepoints.

However, as noted in Refs. 1 and 10 of WCAP-11397, the observed variations in sensitivity factors are generally small over a wide range of conditions.

Selecting the sensitivities at the limiting statepoint is sufficient to assure a conservative DI 7

As is currently done for ITDP, sensitivity factors are determined for l

various statepoints over a range of conditions for which the RTDP methodology will be applicable. With RTDP, the sensitivity factors at each statepoint are used to calculate a DL value.

The highest of these is used g

as the design limit DNBR. This approach removes some unnecessary conservatism from the approach mentioned in the second paragraph on page 14, which combines worst sensitivities from different statepoints.

-YSTfiGOUE D9$P1GGY WSS 2 8.

De dicmnkn before axi after meim (5-1) daals with a siryle Lvelgrim cf tte validity cf usiry meim (2-3)..Tn smaral, "pttti by ecrple" is dxyms, vt=11y Jun cnly cne earple is usa!. Pmu:xbly nsults m17d differ far diEennt wirins, diffenst w1ues cf tk aw?'im irput wriahIes, diffenne cbsign ;m : 4,w, axi diffenre w7= cf tte desigi pumecers Justi$*

tint (2 3) is a cmservative wulm:.ia2 turier cduc passible cirmnu=,7 BESPONSC TO OUESTION 8:

As noted on page 14, the procedure used to* check the form of Eq. (2-3) was the same as that previously evaluated by the NRC in Ref. 10.

The form of the equation is based on observation and experience with the THINC computer program and the Westinghouse DNB correlations, which generally show small variation in tihe S over a wide range of conditions.

The numerical example g

which was made at nominal plant conditions is considered to be a check rather than a proof c f the methodology. A similar check was made with one l

test case at the limiting statepoint conditions where the sensitivities resulted in the highest DL and the second case with all design parameters R

one standard deviation from the limiting statepoint conditions in the direction leading to a decrease in DNBR.

The combined effect on DNBR of these deviations is considerably greater than 1.645 times the RMS of the individual DNBR standard deviations.

If Eq. (5-1) were exact with constant S then it would be expected to agree exactly with the THINC analyses, since g

I the S were evaluated near the limiting statepoints.

This was found to be the case with both the THINC analyses and Eq. (5-1) showing the limiting statepoint ratio of perturbed to nominal DNBR to be 0.834 for the typical I

cell and 0.847 for the thimble cell.

Similar evaluations will be performed if there a;e my changes to the DNBR correlation or in the limiting 1

statepoint.

l 9.

In meim (5-3), din 162't ad1 af tte s(x-nu) te=x te dividd by tk auw,uu nu?

z RESPONSE TO OUESTION 9:

i i

i Yes.

The equation was misprinted and should read:

l l

+... + Sm(ym-pm) + (M/P)

(5-3)

+ S (X "#2)

._Z - S (xi 1)

-9 2 2 1

t 1

2

1. m M/P

c'EP,7KtUSE CWNTTWf Of M 2 The calculated results in Table 5-1 are correct.

The misprint vill be corrected in the approved version of the topical.

10. De d2.stussim ad arprisas en pys 15-16 cf DP-11397 a:e am=xd wid1 Lwip*Lz de akqrry cf de lixar guam in nynrirrs (2-6) ad (2-15). As in cpcsticn 8, se hr.e arte:rs dnr de siliciety cf yar Lustigtsm in das case yas did ansidr seem1 valas of (cr drugs in) tfe x,, b.c all are fcr de sxe sec cf dwig2 puu~erzs Vid2 a find sec af runs ad sc1xtzd &&rrs). Plase jstify Jur arcLcial dat de li:nr agrtxtr:rrims are akga*e ad am:mtive for all pxsible diHe:nt arbimtims cf dsig2 prm.

RESPONSE TO OUESTION 10:

The cases tested are those suggested in the NRC evaluation of WCAP-8567 (Ref. 10 of UCAP-11397).

They provide a confirmatory check that the linear approximation is reasonable and conservative.

It can be shown that, in general, if all the deviations are in the direction leading to a decrease in DNBR, and all 0 are nu:cerically greater than or equal to unity as is generally true, then the linear approximation is conservative.

This is shown by expanding Z in a Taylor's series about the y and including second order terms.

The values of the variables which are major contributors to defining the lower 95% limit of 2/

will be mostly in the direction of negative deviations of Z from Therefore, it is reasonable to expect the linear approximations to be conservative.

11. De shitral dxs roc dmm hu de gnral ::edr t,1cp vill be irple:1rnd in s;xcific a;plic cicts. Carly de scrsitivity f:rms trud to be cMmi for ada nu arrelatia2, ser cf 6sig, puu:ece:s, cr nry aC yli&>ility. It is roc clar Judur it will be nxma:7 to duk de wrics anvrrkrs ad appu+nrims cf de F2P fzt ada glim *!m.

FLmily, tte subr.i ~4 chs roc dunes Irw de dsig, pra~1rir urians will be chri::::&nd. Pla2se disans tirse Lssas nyrdLz de 9 m~im ci de gn=n1 F2P rzdrit>1qy to sprific g1im-L s.

The uvicue:s 7i rnd to krw durfer it is Jst de gnral crdrxblqy dar is tnir nvizw. cr if de Lplairmim ad verifim-1m acptts fir e re!fic 9 mtL,s a v aim to be nvia.tri4prtud j

7i A

i WESTtNQGISE MOPQJETARY C1 ASS 2 4

RESPONSE TO OUEST10N 11:

The general methodology for RTDP will be implemented in the same manner as for ITDP.

Sensitivity factors will be determined for each new correlation, set of design parameters, or range of applicability.

The design parameter variances will be determined on a plant specific basis by the identical procedure currently used for ITDP.

12. Ile first-crder emr JrTi"*fm fr-Ta (2-B) assums ont de nrrf=n varf*7-x..., x are 1

m en~ierem777.hdymdsr. M:lle dds==r7eim ms articrmd in de = Men 7, its zeestrablaiess as tre die-f. Plasse pwide s.rh a rffemicn, nakirg n:fenrre to specific obsig) prnrzras at givirg nascrs W de,r=='em am m nnf 7y rriprrirt

13. Place clanfy Jur s~ennt en pse 4 dat "De cerexal limit tha:rm cf en~selec irrfierm dat de,rr +<^f 7f ty dist:rb eim ftretia: Icr y will qpmf2 a rrrm7 diertheim with rar2 cu(y) ad scxxkxrd dwiacicn siga(y) en, if de irdivitin1 dis:rihrims cf de s, am rzt trc:ral." Clar:17 de cerral li:sc dxtra is nx q7Imhle to "7n~im (2-5). Am)cu basirg this stataart at the

== rreim d (2-3) betry =%=**1y ww ~ ~. #1 by (2-6)? Also, sirne x, am rrt iderien117 dis:r5h raf, a are actrictiw fcr:0 d the cazzal limit dnzza mst le ceidi"1 Firally, de crrmprre to tr=-=7 fry d a linssr cxz:birncien cf rarixn wzdablas is d =dzr en the rurkr cf i

verl+1m ard their irriividal dLstrih t1ms, y]agse argidy gl] g{ tte atCKt p31rts in CladfyL'y ad anrrtfrg yur sentawrs crroemfrg tk cxrezal ll=it thszm ad a1 ww 1. ~ -rzz:md dis:riNefm fcr y.

RESPONSE TO OUEST10NS 12 AND 13:

Questions 12 and 13 deal with the ITDP development, which has'been covered in the NRC review and evaluation of WCAP-8567 (Ref. 10 of WCAP-11397).

The RTDP development is identical to the ITDP development in these areas.

14. 2he ckv.lcp:xrt cf de istsed dsigt 11mLt y asams dat ecy rk!gs,rnm,cw is eif _Mh en1 wid1 a nun value apd tn its xxz:xrni udue. Will this abarys be tk anse, er is itIrmth7e da:

biases cigt adst? If sa, Irw cbs de R:IP fxrd?e the bias adirw x2dd g chape?

l

WiDDMZ WF6TW!CW 2

.I RESPONSE TO OUESTION 14:

l Biases might exist in design parameters.

In such cases, the bias is included separately in the DNBR analysis. The bias is not included 6

indevelvpentofDQ.

t

-?

t f

i

.