ML19350F124

Forwards Proprietary & Nonproprietary Versions of Responses to Questions on Documents Supporting ANO-2 Cycle 2 License Submittal, & Reactor Vessel Open Core Flow Model Test Rept. Proprietary Versions Withheld (Ref 10CFR2.790)

ML19350F124
Person / Time
Site:	Arkansas Nuclear
Issue date:	06/19/1981
From:	Trimble D ARKANSAS POWER & LIGHT CO.
To:	Clark R Office of Nuclear Reactor Regulation
Shared Package
ML19276K212	List: ML19350F124 ML19350F125
References
2CAN068116, 2CAN68116, NUDOCS 8106240240
Download: ML19350F124 (133)

Text

. 4 ARKANSAS POWER & LIGHT COMPANY POST OFFICE BOX 551 LirTLE ACCK, ARKANSAS 72203 (501) 371-4000 June 19, 1981

1. 2 g>gM, k/h s

s m (T) e@

2CAN068116 .

Ea oA /

^

Nh * (

Director of Nuclear Reactor Regulation -

ATTN: Robert A. Clark. Chief Operating Reactors Branch #3

)Kf-s, C'

Division of Licensing U. S. Nuclear Regulatory Com.

Washington, D.C. 20555 SUBJ ECT: Arkansas Nuclear One - Unit 2 Docket No. 50-368 License No. NPF-6 Responses to NRC Questions on the ANO-2, Cycle 2 Reload Report (File: 2-1510)

Gentlemen:

Please find enclosed as Attachment A one proprietary and five non-proprietary copies of a report designated CENPD-157(A)-P, Amendment 1, containing responses to NRC Staff questions on the ANO-2, Cycle 2, reload report which were sent to AP&L on May 5, 1981 and May 6, 1981. All of these responses, except those to question numbers 492.64 and 492.76 were infonnally submitted to the staff on May 18, 1981. The responses to question number 492.64 and 422.76 were infonnally submitted on June 10, 1981.

It is requested that the appropriately designated portions of this report be classified as proprietary information in accordance with the provisions of 10 CFR 2.790. The required affidavit from Combustion Engineering is attached.

Attachment B to this letter contains a demonstration case requested by the NRC staff during a telephone conversation held on June 2, 1981. This information was telecopied to NRC on June 10, 1981.

Attachment C contains one proprietary and five non-p oprietary copies of the vessel flow model test report which was requesteu by Mr. Phillips of the NRC Staff on May 7,1981. It is requested that the appropriately 8106240 F 0

--,ca scum unures svsreu j

' Mr.- Robert A'. Clark, ' Chief June 19, 1981 designated portions of this ' report be classified as proprietary information in accordance with 10 CFR 2.790. The required affidavit from Combustion Engineering is attached. ,

g Very truly yours, 1 N 0. E David C. Trimble

- Manager, Licensing DCT:1p Attachments l

l

.- i

.- 1

. (

L

l AFFIDAVIT PURSUANT T0 10 CFR 2.790 Combustion Engineering, Inc. )

State of Connecticut ) .

County of Hartford ) SS.: . '5

-o I, A. E. Scherer depose and say that I.am the Director, Nuclear Licensing of Combustion Engineering, Inc., duly authorized to make this affidavit, and have reviewed or caused to have reviewed the information which is identified as proprietary and referenced in the paragraph immediately below. I am submitting this affidavit in conformance with the provisions of 10 CFR 2.790 of the Commission's regulations and in conjunction with the

. application of Arkansas Power and Light, for withholding this information.

The information for which proprietary treatment is sought is contained in the following document:

Amendment 1-P to CEN-157(A) - P, Response to Questions on Documents Supporting the ANO-2 Cycle 2 License Submittal.

This document has been appropriately designated as proprietarj.

I have personal knowledge of the criteria and procedures utilized by Combustion Engineering in designating information as a trade secret, privileged or as confidential commercial or financial information.

Pursuant to the provisions of-paragraph (b) (4) of Section 2.790 of

! the Commission's regulations, the following is furnished for consideration by the Commission in determining whether the information sought to be l

withheld from public disclosure, included in the above referenced document, should be withheld.

l I

l O

e

- 1. -The information sought to be withheld from public disclosure are details of the C-E t"ermal margin analysis methodology and thermal hydraulic

-characteristics of C-: cores, which is owned and has been held in confidence by Ccmbustion Engineering.

~

2. The information consists of test data or other similar data concerning a process, method or component, the application of which results in a substantial competitive advantage to Combustion Engineering.

. 3. The information is of a type customarily held in confidence by Combustion Engineering and not customarily disclosed to the public.

Combustion Engineering has a rational basis for determining 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 confidence. The details of the aforementioned system were provided to the Nuclear Regulatory Comission via letter DP-537 from F.M. Stern to Frank Schroeder dated December 2,1974. This system was applied in determining that the subject documents herein are proprietary.

4. The information is being transmitted to the Commission in confidence under the provisions of 10 CFR 2.790 with the understanding that it is to be received in confidence by the Comission.

5. The infonnation, to the best of my knowledge and belief, is not available in public sources, and any disclosure to third parties has been

~

made pursuant to regulatory provisions or proprietary agreements which provide for maintenance of the irformation in confidence.

l

6. Public disclosure of the information is likely to cause substantial harm to the competitive position of Combustion Engineering because:

a. A similar product is manufactured and sold by major pressurized water reactors competitors of Combustion Engineering.

b. Development of this infonnation by C-E required thousands of 5

~~

man-hours of effort and tens of thousands of dollars. To the best of my knowledge and belief a competitor would have to undergo similar expense in.

generating equivalent information.

c. In order to acquire such information, a competitor would also require considerable time and inconvenience related to thermal margin analysis methods development.

d. The information requiv~1 significant effort and expense to obtain the licensing approvals necessary for application of the information.

Avoidance of this expense would decrease a competitor's cost in applying the information and marketing the product to which the information is applicable.

e. The infonnation consists of details of tne C-E thermal margin analysis methodology and thennal hydraulic chanracteristics of C-E cores, the application of which provides a competitive economic advantage.

The availability of such information to competitors would enable thein to modify their product to better compete with Combustion Engineering, take marketing or other actions to improve their product's position or impair I

the position of Combustion Engineering's product, and avoid developing l similar data and analyses in support of their processes, methods or apparatus,

f. In pricing Combustion Engineering's products and services, significant research, development, engineering, analytical, manufacturing, l licensing, quality assurance ar.d other costs and expenses "must be included.

, a . . _ _

4_

Ti:e abilit*, of Combustion Engineering's competitors to utilize such information without similar expenditure of resources may enable them to sell at prices reflecting significantly lower costs. I

g. Use of the information by competitors in the international marketplace would increase their ability to market nuclear steam supply 3 systems by reducing the costs associated with their technology development.

In addition, disclosure would have an adverse economic impact on Combustion Engineering's potential for obtaining or maintaining foreign licensees.

Further the deponent sayeth not.

4

_ suw A.E. deNr Director Nuclear Licensing Sworn to before me this / / day of ' - ' - '

f q-* 'I :s .

Notary Public J CAREY. J. WENZEL, NOTARY PL1RQ state of Connecticut tio. 59962 Ccmm:ss:en Expires March 3P,1935

4 .

AFFIDAVIT PURSUANT TO 10 CFR 2.790 Combustion Engineering, Inc. )

State of Connecticut ) 5 County of Hartford ) SS.:

I, A. E. Scherer depose and say that I am the Director, Nuclear Licensing of Combustion Engineering, Inc., duly authorized to make this affidavit, and have reviewed or caused to have reviewed the information which is identified as proprietary and referenced in the paragraph immediately below. I am submitting this affidai<t in conformance with the provisions of 10 CFR 2.790 of the Connission regulations and in conjunction with the application of Arkansas Power and Light Company, for withholding this i nformation.

The information for which proprietary treatment is sought is contained in the following document:

CEN-167(A) - P, Reaccor Vessel Open Core Flcw Model Test Report.

This docunant has been appropriately designated as proprietary.

. I have personal knowledge of the criteria and procedures utilized by Combustion Engineering in designating information as a trade secret, privileged or as confidential commercial or financial information.

Pursuanttotheprovisionsofparagraph(b)(4)ofSection2.790of the Commission's regulations, the following is furnished for consideration by the Commission in determining whether the information sought to be withheld from public disclosure, included in the above referenced document, should be withheld.

e 0 9

1. The information sought to be withheld from public disclosure are flow and pressure distribution data at the core inlet and exit planes for the 34XX class reactors, which is owned and has been held in confidence by

. Combustion Engineering.

. 5

2. The information consists of test data or other similar data

-c concerning a process, method or component, the application of which results in a substantial competitive advantage to Combustion Engineering.

3. The information is of a type customarily held in confidence by Combustion Engineering and not customarily disclosed to the public.

Combustion Engineering has a rational basis for determining 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 confidence. The details of the aforementioned system were provided to the Nuclear Regulatory Commission via letter DP-537 from F.M. Stern to Frank Schroeder dated December 2,1974. This system was applied in determining that the subject documents herein are proprietary.

4. The information is being transmitted to the Commission in confidence under the provisions of 10 CFR 2.790 with the understanding that it is to be received in confidence by the Commission.

5. The information, to the best of my knowledge and belief, is not available in public sources, and any disclosure to third parties has been made pursuant to regulatory provisions or proprietary agreements which provide for maintenance of the information in confidence.

O e

I

~6. Public aisclosure of the information is likely to cause substantial h' arm to the competitive position of Combustion Engineering because:

a. A similar product is manufactured and sold by major pressurized water raactors competitors of Combustion Engineering.

2evelopment of this information by C-E required tens of -

?

thousands of man-hours of effort and hundreds of thousands of dollars. To the best of my knowledge and belief a competitor would have to undergo similar e" pense in generating equivalent infomation,

c. In order to acquire such information, a competitor would also_ require considerable time and inconvenience related to conducting tests and analyzing data from flos niedal tests.

d. The information required significant effort and expense to obtain the licensing approvals necessary for application of the information.

Avoidance of this expense would decrease a competitor's cost in applying the _information and marketing the product to which the information is applicable.

e. The information consists of flow and pressure distribution data at the core inlet and exit planes of the 34XX class reactors, the application of which provides a competitive economic advantage. The availability of such information to competitors would enable tnem to modify their product to better compete with Combustion Engineer *ng, take marketing or other actions to improve their product's position or impair the position of Combustion Engineering's product, and avoid developing similar data and analyses in support of their processes, methods or apparatus.

i

f. In pricing Combustion Engineering's products and services, significant research, development, engineering, analytical, manufacturing, licensing, quality assurance and other costs and expenses'must be included.

9

l 1

l l

l The ability of Combustion Engineering's competitors to utilize such information without similar expenditure of resources may enable them to sell at prices reflecting significantly lower costs.

g. Use of the information by competitors in the international marketplace would increase their ability to market nuclear steam supply - 2 systems by reducing the costs associated with their technology development.

In addition, disclosure would have an adverse economic impact on Combustion Engineering's potential for obtaining or maintaining foreign licensees.

Further the deponent sayeth not.

ls3 -

A. E. Scherer Director Nuclear Licensing Sworn to before me

/

'I this day of 2 ' 'I 7 -

_ _ %LLu '_ !_

Notary PutIlic . J C.wEl J. % L.sL_L, .NorAny pCBLIC State of Connecucut No. 59962 Commission Expires Marcri 31,195 l

l e

s

e ATTACHMENT D I

(

w e

i

I NRC Question During the NRC audit of the AN0-2 Cycle 2 CPC software there was a demonstration of a Phase II DSVT case with the FORTRAN simulation code on the comparison between the Cycle 1 and Cycle 2 softwcre. git was noted that the hot leg temperature in the Cycle 2 case datag was 2 less than the hot leg temperature in the Cycle 1 case data. Is this 2 difference due to the new pressure -

temperature curve fits for enthalpy-temperature ratios implemented for Cycle 2?

Response

In designing the new STATIC 7eogram for the CPC/CEAC Systems software, pressure and temperature curva fits were developed for calculating the liquid properties required for the T0kC/CE-1 DNBR calculation. The methods used yielded coefficients which calculated properties more closely appronnating the 1967 ASME Steam Table values. The same methods were utilized to determine constants for new pressure-temperature curve fits for the enthalpy-temperature ratios that are consistent with those used for the liquid properties. As a result, the enthalpy ratios based on the new curve fits (TORC /CE-1 correlation) generate a static thermal power value approximately 2% greater than a static thermal power value generated with enthalpy-temperature ratios based on the old curve fits (W-3 correlation), when u.ing the same cold leg and hot leg temperatures.

In order to compensate for this increased accuracy when making a Cycle 1/ Cycle 2 steady state or transient analysis comparison, the initiaf hot leg temperature for the Cycle 2 case data must be reduced approximately 2 F below the Cycle 1 value in order to generate a static thermal power value for Cycle 2 equivalent to that generated for the same Cycle 1 analysis.

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

.\

1-l-

l I

. ARKANSAS NUCLEAR ONE.- UtilT 2 ,

DOCKET 50-368 Cell-157( A)-NP Amendment 1-NP j l

I  !

f l l RESPONSE TO QUESTIONS 0H DOCUMENTS SUPPORTING  !

h' : THE ANO-2 CYCLE 2 I LICENSE SUBMITTAL r

I i

i  !

l JUNE 1981 f i.

l COMP,USTION ENGINEERlfG, IfiC. j NUCLEAR POWER SYSTEMS l POWER SYSTEMS GROUP WIfiDSOR, CONNECTICUT 06095  ;

G  !

p LEGAL NOTICE THIS REPORT WAS PREPARED AS AN ACCOUNT OF WORK SPONSORED BY COMCUSTION ENGit:EERING, INC. NEITHER CO.3USTION ENGINEERING 4 NOR ANY PERSCN ACTING ON ITS BEHALF:

A. MAKES ANY WARRANTY OR R EPR ES ENTATICN, EXPRESS OR IMPLIED INCLUDING THE WARRANTIES OF FITNESS FCR A FARTICULAR PURPOSE OR MERCH ANTASI LITY, WITH i.es?ECT TO THE ACCURACY, COMPLETENESS, CR USEFULNESS OF THE INFOR*.;ATION CONTAINED IN THIS REPORT, CR THAT THE USE OF ANY INFORI.*ATION, APPARATUS, METHOD, OR PROCESS DISCLCSED IN THIS REPORT MAY NOT INFRINGE PRIVATELY OWNED RIGHTS;OR B. ASSUf/.ES ANY LI ABILITIES WITH RESPECT TO THE USE OF, OR FOR DAMAGES RESULTING FROf.1 THE USE OF, ANY lt'E.ORMATION, APPARATUS, METHOD OR PRCCESS D SCLCSED IN THIS REPORT.

()'

/

4 4

j**v h/

&6_Mm ear 4' 'Me446 AL A4L+J.a2-4-m.%

.  ; s

T~ '

j ~

j.

,. t i

i i- ,

a  ;

e

! i l

3 C EN-157( A)-NP  !

l 4

Amendment 1-NP i

i i

t- i

$ i 1

l 1

I l

1 i

! l i

h l

' I RESPONSE T0 QUESTIONS i ON DOCUMENTS SUPPORTIfG

! THE ANO-2 CYCLE 2 j LICENSE SUBMITTAL i

i i

i l

i s

't 1

i l

O  !

1 i

? l l

Abstract

/3 This report contains answers to NRC questions 492.1 through 492.29 on

\)

Arkansas Nuclear One lnit 2 Cycle 2 licensing support documents with the exception of 492.23 and 492.24 which wet e supplied by AP&L. It also contains the answers to questions 492.48 through 492.77 with the exception of a portion of question 492.67 which will be provided separately by AP&L.

J e

O 4

1 f

O

\

Table of Contents 6:

) .

legal flotice j

Title Page jj Abs tract jjj Table of Contents jy introduction l' Responses to NRC Questions on CEN-143(A)-P and CEti-139(A)-P 2 492.1 (A-1) 3 492.2 (A-3) 3 492.3 (A-4) 3 492.4 (A-5) 4 492.5 (A-6) 7 492.6 (A-7) 7 492.7 (A-8) 8 492.8 (A-9) 0 492.9 (A-10) 11

,s 492.10(A-ll) 11

'd 492.ll(A-12) 11 492.12(A-13) 11 492.13(A-14) 12 492.14(A-15) 12 492.15(A-16) -

12 492.16(A-17) . 12 492.17 ( A-18 )- 14 492.18(A-19) 14 492.19(A-20) 15 492. 20(A-21 ) 15 492.21(A-22) 15 492.22(A-23) 15 942.25(A-26) 16 492.26 , 18 492.27(A-2 ) 19 492.28(A-27) 19

(] 492.29(A-28) 20

Table of Contents (continued) 492.48 48-1 rs 492.49 49-1

'/ 492.50 50-1 492.51 51-1 492.52 52-1 492.53 53-1 492.54 54-1 492.55 55-1 492.56 . 56-1 492.57 57-1 492.58 58-1 492.59 59-1 492.60 69-1 492.61 61-1 492.62 62-1 492.63 63-1 492.64 64-1 492.65 65-1 942.66 66-1

{])

)

492.67 67-1 492.68 68-1 492.69 69-1 492.70 70-1 492.71 71 -1 492.72 72-1 492.73 73-1

' 492.74 74-1 492.75 57-1 492.77 77-1 Referentms R-1 1

l CE) 1

l.0 Introduction i  ;

sj C-E% reports CEft-143(A)-P and CEN-139(A)-P have been submitted on the Arkansas Nuclear One-Unit 2 docket in support of the Cycle 2 License submittal. f4RC

~

questions abcat these two reports were given to Arkansas Power and Light and Combustion Engineering, Inc. at the l'. arch 26, 1981 meeting in Bethesda, MD.

These questions were then identified as questions A-1 through A-28. Sub-sequently, a revised list of questions was transmitted by telecopy.

That list contained one additional question, reordored the resulting set of questions and redesignated them as 492.1 through 492.29.

This report contains answers to these questions except the two designated 429.23 (A-24) and 492-24 (A-25). They have been answered in a separate sub-rittal by AP&L.

Additional questions have been received from the NRC on CEN-143(A)-P and O CEN-139(A)-P. They have been designated questions 492.30 through 492.77.

The responses to questions 492.30 through 492.47 are non-proprietary and have not been included in this report. ,

This report does address questions 492.4fi through 492.77. A partial answer to question 492.67 is provided here. The balance of the answer to this question is being developed, and will be provided separately.

/~T U

1

l' \-

r -

('

I 1

l

! i l

l i

i 2.0 Responses to f1RC (uestions on CEft-143(A)-P and CEti-139(A)-P i

i -

f 1

1 l

i 1

1 I l 1

(

S I

i i

{

l l ,

l l i

l l

i d

1 i

f I

I

~

i O

2

~

I I

l tv, l Question 492.1 (A-1)

It was understood that the CETOP code was developed as a C-E Thermal On-Line Program. However, the Appendix A of CEN-143 refers to the CETOP as a design

, thermal margin program. Is the CETOP used as a design analysis tool for the ANO-2 core?

Answer CETOP (also referred to as CETOP-D) was used as the design therTnal margin code for ANO-2 Cycle 2. The CETOP-D code is used to derive and verify the CPC on-line thermal margin algorithm CETOP2.

Question 492.2 (A-3)

Provide a complete description of the CETOP program methodology, algorithm and its usage for ANO-2 Cycle 2 reload.

Answer A complete description of the CETOP (CETOP-D) programming methodology was O~ provided in response to first round questions on CEN-139(A)-P. The description of the CETOP2 algorithm was provided in Appendix B to CEN-143(A)-P.

Its usage for ANO-2 Cycle 2 was described in Section 6.1 of the Reload Analysis Report and in CEN-143(A)-P, Section 2.1.

Question 492.3 (A-4)

In the CETOP program, the transport coefficients of pressure, enthalpy and axial velocity associated with turbulent interchange are used in conservation equations. Describe how these coefficients are obtained. Provide sensitivity studies of DNBR vs. these coefficients. What are the values of these coefficients used in CETOP-2?

Answer Transport coefficients are used to adjust calculations involving a lumped channel for the fact that coolant properties associated with turbulent inter-change and diversion crossflow are not the lumped channel average values.

The application of the transport coefficients to the conservation equations is described in References 1 and 2.

REFERENCES

1. C. Chiu, et al., "Enthalpy Transfer Between PWR Fuel Assemblies in Analysis pV by the Lumped Subchannel Model," Nuclear Engineering and Design, 53, pp.

165-186,(1979).

2. "CETOP-D Code Structure and Modeling Methods, Uncertainties Progrum,"(Responses to F Questions on the Statistical Combination of CEN-139(A)-P), March 1981.

3

~

~

The pressure and velocity transport coefficients' will be' discussed first. These A- coefficients were shown in Reference 1 to have no significant effect on the

.U enthalpy, and therefore, on the DNBR, of the hot channel. Further evidence of the insensitivity of the DNBR to these. values is given in Table 1. The values of these coef ficients used in' CETOP D and CETOP_-2 are typical values calculated from TORC subchannel results. Table 2 provides the values used in CETOP-D and CETOP-2 for AN0-2, Cycle 2. The velocity transport coefficient is (mass velocity in the buffer channel equals the mass velocity in the hot channel.

) This is due to the simplifying assumption that the This simplification reduces the execution ~ time of- the algorithm. Any errors resulting from this simplification are covered by the algorithm penalty factor discussed in response to question 492.15.

The enthalpy transport coefficient has been shown to have a significant effect on the hot channel enthalpy (see Reference 1 and Table 1). In CETOP-D an algorithm is used to calculate cn enthalpy transport coefficient at each axial level. This method is described in Reference 2. In CETOP-2 a constant value is used for'the enthalpy transport coefficient in order to keep the algorithm-

- execution time to a minimum. The value for ANO-2 Cycle 2 is given in Table .2.

I

~

Any errors resulting from this simplification are covered by the algorithm 2

penalty . factor discussed in response to question 492.15.

The use of transport coefficients in the CETOP programs permits substantial simplification while retaining high accuracy. The tuning of the CETOP-D model to TORC over the entire range of operating conditions (See Reference 2) assures O that CETOP-D gives results which are conservative relative to TORC. The CETOP-2 algorithm penalty factor provides a high degree of assurance that CETOP-2 results are. conservative relative to CETOP-D despite approximations such as the use of a constant enthalpy tra_nsport coefficient or the simplification .in the treatment of the buffer channel.

Question 492.4 (A-5)

In the 3-D lumped subchannel modelling, how are the hot assembly and hot channel selected? How is it assured that the selected hot channel is the hottest channel that has minimum DNBR? During an operating transient, how does the model handle the situation where the hottest channel may move to another channel?

Answer When comparing CETOP-D to detailed TORC for a given range of operating conditions the location of the hot assembly and hot channel is important only in the detailed TORC model. The selection of the hot assembly and hot channel in detailed TORC is explained in Section 4 of CENPD-161-P. As a result of the comparison between CETOP-D and detailed' TORC, the inlet flow factor for the hot assembly in CETOP-D is adjusted to yield conservative or accurate DNBR

- predictions relative to detailed 10RC. (The inlet flow factor in 5-TORC was adjusted in the same manner,as described in CENPD-161-P).

4

1 TABLE I UNBR Sensitivity to Transport Coefficients in CETOP-D '

(Response to Question 492.3) .

VALUE OF DNBR

• DNBR DNBR TRAfiSPORT SEliSITIVITY TO SEriSITIVITY TO SENSITIVITY TO COEFFILIEtiT li g fl U u P m.

I 1 _J t >

All sensitivities are relative to a base JNBR of 2.1657 This DNBR was obtained using the following values:

Pressure 2250 psia Inlet Temperature 553 F Core Flow 100% of nominal Power 100% of rated (

li g= self generated by CETOP-D (enthalpy transport coefficient) fiU=[ ] (velocity transport coefficient) ,

f4P=[ ] (pressure transport coefficient)

• The sensitivity of the '1 to li compares g

D"BR's using constant gli values to the '

self-generated case.

L b

)

~

TADLE 2 ANO-2 Cycle 2 CETOP-D/CETOP-2 Transport Coefficient Values 7_

(_l (Response to Question 492.3)

CETOP-D CETOP-2 TRANSPORT VALUE VALUE COEFFICIENT ENTHALPY CALCULATED INTERNALLY .

(Ng )

- NOT VELOCITY - APPLICABLE (N) g PRESSURE * -

(Np )

rs.

x_/

=

• Note that in the code the pressure transport coefficient is given as Cg t

l t

9 f~r, L-]

6

4 ThisLadjustedCETOP-D model is then independent'of the actual location of -the

j-

hot assembly or hot channel within the core since it'has been tuned against 4 . (,)st -theihottest assembly in detailed TORC that could be limiting in DNBR.

ForL transients in which the hottest channel may. move. detailed TORC models

' used for the taning of CETOP-D cover all~ possible potentially limiting . locations ,

of.the hottest channel. .

I

-Question 492.5 (A-6)-

The CETOP code use's a prediction-correction method, as opposed to the iterative

' method used in the TORC, to solve the finite difference equations of the conser- .

t vation laws. - How is it guaranteed that- there is no instability problem? -

- Answer The prediction-correction method used in the eETOP-D and CETOP-2 codes is a non-iterative one-pass method. Therefore, there' are no instability -problems related to convergence.

Thousands of cases, covering the entire range of. operating conditions, have been 4

run comparing CETOP to TORC. Excellent agreement has always been obtained.

Note .lso that the tuning of the CETOP-D model, discussed in response to questions 492.4 and others, conservatively compensates for any small errors due to the differences in numerical schemes between CETOP-D and' TORC.

O . .

Question 492.6 (A-7)

The core inlet ficw distributions are determined from reactor model experiments-for CE type cores. Is the inlet flow split held constant during operating transients?

I Answer -

The hot assembly inlet flow factor (inlet flow split) is adjusted in CETOP-D i to be conservative for all conditions and held constant. This adjusted flow split ~can be different from the value found at any given assembly location.

For transients 'in which the inlet flow distribution mr y change significantly,

, the CETOP-D model is benchmarked against a detailed TGRC model which incorporates the more adverse of the initial and final inlet flow distrioutions as determined by reactor model experiments. The benchmarking of CETOP-D to detailed TORC'is discussed in response to Question 492.7 and the value of the flow split is discussed in response to Question 492.14.

The determination and use of the inlet flow split for CETOP is the same d

- as that described for S-TORC in CENPD-206-P.

1 a

O 3

k 7

4

- , - - -. v-, .

, w --.-r _ w, .e- .-- r , . i, - = , - - .r- , -

m . - - -- y , ,. - 3 ,9

Question 492.7 (A-8)

Provide comparison between the CETOP and TORC results that cover the whole 7-' spectrum of operating conditions. Provide an assessment of accuracy on

() the CETOP code. Justify any reduction in scope of this assessment from that provided in the T&H supplement to CENPD-170 with respect to the original CPC software.

Answer Figure 1 shows the comparison between detailed TORC and CETOP-D for ANO-2 Cycle 2 and other plants. In all cases throughout the range of operating conditions, CETOP-D calculates a DNBR lower than detailed TORC.

CEN-143(A)-P Appendix B Part 2 describes the accuracy assessment for CETOP-2.

As discussed in response to Question 492.15,a penalty factor on core power is determined from this accuracy assessment. The scope of the assessment is not less than that provided for CPCTH in CENPD-170-P Supplement 1-P.

The range of conditions considered are shown in Figure 2.

Question 492.8 (A-9)

In the CETOP-2, two correlations of curve fits used for void fraction calculations fit the Martinelli-Nelson void fraction model . However, there are discrepancies in the range of applicability of these correlations as shown below:

~

(l QUALITY RANGE OF APPLICABILITY Correlation TORC CETOP-2 CETOP-2 Coefficient Table B-1 Programming ALL's 0.01 < X <0.1 ALH's 0.1 < X <0.9 L._

Which is the right quality range of applicability? What is the pressure range?

Justify any simplifying assumptions applied in the CETOP-2 sof tware.

Answer The values of the quality ranges used in determining the void fraction correlation in CETOP-2 are the [ ] The values given in Table B-1 and on page B-7 of CEN-143 are incorrect. The correct implementation of the void fraction correlation is given on page B-26 of CEU-143(A)-P.

The pressure range for this correlation is the same as in TORC.

p.

V 8

~~T -~ ~

C Li T V- 1: . 4

% i}' F Figure 1 g

....4

y. ;+.' .-K . ;m

. ;*.  :: .y . * ;.,;';.;

o

._: ..: 1

,d - - b" e -

COMPARISON IN MDNBR BETWEEN CET0.P-D-(tut 1ED) and r:4

gr_:g:g;

2. . :

n: - . . . _ . . ,

= .. N DETAILED TORC riE-f a

...t~. ---*------1

. :.. ._.....-,..-...r.--.

,. _=-, Ei .

.:.t:c.+ .-- i-:::::_w- r 1-

._+. . . .q - .. -f:1. . 1. < ~. 4i=. .mw'l= 1 Ran9e of Operating Conditions i .:r :-.:. --. ".J'::

.- 6  %. .

j* L r. .z r.1.. :.. n 2:7.. 'i TC C - - i '~ ': 7-

_ -L'

- I :

.# t . i . ~ ' .. _: !? ~L' Inlet Temp. ( F) 465 - 605 --

.g . ; . . . . . . ..;..

3

=-c.2:1+-Ri-iN_:...~Ti System Pressure (psia) 1750-2400 -- :=4ggg=-::v= 'i.5:.i.:j

_.[-

3 .- q = ; z!.-- ~ assel Flow (gpm) 193200-475200 . ~ - 7:5.7:.

?'= "

1. . __i ' _ .i. Axial Shape Index .60 - 4.60 .g-:ii..M_ :bl.ji.+: . . =a. . a E ' "

o O SONGS 2/3 & Calvert Cliffs 1&2

.2 : =  :

+ ANO-2 50 -' -

m N10-2 Points in Table A-1 of v CEft-143(A)-P

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

r (OO ce e

b .

i  ::

5 '?

- n.

e o t--

es LM w* LJ t;I M

~

5o g J-M$

M~

?p*

e xs L. u _  :

a L.+ :s PJ

.?, I-~

us . - .

~ q=w

[ =n r:

L.. L.i

-~

2 :.::..

C aih-

.4

^1 J

t: .

r . :. .

,p -

t _. - -

V L:..  !

[,, , _ _ _

- ; i- > 1 '. !- 1- :! ; - _a;_-

. _-._ .. :;_ q .I .j:

f, j~;

L  ; ~!- DETAILED TORC 11DNBR

---

" j e

P00RDRIGINE

Figure 2 A

'V ,

Range of Conditions Considered for CETOP-2 Accuracy Assessment for Af10-2 Cycle 2 (AnswertoQuestion492.7)

Inlet Temperature e 465'F > Tp ; > 605 F Pressure e 2400 r.,i > P> 1750 psi Flow o 120% > F > 90%

O Axial Shape Index e +.60 > ASI > .60 DfiBR Range o 2.40 >Df1BR > 1.24 O

l 10 l

guestion 492.9 (A-10)

Provide justifications for choosing the Martinelli-Nelson void fraction model (j over other models such as homogeneous, slip flow or drift flux models. Is

'" subcooled boiling considered?

Answer for pressures below 1850 psia, the void fraction is given by the Martinelli-Nelson model. This correlation is used in CETOP exactl3 as in oir approved TORC code (CENPD-161-P) and is further discussed in the CETOP-D description provide / in response to questions on CEN-139(A)-P.

TORC includes a correlation to calculate subcooled void fractions for inQrmation only. The correlation is not used in ccmputing pressure drop or in design DNB analyses.

Question 492.10 (A-ll)

What correlation is used i the two-phase multiplier for frictional pressure drop calculations? Provice a con.parison of data and the result of your curve fits.

Question 492.11 (A-12)

What correlation is used for the subcooled boiling two-phase multiplier for frictional pressure drop.

fm Answer U

The Sher-Green and Modified Martinellt-Nelson correlations are used to determine the two-phase multipliers for frictional pressure drop calculations during local (subcooled) and bulk boiling conditions. These correlation are applied exactly as in our approved TORC methodology and are discussed in CENPD-161-P and in the CETOP-D description provided in response to questions on CEN-139(A)-P.

Question 492.12 (A-13)

Provide a comparison between the saturated liquid properties and the curve-fit results. What is the range of applicability of pressure?

Answer In CETOP-D, exactly as in the approved TORC code, fluid properties are based upon a series of subroutines that use a set of curve-fitted equations to describe the fluid properties in the ASME stean tables. Fluid properties are discussed in CENPD-161-P, and in the CETOP-D description provided in response to questions on CEN-139(A)-P.

O v

11

Question 492.13 (A-14)

In the calculation of core and hot assembly inlet conditions, a flow '

~' measurement adjustment term. MERR, is added to the coolant mass velocity calculation. Is this adjustment in the non-conservative direction?

(V) so, provide justification. If Answer The flow measurement adjustment term, MERR, is entered as a negative n"rber if "a decrease in the coolant mass velocity is appropriate.

Question 492.14 (A-15)

In the core inlet flow split calculation, the algorithm resuits in the same value of hot assembly flow saturation factor (FSPLIT) regardless of operating conditions such as ASI, primary pressure and coolant temperature. Justify thevalueof( 3 Answer CETOP-2 contains the capability for entering two flow split values for two operating ranges. For ANO-2 Cycle 2, a single value is used dver all operating space.

Therefore, FSPLITl = FSPLIT2' The Fspty value,( 3, represents the adjustment factor to ensure CETOP-D always caIculates a lower DNBR than detailed TORC over all operating conditions (ASI, pressure, temperature, flow).

Question 492.15 (A-16)_

' How is the value of power uncertainty factor of [ lobtainedforDNBR calculation?

Answer Thepoweruncertaintyfactorof[

casesofCETOP-DandCETOP2asshowni]nFigure3.results from the comparison It represents the penalty applied to core power in CPC to ensure that DNBP results from CETOP-2 have a 95/95 probability / confidence level of being conservative relative to CETOP-D. A similar factor was determined for CPCTH, the corresponding CPC algorithm for ANO-2 Cycle 1, in CENPD-170-P Supplement 1-P to ensure that DNBR results from CPCTH have a 95/35 probability /

confidence level of being conservative relative to BUL'_ cnd COSMO.

Question 492.16 (A-17)

What is the value of the addressable DNBR uncertainty factor, BERR1, used in thn calculation of heat flux at full power?

Answer n BERRl the addressable DNBR uncertainty factor, is calculated at the conclusion

() of the CPC software modification effort.

Phase II test report requested in Quest-ion 492.24.It can be provided along with the 12

, - . = ,__ _ - _ .

. ~ '

h -

. l r .

I -

? G _

O R I A T

l M _

i .

B R _

E, I '

R .

T ) 0'. _

P S 5 _

I 1 R _

D E .

2 V -

R 9 O .

O 4 R D n - .

o P

- i O R D t T U - s E G P e C I O u /

F T Q 2 E -

C o P t O T

S r E V e C 2 s w /

F

- n O P A O ( O T I E T C_ A R .

h _00 3

0 5

2 0

0 0 5

0 0

0 2 1 5 1

D=EEE

""  ; 1 ' ! l,

BERR1 was calculated in Cycle 1 by a combination of statistical and determints-tic methods. As discussed in CEtiPD-170 Supplement 1-F and CEti-35(A)-P rs (answer to question 222.129), CPC DNB and power distribution algorithm V uncertainties were determined by stochastic simulation. Detector noise, CEA position measurement errors, and certain processing errors were included in the simulation. The resultant uncertainties were then combined statistically by the root sum square (RSS) method with other uncertainties such as radial peak measurement errors and engineering factors. Other uncertainties including pressure, terrperature, and flow measurement uncertainties were treated deterministically by multiplication of individual components.

A numerical example of such e calculation was provided to L. Beltracchi of HRC following the uncertainty analysis audit of June 14, 1977.

For Cycle 2, BERR1 is being calculated by applying the more realistic statistical method, stochastic sirnlation, to calculate and combine CPC DNB and power distribution uncertainties, CEA position measurement errors, detector noise, processing errors and pressure, temperature and flow measurement uncertainties. The simulation technique used is similar to that described in CENPD-170 Supplement 1-P.

Engineering factors have been accounted for by increasing the fiDNBR limit as described in CEN-139(A)-P and discussed in response to Question 492.25.

Question 492.17 (A-18)

In the linear heat distribution calculation, the P2, P3, and P4 are defined g as the corresponding channel power relative to channel 2. Explain the V algorithm in the equations or, page B-9.

Answer P2, P3 and.P4 are only used in the form of ratios. Therefore, they can be normalized to any common value. The power in channel 2 is cbsen for convenience.

i Question 492.18 (A-19)

In the transverse momentum equation, which crossflow resistance correlation is

1. used in CETOP-27 Is the crossflow resistance the same between core region -

hot assembly gap and buffer channel - hot channel gap?

Answer T'

crossflow resistance correlation used in calculation of the core region -

hot assembly crossflow is the same as that used in TORC (Option 2, Section 3.4 ofCENPD-161-P).

The crossflow resistance appropriate for the buffer channel - hot channel gap is small. For the range of interest, the actual value chosen has a negligible effect on the DNBR, as si.own in CENPD-161-P. Therefore, for simplicity, this O term is set to zero ia ct'or 14

,r=-

Question 492.19 (A-20) .

How is the value of turbulent interchange constant obtained? Provide a (Vl sensitivity study of turbulent interchange on DNBR.

Answer The turbulent interchange constant (inverse Peclet number, .0035) was derived from cold water dye mixing tests. It was verified for 14X14 and 16X 16 assembifes from test data obtained at Colunbia University (see CENPD-162-P-A.) A sensitivity study of turbulent interchance on DNBR is given in Appendix F of CENPD-162-P-A. Both CETOP-D and TORC use the same constant as is evident by comparing Table 4.1 of CENPD-161-P and Section 2.7 of the CETOP -D description providedinresponsetoquestionsonCEN-139(A)-P.

Question 492.20 (A-21)

On page B-13, lines 3 and 6 "Section 2-11" and "2-12" should be "Section 2-12" and "2-13" respectively.

Answer, Correct Question 492.21 (A-22)

Q Justify the use of the Newton difference fonnula and Bessel's interpolation formulatoconvert[ 3 -node axial power distributions to [ ]pointpower distributions.

Answer '

The Newton difference - Bessel interpolation scheme is a second order technique.

It provides a better representation of the true flux shape than can be obtained by linear interpolation - extrapolation. The Newton difference-Bessel interpolation scheme is the Newton's divided difference formula *, adapted for use in the on-line CETOP-2 algorithm to obtain the required [ ] point power distributionfromthe{ ,) node power distribution obtained from the on-line POWER algorithm. A typical result of applying this technique is shown in Figure 4.

Question 492.22 (A-23)

Provide a comparison of the CPC transient calculation to Cycle 2 design safety analyses for the loss of flow transient, the comparison safety analyses should be based on (a) CETOP/CE-1 (b) TORC /CE-1, and (c) COSliO/W-3.

i O

• B. Carnanahan, H. A. Luther, J. O. Wilkes, Applied Numerical liethods, Wiley and Songs, New York (1969).

15

- 5 0

O 1 8

S T

O L 6 P

N O

S )

T I 1 R 2 H A G I

P 2 E M 9 H)

O 4 T '

C EE n

O N o RL O i ON 4 I t CI T s FM E U e R B u OO

. U I Q 4 R G R NF I

T o O(

F I m

_ S t -

I T D e C -

s A -

R n R -

E o F p

0'd s -

P e -

R -

L (

A I

X A .

2 .

- m 0

0 8 6 4 2 0 8 6 4 2 O e

2 1 1 1 1 1 i _

$t$ eW2 l:R< _

_cn t _

l  ;.4'  ;

\

Answer

,3 Comparisons between C0510 and TORC were presented in CEtiPD-161-P (Table 7.10).

U The CE-1 correlation is compared to W-3 in CEtiPD-162-P-A (Section 7.2). These comparisons apply to the DfiBR's calculated during a loss of flow transient analysis.

The response to question 492.7 provides a comparison between TORC /CE-1 and CETOP/CE-1 over the whole spectrum of operating conditions. As discussed in the response to question 492.7 CETOP calculates D!iBR lower than that calculated by TORC over the entire operating range. In addition, the response to question 492.27 provides comparisons of DNBR calculated by TORC and CETOP at the point of minimum DNBR during the loss of flow and CEA withdrawal transients.

A comparison of the CPC transient and design transient calculations for certain transients will be provided for ANO-2 Cycle 2. The design DNBR code will be CETOP and the NSSS simulation code will be CESEC. This comparison will be similar to the one performed for AND-2 Cycle 1 and will consist of five transients. For Cycle 2 the transients will be:

1. Four pump loss of flow

2. One pump coastdown from four pumps ru' ming

3. Full length CEA drop

4. CEA bank withdrawal from 1% power

5. Pressurizer spray malfunction O The results that will be provided are:

1. Traces of the CESEC analysis DNBR (calculated by CETOP) vs. time

2. The required trip time determined from the CESEC analysis.

3. The latest expected CPC trip time as simulated by the CPC FORTRAN Since the CPC FORTRAN Simulation code models CPC System calculational delays, the compa'rison cannot be completed until the CPC' software disk is generated. The results will be provided with the CPC Phase II Test Report requested by Question 492.24.

Que,s_tipn 492.25 (A-26)

Provide a comparison table of values for CPC data base constants based on statistical combination of uncertainties (SCU) versus the values and uncertLinty bands for the same constants without credit for SCU.

Answer The use of Statistical Combination of Unce:tainties (SCU) in treating systera p3rameter* uncertainties as described in CEN-139(A)-P affects the minimum DNBR (fiDNBR) limit in CPC and the various system parameter uncercainty factors in the ANO-2 Cycle 2 TORC and CETOP-D models and CPC DNCR algorithm (CETOP-2).

O V System parameters are those that describe the physical system and state parameters are those that describe the operational state of the reactor. State parameters and monitored during operation while system parameter:. are not.

17

O As discussed in CEN-139(A)-P Section 2, the deteministic approach would V involve applying system parameter uncertainties to the limiting subchannel in the CETOP-D model in the adverse direction. This is equivalent to assuming that all adverse deviations occur simultaneously in the limiting subchannel.

On the other hand, the statistical method of CEft-139(A)-P being used for Cycle 2 accounts for system parameter uncertainties by incorporating them into a re-vised MDitBR limit for CPC and the safety analysis. A best estimate CETOP-D model is then used in the safety analysis and in the derivation of the CETOP-2 DitBR algorithm and constants. The use of this model and the revised

!CNBR limit ensures to a 95/95 probability / confidence level that tha limiting fuel pin will avoid Drib if the predicted MDNBR is not below the MDNBR limit.

As a result of the analysis presented in CEN-139(A)-P, the MDNBR limit for This corresponds to approxi-ANO-2 Cy(cle mately 2 was increasedItfrom 3overpowermargin. 1.19 to 1.24.

is estimated that the effect of the system parameter uncertainties treated by CEN-139(A)-P (Table 5-1), and the wouldyiuldapenaltyofapproximately[2% rodoverpower bow penalty discussed margin. The net in Section 6.2,]if overpowermargin9uinisthus[ ].

Treatment of state parameter

• uncertainties in CPC is independent of this statistical treatment of system parameter uncertainties and independent of CEN-139(A)-P. The treatment of state parameter uncertainties is discussed in response to Question 492.16. The only imoact of CEN-139(A)-P on CPC data base constants is the change in the MDNBR limit to account for system parameter s uncertainties and the corresponding removal of dete;ministic system parameter uncertainties.

Questien 492.26 Explain how the application of SCU on the Cycle 2 differs from the uncertainty treatment in the Cycle 1 and its impact.

Answer Statistical treatment of uncertainties has been employed in the ANO-2 Cycle 2 analysis in two independent areas.

Thermal-hydraulics system parameter uncertainties were treated statistically as described in CEN-139(A)-P. Reponse to quertion 492.25 discusses the impact of such statistical treatment.

The treatment of state parameter uncertainties and other factors that need to be applied to the DNBR calculation by CPC is discussed in response to question 492.16.

G V

18

~

, , Question 492.27 (A-2)

("') Provide safety analyses based on an approved version of TORC /CE-1 for the loss of coolant flow and CEA withdrawal events.

Answer Minimum DfiBR (MDriBR) predictions with detailed TORC were compared to CETOP-D results for the loss of coolant flow and full power CEA withdrawal events.

Comparisors were made at the operating conditions ccrresponding to the point of MDNBR in the transient. The detailed TORC results in Table 3 indicate that the MDNBR limit (1.24) is not violated and that there is conservatism in the CETOP-D results relative to detailed TORC results.

Table 3 MDNBR Comparisons Between Detailed TORC and CETOP-D For loss of Coolant Flow and CEA Withdrawal Events (Response to Question 492.27)

MDNBR AMDNBR Conservatism Transient Detailed TORC CETOP-D in CETOP-D

(~~') ,

" Loss of Coolant Flow 1.240 full Power CEA Withdrawal 1.240

+

Initial conditions are defined in the Reload Analysis Report Table 7.1.8-1 for the Loss of Coolant Flow transient and Table 7.1.6-5 for the CEA withdrawal transient.

Question 492.28 (A-27)

Compare the initial values of peak linear heat generation rate (kw/ft) used in Cycle 1 and Cycle 2 safety analyses for loss of flow and CEA withdrawal event. How are worst case initial conditions determined?

Answer There has been no change in the peak linear heat rate (PLHR) LC0 of 14.5 ku/ft or the fuel centerline to melt trip limit (21.0 kw/f t).

The loss of flow and CEA withdrawal are DNB limites events; therefore, PLHR O does "ot eater into the a"airsis- The dirrerc"ce 4" ons overnower mer94a associated with the change from Cycle 1 to Cycle 2 can be directly converted into a PLHR increase during steady state operation if the plant operates at its LCO's.

19

3 For example, during Cycle 1, the PLHR calculated by CECOR has ranged from 9 kw/ft to 11 kw/ft. The plant has been operating with a COLSS power

( ,)

operating limit (POL) near 110% power. Theoretically, the PLHR could incrase another 10% before reaching the DNBR LCO. One could consider a DNB overpower margin gain of X% for Cycle 2 as a potential allowed increase in PLHR by X%.

Answers to questions 492.25 and 492.29 discuss margin gains for Cycle 2 which can be substituted for the "X" in the above paragraph.

Question 492.29 (A-28)

Provide a quantitative assessment of DNBR margin (and equivalent power margin) gained as a result of proposed methodology changes for ANO-2 Cycle 2 versus ANO-2 Cycle 1. The assessment should include a tabulation of the individual components of the gain (e.g., use of SCU, CETOP/CE-1 vs. COSM0/W 3, etc.).

Explain the impact of the margin gain on plant operating limitations.

Answer The comparison of TORC to COSMO was presented in CENPD-161-P (Table 7.10).

A comparison of the CE-1 critical heat flux correlation to W-3 was presented.

in CENPD-162-P-A (Section 7.2). For past reloads, we have seen that replacing COSM0/W-3 with TOP.C/CE-1 provides an overpower margin gain of ( 3 As shown in the response to Question 492.7 CETOP-D calculates DNBR lower than

,O that calculated for TORC throughout the entire operating range. Therefore, U

use of CETOP results in no margin gain relative to TORC.

The margin relating to methodology changes in the treatment of system parameter uncertainties (CEN-139(A)-P) is discussed in the response to Question 492.25.

Any margin gain from TORC /CE-1 or SCU will balance increased radial peaks for Cycle 2 or allow wider ranges in axial shape , temperatur , pressure or flow before reaching a COLSS limit or CPC trip. However, these ranges are limited by LC0's which prevent operation beyond the bounds of the safety analysis and any increase in margin to trip is reflected in that analysis.

None of these changes have affected the trip criteria for ANO-2. The fuel centerline melt limit remains 21.0 kw/ft. The DNBR limit for Cycle 2 will be the approved limit for the CE-1 correlation with adjustments for rod bow penalties and . system parameter uncertainties as described in CEN-139(A)-P.

.Q v )

20

+

1 Opestion 492.43 l l

j ls there a typographical error in the momentum equation (1.7) on paqe 1-4 of the .

CETOP-D topical ? Ii :t, verify it. l Resno_nse_

There is an error. Equation (1.7) of Reference 1 should read: l i

1 8

j -Fj dx + p jdAg - gAj pj dx + p Aj j - (p Ajj+ 3x P b dx) =

ii l

. (

i i -mj ug + (m ujg+'3 mj g u dx) -w'jj ju dx + w'jj gu dx + w g u*d- .

1 i

i

> . s e  !

b 4

I 1

I I

i i

i 48-1 l O i i

i I

, question 492.49 In the axial momentum control volume representation as stIown on both in Figure lunped 1.3 on page channel 9 1-14 of the CETOP-D topical, the vector oirections and averaged channel are pointing downward.

of p;dAj?

Is it an error BeSR9ESS.

Yes, the directions of P dAg j in Figure 1.3 should be pointing upward. The corrections are shown in Figure 49-1 e

1 f

O 49-1 i

i J

< -. -. - , -- e - ,- . - - , . - - , . , ,

, m;u;+~ m;u;dx [

. , P:A; + d_ p A;dx

. , L _ P_ _ _ _ .

3 V I i ll h

e 1

! i CONTROL / I

/ -

lr F*-li g i w;;u'dx I

g VOLUME l i i I .

Y i, Cli ANNE L il 1- I CilANNEL j F;dx i  :  ; I I gA; p ; dx  ;  ; I dx '

I '

l 1  ;

+

l Wji jU dx .

I I_I' _

w;;U;dx ll Pi dA; l l

m;u; i I

! = qA f

_ _ g= T ~~- J I

i y BOUNDARY SUBCHANNELS PiA; (A) CONTROL VOLUME FOR LUf.iPED CilANNEL m u; + 3 m;u;dx O

0*

A__.{__  :

I l 1 & '

p A; + O_. p;A i dx CONTROL ..

VOLUME l I_.-t- w;;u'dx p I ,

( f idx .

CifANNELi CHANNELj i

, j gA;p;dx dx i

l 4-

{ l w[;u;dx l

<l I ,

f p;dA; ii"i

!  ! m;u; 1

I Pi ^i I i

L ._ _A Y f }

j __ _ _

,i I .

] (G) CONTROL VOLUME FOR AVERAGED CilANNEL 1.0 0 Figure 49-I j

CONTROL VOLUt.1ES Foil AXIAL t.10MENTUT.1 EOUATION t, i 4

2 P00RORBIR l

l l

Qupstion 492.50 Is there a typographical error in the Dittus-Boelter Correlation (2.2) on page 2-l'

$. of the CETOP-D topical? .

Responst There is a typographical error. Equation (2.2) should read:

(K) (Re)0.8 (pp)0.4 h=ke where K is the coolant thermal conductivity.

e O

I O

50-1

t i.

1 i

e  ;

i l

l s  ;

i-Question 492.52 1

I i

ls there a typographical error in formula 2.9 on page 2-3 of the CETOP-D topical

! for the calculation of the coefficient B 2?  !

I i l Response Yes, the equation for B2 should read:

-3 p + 2.618 x 10 -7 p 2 -6.893 x 10-12 p, 3

! B2 = -1.060 - 1.194 x 10 1

i

?

}

1 l

9 f

l l

I 1

,1 k

i O 52-1 i

i

_ _ _ _ _ _ - . _ . _ _ _ - . , . _ _ _ . . . - _ _ _ _ _ _ _ _ . . . ,,, _._. - ._. . . . - . . , , - _ . _ . _ _ . - - , . _ _ _ _ _ . , ~

,questi_on492.53 Is the. algorithm of calculating the subcooled boiling and two-phase flow pd friction factor multipliers on pages B-29 through B-33 of CEN-143(A)-P the same as that described in Tables 1 and 2 on pages 2-8 and 2-9 of the CETOP-D topical?

Response

Yes. The generic algorithms of calculating the friction factors in CETOP-2 and CETOP-D are the same. However, the fonn of the correlation of the sub-cooled boiling and two-phase friction factor in CETOP-2 differs from that used in CETOP-D and TORC. The differences betwe6n CETOP2 and CETOP-D are as follows: ,

(i) CETOP-D calculates the two-phase friction factor multiplier at mass velocities of greater than and less than .7

• 106 (Table 2.1 on page 2.8 of the CETOP-D topical). The option of calculating the friction 6 2 factor at mass velocities of less than .7

• 10 lbm/hr-ft does not exist in CETOP2. This is because the CPC operational range on flow is greater than 90% design flow. The mass velocity will thus never Q be less than 2

• 10 6 lbm/hr-ft2 . A two-phase friction factor for mass velocities less than .7

• 10 6is thus unnecessary.

(ii) Some simplifications of the CETOP-D correlation are made so that it can be used in CETOP2 with less computer running time. The simplifications replace exponential terms with polynomial fits as shown in Table 53-1.

The accuracy of these simplifications, as individually stated in the last column of the table, results in.no more than a [ ]overall error in the friction factor.

All the above named simplification errors in CETOP2 are covered by the penalty factor between the CETOP2 and CETOP-D codes.

O G

53-1

TABLE 53-1

_SU_BC001.ED _B0lll_t G__AfiD TWO_ P_IIASE_ FRICTION FACTO,R CURVE _ FITS .-

TERM CORRELATION

• ACCURACY (%)

O x,B B

2-1 XB.,a-..

B 4 .

i  :

i .

l x(B3 .48) 4 B2 . G 1425) y(B3 .6099) ,

'B4-(.11526) 1 1+7X3/4/g 1+X**

(for 0$4 4.02) t O 1+7X3/4/g 1+X (for .02< t 6.20)

ItX 1+7X 3/4 4 (for .20 <0 4.40) p -0.2

" *** 1 q g =0.25 I

I 2/3

V _ -l _ _

• These equations are used in section 2.13 of CEN-143( A)-P

.** Gp = .0036 G and 4 = local coolant quality n n

• • • q g

= .0036 - 4 W 53-2

, ques _ tion _492.54 x Provide a figure showing the noding scheme on the finite difference of k conservation equations. What kind of finite difference techniques is used for the continuity, energy, and axial pnd lateral momentum equations? Is it forward or backward difference, or a mix of both?

Response

The noding scheme used in CETOP is illustrated in Figure 54-1. CETOP genera tly uses both forward and backward finite difference schemes.

There is a typographical error in equation 3.3 on pages 3.1 and 3.2 of CEN-139(A)-P. The correct axial momentum equation should be:

pg (J) - p4(J-1) q

/ u.' - u. .

3-A j =-F$-A$ g p4(J) -J wjj y g" (3.3)

. lu.' + u. u. - u )nJ 2u g w$ 3 + J~ + (~l2 "i ,

2 O

Section 3.1 of CEN-139( A)-P shows the finite difference form of the conserva-tion equations used in CETOP-D. The continuity equation and the axial momentum equation utilize a backward difference technique while the energy equation uses a forward difference technique. The lateral momentum equation relates the

" lateral" pressure difference to the fluid properties and flow conditions of the channels of interest. Thus, this numerical technique can not be categorized as either forward or backward difference.

T

• Typographical error in CEN-139(A)-P.

i O \

54-1

)

(] O (])

.a FIGURE 54-1: FINITE DIFFEREt;CE fiODAL SCHEME FOR THE SOLUTIOTl 0F THE CONSERVATION ACTIONS 1 2 3 4 m)(J+1) m)(J+1) A"3(U[I) "4(J 1)

J+?

(1,J+1) ( 2,J +1 ) (3,J+1) '

(4 J+1) ,

W1250 ) "23 ) "345d+ )

A A m)(J): m2 IU) ^ *3(J) m4 (J)

J  : e -

y (1.J) (2,J) (3,J) '(4,J)

" ~

---o- >  :.

"12(U) "23 (,U ) "34(U)

A m) (J-1) f m2(U-I) "3 (U ~I ) "4IU I)

J-1 - - -

(1,J-1 ) l

( 2,J -1 ) (3,J 1) (4,J -1 )

- -c- - --n. y w w l2(J-l ) 23(U~I) "34(J-l)

I l

1

f L)

Question 492.55 -

As stated on page 3-4 of the CETOP-D tqpical. *The success of this non-iterative, prediction-correction method lies in the fact that the lateral pressure difference, P) (J) - P2 (J), using the " guessed" diversion crossflow, W is a good approximation // However, the diversion crossflow is guessed based on the assumption of no lateral pressure difference downstream, i.e., P) (J+1) =

P2 IU+I) is the guessed diversion crossflow as well as lateral pressure difference prediction accurate? Justify (quantify) it.

4

Response

Yes. The guessed diversion crossflow and the resultant lateral pressure difference are accurate enough to yield an accurate solution of the axial s finw which has a strong bearing on the calculated overpower margin. This fact is shown through the demonstrated good agreement between the O c " "d t"* i tere t' v" S-' "' c d** -

In this question we are implicitly asked to justify the use of a predicted crossflow based on the assumption of zero lateral pressure difference at the .

node exit. This predicted cross flow at J+1 is used to calculate the lateral pressure difference at the exit of node J. The lateral pressure difference at node J is then used to calculate the corrected crossflow at node J.

The accuracy of this predici: ion-correction scheme will be much better than the accuracy obtained by using an assumption of zero lateral pressure difference at each node exit. The error in the predicted crossflow (J+1) will depend almost entirely on the relative importance of the lateral pressure difference term (J+1) in the crossflow equation. If this term has only a minor effect on the crossflow then the predicted crossflow (J+1) will be accurate. The error in the predicted crossflow will result in the same percent error in the calculation of the lateral pressure difference at node J. The final error in the corrected cross-flow will depend on the percent error in the lateral pressure difference and Q the importarice of this tenn in the crossflow equation at J. Therefore the corrected crossflow will have a much smaller error than the predicted crossflow.

l 1

l l

55-1 i

x

• an exampje.

() Assume the arn exclusion of then translates o troximately the lateral pressua 30f e Assuming that a 30% errror in the predict also the re difference tenn a results in exclusion of ror ined thecrossflow lateralat node J+1 ccounts for corrected is crossflowa 30% error, andthe lateral presspressure e This error diff there is a 30 ure difference terma t be less then 0.5% (ismall relative to

.e.,

e axial only a 9% error. % error in this 9% x 5%). flow (< 5%), the Finally, since the at.. ten ode J error in the crossflow.

mass flow will The numbers the accuracy ofused inexample this are in the predicted andthe prediction-cor only intended at each node. used method will be flows In virtually all cacorrected will general. crossflowsrecti to illustrate The C in using assumption accuracy is the significantly more ses however, the predid conditions of zero lateral presaccurate a than just methods found in COBRobtained ess with much culatin cross-l The tuning of A-IIIC or TORC. computing timeThis than thsure d increase compensates for thethe CETOP-D model e iterative errors in crossflow theto the detailed TOR calculation.C model vely conserv i

V e

Question 492.56 in question 492.5, the concern over numerical instability is the propagation C and amplification of error throughout the subsequent calculations. As stated by Carnahan, et al,* "For a predictor-corrector method, if the corrector is iterated to convergence for each integration step, then any error in the observed solution can be attributed solely to the corrector. Hence it is sufficient to study the propagation of error by corrector equation alone.

When the corrector is not iterated to convergence, but is employed just once,

'the stability analysis must include errors generated by both the predictor and corrector equations." A stability analysis is required accordingly.

• B. Carnahan, H. A. Luther, J. V. Wilkes, " Applied Numerical Methods",

John Wiley and Sons, Inc.

Answer The W ediction-Correction scheme used in CETOP is a fast method for solving the crossflow equations in a marching solution. Although it is similar in some respects to the Predictor-Corrector numerical scheme described in Carnahan et al, the two methods are different.

The Predictor-Corrector scheme described in Carnahan et al is a general purely numerical scheme for linear difference equations. The example in Carnahin et al which is used to illustrate the stability analysis of the predict 1r-corrector method uses a fourth order Milne method to integrate the difference equation. The derivatives used are assumed to be linear.

The results of this analysis however are not applicable to the Prediction Correction method. The equation which is being solved in CETOP using the Prediction Correction method is the conservation of mass equation. i.e:

F$ (J) = Fj (J-1) + S Ax The derivative term in this equation is defined to be the crossilow dFj-(g)  ;

dx

-W ij(J) v The solution to this equation is not a linear function of F. It is a very complex solution involving a simultaneous solution of the noss, axial momentum, 56-1 j 1

1

\

and transverse momentum equations. Since' this solution is non-linear it does not lend itself readily to an analytic stability analysis. However,

( ])- several observations can be made which will indicate the stability of the prediction correction method.

As stated in the response to 492.5 we have run thousands of cases comparing CETOP and TORC and in all of these cases the accuracy in OPM has always been excellent. If a serious instability in the prediction correction method occurred it would show up directly in the value of the OPM, over the ranges of operating conditions of interest.

1 4

0 a

4 P

56-2

r^x Question 492.57 9

The enthalpy rise factor is not applied to the hot chcnnel enthalpy calculation as is normally done. Rather the enthalpy rise factor is included in the " modified effective rod diameters". Are these treatments equivalent? If so, prove the equivalence.

Response _

The,enqineering enthalpy rise factor accounts for as-built variations in U-235 rod cont.ent. For the TORC code, this factor is applied by increasing the heat input of the. rods adjacent to the hot channel (Reference 2). This treatment has been reviewed and approved by the NRC. Similarly, for CETOP-D the enthalpy rise factor is applied to the heated effective rod diameters in the hct channel to increase the heat input. These effective rod diameters are used in CETOP-D to increase the rod heat inout, not the geometry of the rods. Sir ce the rod heat inputs in CETOP-D and TORC are adjusted by the same amount, the treatment of the enthalpy rise factor is equivalent.

As a result of the May 7,1981 meeting the NRC requested C-E to provide additional justification that tha treatment of the enthalpy rise factor is equivalent between CETOP-U and TORC. This is given below.

The equivaler.co is proven by showing how the heat input into the hot channel

, is calculated to be the same for both codes, with and without the enthalpy n rise factor. Using Figure 57-land the information in Section 4.3 of Ref.1, the heat input into the hot channel is determined by the following equations.

Without the Enthalpy Rise Factor:

9 C * (IRl'1 5

• fR2(4)*f R35 (4)) 9" I rdAZ Z (1) 9 (I R(4) D(4)) q" F HAZ E OP Z (2) where:

FR (4) " I Rl# I I4)#IR2'2(4)*IR35 3(4) (3) c)(4)+c2(4)+C3(4)

D(4) = (cj(4)+c2 (4)*'3(4))d (4) and:

fRj = radial power factor for rod j c)(i) = fraction of rod j depositing heat in channel i O d = red diameter. ft.

AZ r distance between nodes in model, ft.

ii' e core average heat flux, Btu /hr-f t 2 57-1

q(i) = heat deposited into channel i. Btu /hr b_ F Z

= axial power fador at locadon Z Equations (3) and (4) are developed based upon equations (4.11) and (4.15) of Ref. 1. Substituting equations (3) and' (4) into eq. (2), it is readily seen that equations (1) and (2) are the sane, therefore, the heat deposited into the hot channel is the same for both codes without the enthalpy rise i

factor.

When the enthalpy rise factorHf is applied, the fractions of heat from the rods adjacent to the hot channel are nultiplied by f . In TORC the fractions c are easily adjusted in the input. In CETOP the fraktions c are not contained in the input, however, the effective radial power factor F (4) and the effective rod diameter D(4) are alternatively adjusted in the inpht to account for changes ,

in c. These adjustments in CETOP do not affect the local heat flux used in DNBR l calculations or the hydraulic dimensions of the hot channel. The following equations are used with the enthalpy rise factor. t 1

With the Enthalpy Rise Factor:

H 1(4)*IR2 fHC2 I4)*f R3 H 3(4)) 9" FZ'""

fC q(4) = (f RifC TORC q(4) =(FR (4) D(4)) q" FZ dZ (6)

CETOP O where:

Fg (4) = Rif fCI4)+f Hl R2fC C H 2(4)+I R3 H 3(4) (7) f gt)(4)+f CH2( )* 113(4) 6 (8)

D(4) = (fgc)(4)+fg2 t (4)+I H3 C (4))d If equations (7) and (8) are substituted in equation (6), equation (6) will be-

'come the same as equation (5), therefore the treatment of the enthalpy rise f actor in both codes will be the sane for the hot channel.

The remaining channels surrounding the rods of the hot channel will also be affected by the enthalpy rise factor fH . The procedure for applying gf in the remaining charnels is the same as that applied to the hot channel. Therefore, the treatment of f gin these channels will again be the same in both codes.

1 57-2

~

\- [ .[\ [\ h J %) (G %J

}J '

>a ,

) .

OOOO rs,e c arI,at.g i rise Sartor are atiled + rod .1.,1,nnd 3. Tie ~ ~ s inaic<1e w!s.rl, cisa.unch ~e. al%ied b y Sn.

~,

~

e,.3 P00RBRISlWJ1

~

q gues, tion 492.58

(_ / - -

The heat flux factor is applied to the final DNBR calculations only. The effect of heat flux factor on coolant condition is not included. Has this effect been taken into account by the enthalpy rise factor?

Epsponse Yes. As explained in ' Question 492.57, the enthalpy rise factor af fects the coo,lant condition in the hot channel by increasing the entire rod heat input.

The application of the. engineering factors on pitch and clad 0.D. also has an effect on toolant condition. The local heat flux factor is used to take into account the random variations in pellet enrichnent, pellet density, pellet diameter and clad outside diareter. This heat flux factor is appropriately applied only to the final DNBR calculations and does not affect coolant condition.

e P

O o

V 58-1

'\ .

,Que_s_ tion 49_2.59 gr. Is there a typographical error on page 2-6 of the CETOP-D topical in the equation for Tong-F factor for non-unifonn heat flux CHF correction? If not, justify it. Also provide a deriviation of the F-factor fomula (Fk) used in CPC software algorithm on page B-23 of CEft-143(A)-P.

Response

There is a typographical error. The Tong-F factor on page 2-6 should read:

X (J)

FS(J)= C(J) fq"(x)e-C(J)(X(J)-x) dx (J) (1-e

~

) b)) U The F-factor used in the CETOP-D topical is derived from the correct Tong F-factor fomula. The derivation of the F-factor used in Cell-143(A)-P follows:

X(J)

O V

FS(J) = ,,

C(J) q (J) (1-e-CW) X(J))

g

[ q"(x)e -C(J) (X(J) x) dx

., = h WHERE J = N0DE OF INTEREST =[ ].

q (J) = LOCAL llEAT FLUX AT NODE J

-I ft C(J) = 1. 8 (1 - 4 CHF)"' _

for the CE-1 CHF correlation 6

(G/10 ) 0.478

, 4 = QUALITY AT THE CHF LOCATION 2

G = LOCAL MASS VELOCITY AT CHF LOCAT10!1, lb/hr ft X(J) = AXIAL HEIGHT OF NODE J Q AX = N0DE HEIGHT FOR[ ] AXI AL N0 DES 59-1

'rQ . l v l 1

q({-1);______ _ _ _ _y ,

q (t )

I l

' - I u l 4 (x). I i I i i i i I  !

I I i i ,

' I o .

x(1-1) x(L) X(J) y Il Increments used to approximate q FIGURE 59-1 Let the interval between 0 and X(J) be broken up into equal increments -(Figure 59-1)Then assume a linear variation of the heat flux within each increment. Thus in the increment preceding f-1 .

4 (x) = q (E-1) + h ~x-l{ 9 (E) - 9 (#~1) f n n

=q (R-1) +

~

L #" ~ *(t~l) - ~

( )

_x -

i

=

iu q (f-1) - x(t-1) j- _

t-M \,. tj-

-11)x @

n v

=a g

+bxg 59-2 i

4

_ . . , _ , . _ . , - . . - ~ _ . , , --

Usingequation@wehave x(t)

[q"(x)e-C(J)(X(J)-x)dx= (ag+ bgx)e -C(J)(X(J)-x)dx x(t-1) x(t-1) t) * ( E()

=e -C(J)X(J) ae C(J)x dx + bg xe (U)*dx g [

x( -1) x(1-1)

(J)x

=e -C(J)X(J) age , b ge ( )* (C(J)x-1)

C(J) (C(J))

x(g,3)

C

,, c_-C(J)X(J)

~

C at e (J)X(1) e (J)x(E-1)

C(J)

O b

+ eC(J)x(t)(C(J)x(t)-1)-e C(J)x(E-1)(C(J)x(L ,)-1)

_TJT C /

a Inserting equation @into(I)the form of the F factor at node J becomes':

h FS(J) =

C(J) ,

e'(U}*(U} f a*(eC (J)x(t) e C(J)x(R 1);

q (J) (1-e-C(J)X(J)) C(J) t1

+ C C CJ) e (J)x(t) (C(J)x(t) -1) -e (J)x(E-1) (C(J)x(t-1) -1))

o 59-3

/

Rearrangingequation@

O

\,,,'

J C

FS(J)= 1 a e (J)x(t) e ( )*(**I

- + t e ( )*(E)(C(J)x(t)-1) t=1 ( f C(UT

-e (U)*(E~I) (C(J)x(t 1) -1) q (J) (eC(J)x(J) ,3) @

In both CETOP-D and CETOP-2 the F-factor is not calculated at the first node, Therefore, FS(J = I ag e }*(} -e (J)x(L 1) +

J) e ( ) ( ) (C(J)x(t)-1) o .e C -1>

(mJ> . 1).1) g<m)<e meo> .

1) e O

59-4

i

! Changing to the nomenclature of the equation for the F factor i in section 2,10 of CEN-143(A) -P '  !

4 i

Where ,

w l 1 4.

! e 4

Therefore p.

l f

i  !

I i

i l n h i And thus ,

i ~

i 1

4 t i @ l I

f i

i I

i l

O 59-5

i O Since [ .],FS(J)becomes 7

1 i

IJ 1

}

I e

13 4

i M

3 Where J

4 9

i

~ @

i -

y-- w we 1

O I

59-6 l

. . , . , - . , - , - , - . , , . , .,.,-,---.n.n,-,-, , , , _, , , - - n,-, - . , - - - - - , - . . - - . . - - , - - - - - - . . - - . , .-- - - , - _ - - - - - - - - - , , , , - - - - -

\

Comparing equations @ and @ and' equations @ and @, we obtaia:

i With the above equations, and designating Fk forFS(J),equationh changes to the following form:

1 1

!O E

N After equation insn(6 becomes: ting the engineering critical heat flux factor (SKECDK)>

.M 4

~

^

for k =[ ]

O 59-7 1

l 4WW V

Ques,t_ ion 492.60 3 Does minimum DNBR always occur in the subchannel next. to a guide tube? If the answer ~is no, justify the use of a pseudo-hot channel modeled for the guide-tube subchannel for CPC analyses. If the answer is yes, prove it. Also explain how the guide tube subchannel boundary is formed. What is the effect of different guide tube subchannel boundaries on DNBR?

Response-

~

i

~

,_s The type of boundaries usad for the guide tube channels in detailed TORC,

() CETOP-D and CETOP-2 rodels hre shown in Figure 60-1.The boundaries are formed by straight line segments connecting centers of adjacent rods and guide tubes.

This is conventional industry-wide practice and, for CE, is justified by the fact that the DNB data (Ref. 4) upon which TORC /CE-1 methodology is based, were correlated by consistent use of this convention. Use of a different convention for guide tube subchannel boundaries would require a DNB correlation based upon that different convention.

As a result of the 11ay 7,1981 neeting the NRC requested a description of what type of pin - power distribution, used it our detailed TORC analyses could shift the location of MDNBRL ] This description is provided below. -'

r-(~h .

%-)

60-1  !

~

.~

' For ANO-2 Cycle 2 and other thermal margin analyses using CETOP-D, MDNBR has al ways occtred in corner guide tube channels in detailed TORC. Using power distri-butions derived for present fuel management schemes, the likelihood of finding a detailed TORC case which shows the MDNBR in a location other than guide tube channel is very small when using the CE-1 DNB correlation. If a case is found, the bencinarking of CETOP-D against the TORC results will still provide conserva-tive DNBR predictions in CETOP-D.

O O.

60-2

s v

e I g E g

~ l

. \

f , I.

a

~ ~~ l , , .s 1

0 -

{.-

-lv\ .).

v I #

s . 1 Nu -

. , ~%~ -J &

J O m u,e c0.n cosae ,use suuaiaooe, couooeries 5

60-3

Rod Radial Power factor

,o.,

(,,/ .

Channel Number I

g..

j

~

%.) \ f s.

)

w g..._..__ - .

e es O

v

[sure60-2: Ouadrant Pin-Pouer Distribution of Center Assembly in Core 60-4

Qugs_ tion 492.61 L

p!

i In Table 5.1 of the CETOP-D topical a comparison is made for the CETOP-D results using

~

inlet flow factors of L J and 0.8. Explain T, uhy the differences in the MDflBR's are whereas the differences are about around[ ] percent for cases with ASI 7 percent for cases with ASI of L of[ 3a:nd C J

[,

1

Response

As seen in Table 5.1, a change in the inlet flow factor in CETOP-D produced a significant change in MDilBR for the C- ]ASI. This change is expected, since in detailed TORC the calculatio' of MDNBR was also found to be quite sensitive to changes in operaling conditions, (Inlet Temperature, Pressure Flow, etc.),

when using theE JASI. This large sensitivity has been discussed in Section 3.1.1 of Reference 3 and in the response to Questions 2.b and 2.c of the Battelle questions (March 27, 1981). -

9

's -

61-1

\ .

,uestion Q 492.62

Expand the safety analyses comparison for AN0-2 . Cycle 2 in response to questions 492.22 and 492.27 to include comparisons to CETOP-2 and COSMO.

In particular, the comparison should be made among CETOP-2, CETOP-D, COSMO and TORC code for the loss of flow and CEA withdrawal events with the following initial conditions:

ASI = -0.6 Coolant Inlet Temperature = 560 F Primary System Pressure = 2200 psia

-Flow rate = 60%, 80%,100% and 120% of rated flow

.The initial power level for each event is such that the minimum DNBR calculated by the CETOP-D is 1.24. Tabulate the value of PLHR (peak linear heat rate), the transient time and the value of minimum DNBR calculated by each code for each event and initial condition.

Response s As discussed at the May 7,1981 meeting, the conditions listed above are outside of the LCO bands for ANO-2 Cycle 2. In particular, the ASI of -0.6 is beyond the + ,.3 limits of proposed Tech Spec 3.2.7, the inlet temperature of 560 F is beyond the limit of Tech Spec 3.2.6, and any flow rate below 100%

of rated flow would violate Tech Spec 3.2.5. Therefore, a loss of flow or CEA withdrawal event initiated from these conditions wculd be beyond the scope O' of the Cycle 2 safety analysis.

Minimum DNBR (MDNBR) predictions with detailed TORC were comoared to CETOP-D results for the loss of flow and full power CEA withdrawal wents is, response to Question 492.27 (CEN-157(A)-P). The comparison between detailed TORC and CETOP-D was made at the operating conditions corresponding to the point of MDNBR for the transients presented in the Reload Analysis Report (Section 7.1.8 for the loss of flow and Section 7.1.6 for the CEA withdrawal).

The initial conditions are given in Tables 62-1 and 62-2 . respectively, Additional MDNBR predictions for identical conditions were made with COSM0/W-3 and CETOP2/CE-1 to compare to the detailed TORC /CE-1 and CETOP-D/CE-1 results reported in the response to Question 492.27. The er 'oarison is shown in Table 62-3.The predictions for COSM0/W-3 were e lated based upon past COSMO studies. The detailed TORC results ir ute that the MDNBR limit (1.24) is not violated. Also, Table 62-3 shows that CE10P2 calculates a more conservative MDNBR than CETOP-D or detailed 10RC.

l l

l n

v j

I l

l 62-1 1

In addition to the comparison presented in Table 62-3, a comparison of CPC transient and design transient calculations for five transients will be (q

! provided for ANO-2 Cycle 2. Details of the information to be provided were presented in response to Question 492.22.

A set of 4080 initial conditions were investigated for the loss of flow analysis in order to calculate the COLSS and CPC setpoints for ANO-2 Cycle 2. By choosing the setpoints to bound these 4080 cases, we provide assurance that a LOF from initial conditions within the LC0's will not violate the DNBR SAFDL.' Therefore, there is no single " worst case" set of initial conditions for the loss'of flow.

The ' case presented in the license submittal is only typical of those cases which approach the DNBR SAFDL. For the particular illustrative example, the initial peak linear heat rate was 13.2 kw/ft. The case presented in Section 15.1.5 of the FSAR had an initial peak linear heat rate of 13.4 Lw/ft.

Since the initial conditions, especially axial shape index, for the Cycle 1 and Cycle 2 cases were different, the PLHR's are not directly comparable.

The CEA withdrawal analysis case preser;ed for Cycle 2 had ar, initial PLHR

, of 14.8 kw/f t while the Cycle 1 case haci an initial PLHR of 13.4 kw/ft.

In all cases, LOF and CEA withdrawal for Cycles 1 and 2, the initial power level was 103% rated thermal Power and therefore the core average LHR was 5.71 kw/ft. Initial PLHR data for these four cases are summarized in Table 62-4.

Initial PLHR for a transient analysis depends not only on the DNBR margin available but also on the particular set of initial conditions chosen. Many sets of initial conditions could be chosen which correspond to the same initial DNBR margin but very different initial PLHR's. The PLHR is limited by the LCO of 14.5 kw/ft independent from the DNBR LCO.

The limiting linear heat rate transient case for each event was not presented in the Reload Analysis Report for the following reasons:

a. The CPC local power density trip prevents violation of the PLHR SAFDL since CPC is able to measure and project all parameters of interest,Fxy, Fz, and power level,during a transient, and

b. the degradation in PLHR margin during a LOF or CEA withdrawal event is small compared to the degrad: tion in DNBR margin.

The LOF case presented in the Reload Analysis Report shows no noticeable increase in core power, heat flux, fxy or Fz during the transient. Therefore, the PLHR will not increase above the LCO limit of 14.5 kw/ft.

4 The full power CEA withdrawal case presented in the Reload Analysis Report shows an increase in core power from 103% to 124%. For that case, Fz increased by 14% while Fxy decreased by 125. Therefore, if the PLHR initially were at the LC0 limit of 14.5 kw/ft, its peak during the transient would be no greater Q, than 17.8 kw/ft.

62-2

TABLE 62-1 s.J KEY PARAMETERS ASSUitED IN THE ANALYSIS OF UNC0tiTROLLED CEA WITHDRAWAL FROM FULL' POWER Pa ram.?ter Units FSAR Cycle.2 Initial Core Power (f1Wt) 2900 2900 Level Core Inlet Coolant (F) 556.7 556.7 Temperature Core itass Flow Rate (106 1bm /hr) 116.2 116.2 Reactor Coolant System (psia) 2250 2200 Pressure

.;, team Generator Pressure (psia) 923 939 floderator Temperature (10-4ac/F) +0.5 to -3.5 Coefficient +.5 O)

( Fuel Temperature 0.85 .85 Coefficient Multiplier Minimun Available CEA (10-2ac) -5.4 Uorth on Trip -5.4 Steam Bypass System -Automatic flanual Feedwater Regulatin9 Automatic Systen Autonatic fiaximun Reactivity (10-4ao/sec) 7 .5 Addition Rate S

nj 62-3

i TABLE 62-2 KEY PARAMETERS ASSUMED If4 THE LCSS OF COOLANT FLOW AllALYSIS

\

l Parameter Units FSAR Cycle 2 Initial Core Pcwer f1Wt 2900 2900 Level Initial Core Inlet *F 556.7 556.7 Coolant Temperature Initial Core 11 ass 106 lbm/hr 116.2 116.2 Flow Rate Reactor Coolant System psia 2250 2200 .

Pressure Moderator Temperature 10-4ap/F +.5 +.5 Coefficient fuel Temperature --

.85 .85 Coefficient Ilultiplier CPC Trip Response Time sec .75 .75

(-) CEA Holding Coil Delay sec .3 .3 v

CEA tine to 90!; sec 3.0 3.0 Insertion (Including Holding Coil Delay) ,

CEA Worth at Trip 10-2ac -5.41 -8.00 (all rods out) i 4-Pump RCS Flow Coastdown Figure 15.1.5-1 Figure 7.1.8-1 i

l O

62-4

O Table 6,2-3, MDNBR Comparisons Between Detailed TORC, CETOP2, CETOP-D and COSMO for the Loss of Flow and CEA Withdrawal Events (Response to Question 492.62)

MDflBR Estimated O Transient Detailed TORC /CE-1 CETOP-D/CE-1 for COSM0/W-3 CETOP2/CEi Loss of Coolant Flow 1.240 ,

i Full Power CEA Withdrawal 1.230 ,

~ - -

1 l

O 4

62-5

O TABLE 62-4 Initial PLHR Data for LOF and CEA Withdrawal Analyses Loss of Flow CEA Withdrawal Cycle 1 Cycle 2 Cycle 1 _ Cycle 2 4-Initial power level 103% 103% 103% 103%

Initial Fxy

• 1.63

• 1.57 Initial Fz

• 1.42

• 1.65 Init1}al Fq 2.35 2.32 2.35 2.59 Core Average LHR (kw/ft) 5.71 5.71 5.71 5.71 Initial PLHR.(kw/ft) 13.4 13.2 13.4 14.8

~

, *Not available for FSAR cases I

d e

O I

62-6

) . Question 492.63 Provide test reports for the b6 sis of. core inlet flow distribution measurement uncertainties which are the basis for the core and hot channel inlet flow reduction adjustment.

Response

The core inlet flow distribution and exit pressure distribution shewn in CEN-139(A)-P were based upon the flow model test report. The inlet flow distribution uncertainties given in Table 5-1 of CEN-139(A)-P were based upon the measured variances of the test data. As requested, the test report is being prepared for submittal.

S

%J s

. (_)

63-1

Question 492.64

,,3 - Provide detailed justification of plant meas

\_.) as flow measurement, plant calorimetric urement uncertainties such

, e tc .

Response

\

Afi0-2 Cycle 2 Plant lieasurement n es Uncertai ti The state variables that e are measur d di system have uncertainties which are taken irectly or calcula adjustment factors, BERR1 and BERR3 nto account in the CPC power uncertainty factor. BERR3 is the local power density from Cycle 1 and is documented in sR fThe meth unchanged contained in Reference 6. e erence 5 with supplementary informatio n The methods used to obtain BERR1 for C relative to Cycle 1 as discussed in resycle 2 are modified in t the system parameter uncertainties ponse to question 492.65.

First, (Reference 7).

Second, a rational extension of thwere mt inc technique used in Cycle 1 (Reference 5) e stochastic simulation state variables and their uncertainties was employed to simulate add The measured state variables wo types: are of t A. Directly fieasured Variables 1.

2 Pressurizer pressure

3. Cold Leg temperature

4. Hot leg temperature 5 Excore CEA Detector / neutron power position 6

Reactor Coolant Pump speed B. Indirectly lieasured Variables 1.

2. Core mass flow

3. Core thermal power

4. Radial peaking factor Axial power distribution Only three of these state variables ha O unlike that of the Cycle 1 analysis ve uncertainties combined in a mann V .

pressure, and core mass flow. They are cold leg temperature

, pressurizer 64-1

i The Monte Carlo simulation analysis used to combine the individual uncertainties s, receives inputs of cold leg temperature, RCS pressure, excore neutron detector J

signals, Control Element Assembly (CEA) positions, core flow and their associated uncertainties. The uncertainties associated with the on-line sensors pertaining to these variables are given in Table 64-1. The basis for these values are instrument specifications, type tests, and calibration procedures.

. For the directly measured state variables, the listed values are the basis for the uncertainties used in the simulation. ,

A. Directly Measured State Variables The following sections describe the sources of uncerttinties for each of

() the directly measured state variables uded by COLSS and CPC on ANO-2. The reference types denoted A, B, C, and D in the following tables are defined below:

REFERENCE TYPES A. Design Specification - Specification to which the equipment manufacturer designs the equipmer.t.

B. Error Analysis - Analysis of uncertainty based on the Manufacturer's specification of the equipment.

C. Design Analysis - Analysis of uncertainty based on the design of the specific equipment.

D. Acceptance Criteria - Specification of accuracy to which the plant operators calibrate the equipment.

I) v 64-2

1. Pressure Measurement Uncertaintt Table 64-2 displays a detailed breakdown of the uncertainty components

)

associated with the reactor coolant system (RCS) pressure measurement. The total error in this measurenent is _.

2. Cold leg Temperature Measurement Uncertainty Table 64-3 displays a detailed breakdown of the uncertainty components associated with the RCS cold leg temperature measurement. The total error in this measurement is .

3. Hot Leg Temperature Measurement Uncertainty Table 64-4 displays a detailed breakdown of the uncertainty components associated with the RCS hot leg temperature measurement. The total error in .this measurement is .

4. Excore Detector Measurement Uncertainty Table 64-5 displays a detailed breakdown of the uncertairity components O

associated with the excore detector neutron flux power measurement. The total error in this measurement is .

5. CEA Position Measurement Uncertainty Table 64-6 displays a detailed breakdown of the uncertainty components associated with the CEA position measurement. The total channel uncertainty is .

6. RCP Shaft Speed MeasuIement Uncertainty Table 64-7 displays a detailed breakdown of the uncertainty components associated with the reactor coolant pump (RCP) shaft speed measurement.

for the 28" disk, and The total error in this measurement is

[ ] for the 17" disk.

b-)

64-3

B. Indirectly Measured State Variables The following sections present the sources of uncertainties associated t

with each of the indirectly measured state variables used by COLSS and CPC in ANO-2.

1. Flow Measurement Uncertainty The total core mass flow uncertainty includes both the uncertainty in determining the best estimate flow rate and the uncertainty in the CPC measurements and algorithm approximations used to infer the actual flow rate.

The uncertainty in the best estimate flow rate includes the uncertainties in factors such as field ano vendor tests and pump characteristic curves. The CPC flow determination uncertainty includes the uncertainties in reactor coolant pump speed, coolant density determination and the flow algorithm.

Core mas,s flow is the only indirectly measured state variable used in the Cycle 2 uncertainty analysis in a manner different from Cycle 1. The RCS

.a V flow uncertainty is of design flow. Tao components are included in this uncertainty. The first component is obtained from a detailed uncertainty assess-ment of the RCS calorimetric flow measurements. The value of this component is

The second component is a calibration uncertainty. The procedurc for per-forming the flow calibration tests is described in Reference 8. The acceptance criteria, described in Section 2.0 of Reference 8, specifies the value of the calibration uncertainty to be .

' ~

The third component is the RCP speed sensor measurement uncertainty. There are two speed sensor disks used in the RCP speed sensor system of ANO-2. The un-certainties associated with each disk are displayed in Table ti4-7. The large value

-of [ ] is equivalent to a flow rate uncertainty of [ ].

I 64-4

, These three components are [ root-sum-squared to produce a flow measurement i

V uncertainty of 13.6%]. This is conservative with respect to the_rCS flow uncertainty used in the overall uncertainty analysis.

The COLSS and CPC flow rate measurements are calibrated against a flow rate determined by the secondary side calorimetric measurement method, using the relationship:

BULK OCCRE = "VESSg9 (h -hBULK) where QCORE is the core power determined from secondary side measurements, primary side heat losses, and primary coolant pump input, BTU /hr, W is the vessel mass flow rate, lb/hr.

VESS

. ,' BULK is the bulk enthalpy of the hot leg flow, BTU /lb, HOT t

h BULK is the bulk enthalpy of the cold leg flow, BTU /lb.

O)

(_ COLD The bulk enthalpics are determined from the relationships, BULK AYO h = f (TH0T' PPRESS) hut h BULK = f (TAVG COLD COLD' pPRESS) where T is the average hot leg temperature, F, 0

is the average cold leg temperature, UF.

OD P

PRESS is the pressurizer pressare, psia.

The uncertainties associated with the parameters used in the equation to determine flow rate are given in Table 64-8. The listed uncertainties include oluipment tolerances, A/D conversion errors, environmental effects, and instrumentation drift.

O 64-5

The average cold and hot leg temperatures are determined by taking the average of the measured RTD temperatures in the cold and hot legs, respectively.

(6) Since the cold leg temperatures are measured just downstream of the primary coolant pumps, the average of the cold leg RTD temperatures gives a good representation of the cold leg bulk coolant temperature. -

g _.

Since an additional uncertainty is included on the hot leg

~

temperature, as indicated in Table 64-8. _

The uncertainties are combined . The distribution of the uncertainty for each variable in Table 64-8'is defined

. The resulting total flow measurement uncertainty has a range of This value supports the allowance of specified for the calibration flow measurement uncertainty.

O 2- core Tnerm8' rower "ees" rem "t uncertainty The treatment of power measurement uncertainty is not changed from the Cycle 1 analysis. This uncertainty is not statistically combined in Cycle 2. It in-cludes a calorimetric component and a CPC calculational error component. A value of 2% for the calorimetric error component is applied at 100% power. I The actual calorimetric power level uncertainty at 100% power is less than 2%, but CPc must use not less than 2% to assure consistency with 10CFRCD, Appendix K.

The CPC calculation error component is an output of the uncertainty analysis and is not presently available. _ _

The total power measurement uncertainty in Cycle 1 was and resulted from a consideration of uncertainties in both core thermal power (AT) and neutron flux power. The Cycle 2 value is not expected to be substantially different.  ;

64-6 l

\

l

\

3. Radial Peaking Factor Measurement Uncertainty

The treatment of radial peaking factor uncertainties is not_ changed from the (3

(_,/ Cycle 1 analysis. A detailed description of the treatment of these uncertainties is given in References 5 and 9. The uncertainty in CPC radial peaking factor

!. election due to uncertainties in CEA position indic '. ions is directly simulated in the overall analysis.

4. Axial Power Distribution Measurement Uncertainty The treatment of the axial power distribution uncertainty remains unchanged from the Cycle 1 analysis. A detailed description of the treatment of tnis uncertainty is given in Reference 5.

L) n m

64-7

Tabb4-1 - (~)

'v' ANO-2 PROCESS INSTRUME'NIATION UllCERTAINTY COMPONENTS TOTAL IriSTRUMENT CALIBRATION A/D ERROR TOTAL SIGNAL REFERENCE ACCURACY ENVIRONMENTAL EFFECT DRIFT i.

ASTERISKS INDICATE THAT THE EFFECT IS NOT APPLICABLE ALL VALUES ARE THE TOTAL RANGE OF VARIATION REFERENCE ACCURACY = EQUIPMENT TOLERANCES, NORMAL ENVIRONMENTAL EFFECTS, FIELD CALIBRATION g ENVIRONMENTAL EFFECTS = WORST CASE ENVIRONMENTAL EFFECTS, SEISMIC EFFECTS E DRIFT = EQUIPMENT DRIFT BETWEEN PERIODIC PLANT TESTS '

TOTAL INSTRUMENT = COMBINED EFFECTS OF REF.ACC., ENVIRONMENTAL EFFECTS, AND DRIFT A/D ERROR = ANALOG TO DIGITAL SIGNAL CONVERSION ERROR

~

Table 64-2 PRESSURIZER PRESSURE UllCERTAINTIES A. Reference Accuracy REF. B REF. A I _

B. Environmental Effect

_ q REF. B REF. B

~

C. Drif t (18 Months): REF..B D. _ Total Instrument Uncertainty: _

E. ,A/D Error -

REF. B O ReF. B REF. B REF. C F. Total Error After A/D Conversion:

E _

l 0 -

64-9

s Table 64-3 COLD LEG TEMPERATURE UNCERTAlflTY A. Reference Accuracy: REF. B O _ ._

- B. - Environmental Effect _

REF. B REF. B C. Drift (18 Months): REF. B D. Total Instrument Uncertainty: ._

E. A/D Error REF. B REF. B

, REF. B REF. C

O F. _ Total Error After A/D Ccaversion

6 _

J l

O 64-10

,. . - - . _ - _ . - - - - .. . = - . . . - - _ = . . . . . _ _ - . .- _. - - - - _ . _ _ _ - _ _ _

Table 64-4 Il0T LEG TEMPERATURE UNCERTAINTY A. Reference Accuracy REF. B REF. B B. Environmental Effect -

l

~

. REF. B REF. B REF. B C. Drift (18 lionths): _

D. Total Instrument Uncertainty 4

1 l

i s

h _

REF. B i .

( E.

A/D Error: ._

Total Error Af ter A/D Conversion:

t F. b - -

4 O

1 64-11

. , _ . . - . . , _ - . _ _ _ _ _ - . , _ _ ...._.__.,__....,_.-_.m., . _ , _ . _ . _ _ _ . , , - ,

7 Table 64-5 EXCORE DETECTOR Uf!CERTAlflTY i

l A. A/D Error: _

REF. B s REF. B REF. B REF. C B. Reference Accuracy:

REF. D l'

i REF. C i

. REF. B 1

0 C.

Long Term Drift:

REF. B g

D. Total Instrument Uncertainty: ~

b _.

E. _ Total Channel Uncertainty:

L i

i ._

O -

i 64-12

r.

L TABLE 64-6 CEA Position Uncertainties

% FULL COMP 0tiEtt r SCALE ItiCHES REFEREtiCE A. _

Reed Switch Position Transmitter _

A B

B B

B D

B. A/D Processing ,

~ B B

B C

, h C D

B Total Channel Uncertainty i

• = Applies to CEAC path only .

1 O

64-13

h;.

t

Table 64-7 RCP ;l! AFT SPEED titlCERTAlflTIES O

j A. Probe - Disk Interface (28" Disk) 17" Disk _

! REF. B i REF. B I REF. B

! REF. B 4

4 t

l _

) B. Sensor to Register: REF. B

! ~l ,

i

i j * . - i C. _ Total Speed Error:  ;

~

lO

~

riOTE:

1 4

f i

t t

I i

O 9

64-14 ,

l

r-Table 64-8 UtlCERTAlflTY COMP 0flEllTS FOR VESSEL FLOW L'l.)

MEASUREMEllT - SEC0!!CARY SIDE CALORIMETRIC TECliflIQUE EQUIVALErlT UllCERTAltlTY IMGflITUDE OF Ill CORE FLOU RATE, PARAMETER UtiCERTAlflTY_ _,  % OF DESIGil FLOW

- . i fl0TES:

(1) Listed uncertainty is for the average of four RTD signals; the uncertainty on an individual RTD signal is assumed to be [ ]. This is conservative with respect to the actual uncertainty shown in Table 64-3.

(2) Listed uncertainty is for the average of eight RTD signals; the uncertainty on an individual RTD signal is assumed to be [ ]. This is conservative with respect to the actual uncertainty shown in Table 64-4.

l l

[a g

\

64-15

, i

, Question 492.65 v

Provide details on operating state parameter uncertainty justification and the effect of the additional SCU on the ANO-2 Cycle 2 thennal margin compared to Cycle 1.

Response

The response to question 492.16 provided a comparison of Cycle 1 and Cycle 2 methods of combining state parameter uncertainties with CPC DNB and power distribtuion algorithm uncertainties in calculating the addressable DNBR uncertainty factor BERRl. A brief description of the changes in the method of analyzing state variable uncertainties from ANO-2 Cycle 1 to Cycle 2 is giveninthesubsecuentpages. CENPD-17n provides a description of the CPC uncertainty assessment.

O ine enswer to rSAR question 222.i29 documented 4n CEN-35(A)-e discussed deterministic and statistical (root-sum-square, RSS) methods of combining these uncertainties. Calculations using the stochastic simulation technique for Cycle 2 and the statistical (RSS) approach described in CEN-35(A)-P would yield approximately equal values for BERRI. However, the actual Cycle 1 approach was a combination of stochastic simulation, RSS and deter-ministic approaches. A numerical example of such a calculaticn was provided to L. Beltracchi of NRC following the uncertainty analysis audit of June 14, 1977. The Cycle 2 method is estimated to represent [ .Ioverpower margin gain relative to the actual Cycle 1 method.

65-1

STATE VARIABLE UNCERTAINTY AtlALYSIS ftETHOD CilANGES AN0-2 CYCLE 2 Introduction This dicussion compares methods used to evaluate CPC

" power uncertainty f actors" for static DNBR and tw/f t for AN0-2, Cycle 1 and ANO-2, Cycle 2. The two power uncertainty factors are defined as:

1. BERR1 - Power uncertainty factor for static DNBR calculation.

2. BERR3 - Power uncertainty factor used in LHR(kw/ft) calculation.

The methods used to calculate BERR3 are essentially unchanged from Cycle 1.

Any detail changes will have no significant effect on thermal margin.

The methods used to obtain BERR1 for Cycle 2 are modified in two respects relative to Cycle 1. The first is the incorporation of system pararoe.ter uncertainties directly into the DfER trip setpoint. The second is the direct stochastic simulation of the state variables and their uncertainties.

BERR1 Calculational Methodology - Cy_cle 1 The simulation and other statistical methods (Root-Sum-Square) used to com-bine DiiBR related uncertainties are described in CENPD-170. The uncertainty

' (3 components include:

%.)

1. pseudo - hot pin synthesis error

2. DNBR algorithm error

3. Radial peaking factor error Section 5.60; CENPD 170 Sup. 1-p A spectrum of possible combinational methods for the above errors, state variable (pressure, temperature, flow) " measurement" uncertainties, and other less significant uncertainty components, is presented in CEN-35(A)-P.

Excerpts from CEN-35(A)-P are included on pages 65-4 through 65-12. The methods range from a " statistical" combination of all factors to a multiplicative " worst-case" combination of the uncertainties. The statistical method is illustrated in the equation for EyN3 on page 65-6. The deterministi

" worst-case" method is illustrated in the equation for Tf.g on page 65-7.

The method actually used for Cycle 1 was essentially the " worst-case" method.

The documentation of the detailed methodology was reviewed by the NRC prior to Cycle 1 criticality.

o 65-2 r

BERR1 Calculational riethodology - Cycle 2 O The method used to eveiuete BERa1 for nr:0-2, Cale 2 is similar to Cycle i in all but two respects:

1. The system parameter effects are incorporated directly into the DilBR trip setpoint. This methodology is described in detail in Cell-124(B)-P,Part2.

2. The state variable effects (P, T, M) and those of their measurement uncertainties are directly simulated in Cycle 2.

The CPC system allows operation over a wide range of pressure, temperature and flow. In the Cycle 1 analysis, all Drib calculations were perforned at a nominal set of state variables (To, P o, Mo). The Cycle 2 analysis performs the equivalent set of calculations over the entire range of permitted state variable values. The values for an individual computation are obtained by randomly sanpling from the pressure, temperature, and mass flow PDF's.

The PDF's for the 'above state variables are assumed uniform over the allowable range. .

The effects of " measurement uncertainties" in the state variables are also directly simulated by adding randomly sampled " noise" components to the state variable set described above. The set of state variables with " noise" is processed through the CPC algorithm, the unperturbed set is processed through the FLAIR /(DliBR) alo ~ithm. The comparison between the two calcula-p tion! provides one datum point for the Di1BR uncertainty distribution. The v pro ess is repeated 1200 times yielding a Lt1BR error distribution which ircludes the above uncertr.inty effects.

This methodology for BERR1 calculation is a logical extension of the BERR3 calculational methodology documented in CE!:PD-170. It has the advantage of greater physical consistency than the RSS method described as " purely statistical" in CEft-35. The stochastic simulation method directly addresses the variable sensitivity coefficient difficulty intrinsic to the RSS method.

65-3

3~

DiscussionfromCEN-35(A)-P

'ihe CPC's can be thought of as consisting of two classes of calculations;

" static" and " dynamic". 1hese two calculation types operate in conjtaction to perform a trip decision as shown in Figure 65-1. The static c11culations are

' designed to give high accura'cy during steady-state operation and nornal operational conditions-and use " snap-shots" of core conditions to ascertain the margins to the DNBR and fuel centerline melt specified acceptable fuel design limits. Dynanic calculations are designed to give conservative indica-tions of the margin to the fuel design linits and are rclied upon to rapidly

" update" the static calculations, and anticipate the core response to pro-vide protection for the anticipated operational occurrences that are part of the CPC design basis events.

. Uncertaintics and allowances are applied to the results of these calculations

• to account for the following items: ,

1) Ficasurement Unt.crtaintics, .

i 2) Algorithm tiodellir.g Uncertaintics (Calculat onal Unt ertainty),

3) Algorithm Const ants Uncert ainties (Calculational U .ccrtainty),

4) CPC (Computer) Processing Uncertainty (Calculational Uncertainty),

5) Scotic Allowances,.and-

6) Dynamic Allowances.

o -

-Items 1 through 5 apply to the CPC " static" calculations while item 6 applies to the " dynamic" calculations. The dynamic calculations employed by the CPC (dynamic compensation, projection techniques, etc.) are designed to accommodate

-cffects such as sensor and measurement channel dynamic response, CPC delay time, RPS delay time, CEA insertion time and NSSS dynamic response. A dis-cussion of the methods employed for dynanic compensation has been provided in response to questions 222.66 and 222.67 Further information on dynamic

allowances will be provided in response to question 222.7S. The following 4

discussion will concern itself with the uncertaintics and allowances that -

. apply to the " static" calculations.

'lhe CPC's perform five basic " static" calculations:

1) Core Average Power -

2) Core !! ass Flow Rate

3) llot Pin and flot ' Channel Power Distribution

4) Steady-State kw/ft Trip Sctpoint

, 5) Static DNER i.

o P00R ORGK  ;

. 65-4 I

1 j

l

/~ F.adi calculation is affected by either measurement or calculational uncer-

~

taintics, or allowances, or all of the above- as shown pictorially in Figure '65-2 Heasurement uncertaintics are the effect on the CPC response due to sensor and'mcasurement channel characteristics. This includes consideration of calibration techniques , drif t, linearity,' environmental effects, the ability of 4

the sensor to measure the variabic of interest and the accuracy of analog to

' digital conversion techniques. Information on these uncertainty components are obtained from calculations, instrument specification. . type testing and previous operating experience. The uncertainty conponents are charactcriaed (i.e. , randon, systematic, dependent or independent) and quantified using

.the above information. The contribution to the CPC DNBR and kw/f t uncertainty can then be determined for the components by observing the diange in CPC result with a measurement uncertainty applied to that assuming a perfect measurement. Measurement uncertainties will be discussed further in the response to' question 222.126. .

' Algorithm modelling uncertaintics address the accuracy with which CPC algorithms replicate the results of "best estimate" measurements and/or "best eatimate" calculations. An exampic of how this is performed is given in Section 6 of CENPD-170-P where the synthesi:cd three-dimensional peaking factor is compared ~

to, in effcet , measured results. As with measurement uncertainties, modelling uncertaintics are also characteri:cd for later inclusion in determining the i total CPC. uncertainty. These uncertainties are then further augmented to l account for the accuracy of the CPC constants used in the algorithms.

Algorithm constants uncertaintics address the accuracy of the measurements  !

and/or calculations used to obtain the constants used in the above assess-pent. An exampic of this technique is again shown in Section 6 of CENPD-170,P where the CPC modelling uncertainty is increased to account for the uncertainty

.in the planar radial peaking factors determined by measurement (in this case INCA). These types of uncertaintics depend upon both the accuracy of the instruments used in the rucasurements as well ,s the technique used to prccess the measurements and can consist of both syM, .:atic and random components. -

I

' The 1.'st uncertainty component to be considered is that attributable *o CPC (computer) processing. This component addresses the effectTesting that staling,of round-off and bit manipulation has on the CPC computed result.

L the CPC's and calculations provide the information needed. Comparison of actual CPC response to a known result obtained using the CPC algorithm with a higher resolution computing facility provides a nechanism for quantifying '

and characterizing the processing uncertainty.

f

.O P00R3g3lyp[

65-5 ,

l

_ _ _-- ~ _ ..

L 1

LO We

%c above four-uncertainty components are consideredDNto be independent.

I Tl 'PU, are then func .

total DNUR and local power density uncertaintics ETEc.B. CPC's are designed wi.h tions of the conbination of these uncertainties.

the capability of accommodating uncertaintics in a variety of ways with the (systematic choice .being dependent upon the nature of the uncertainty component Measurements can be biased prior to use in calculations to account or random).

.for t~icertaintics and/or allowances, calculated results can bc_ individually modified, or selected calculated values can be modified to account forItthe is effect of all of the above components on the final trip comparison.

4

. presently intended to convert the above components to common units Wis of percent of rated power and equivalent relative power margin (overpower margin).

infers that parametric analyses be performed to convert these components to the common units. An example of how this conversion will be achieved is l' -

discussed in response to question 222.128.

Figurc 65-3 provides a brief synoosis ne method chosen to combine

. -of the presently intended evaluation process,uncertaintics d with deter- depends upon the be verified as part of the quality assurance program associate -

minin;; the CPC uncertaintics.

' By way of example, Figures 65-4 and 65-5 li,st typical magnitudes of DNBR an related uncertainties, with their overpower margin equivalents, Taking each encompassing component - ,

i (mi=0) (except wht.. c noted)

L O .the components as independent, and tw -siddiscussed d with

cro meanin theleast errors abovethe samefour categories proaability and assuming that all uncertaintics are given to atand confidente level; by combining the systematic and random components as shown below:

. , 1/2

• • ' DNB B .m 3~

s

[8 lio; 2}

• I'08

i - '*

E =1+ I- 100 . (100) 1=1 1=1 T

~ '

i/5 .

i/5 ) . .

and 1/2

)

. ( *

. : 4 r,

"i + I .(

ko;) 2

• I*II

. .E.LPD =1+ I g

)

7 ,

1=2 1=2 .100 .

x.1 = "i + Ko.1 where .

and Y; J "i + ka j

(. .

O .

65-6

As an additional exampic an .cstimate of the CPC uncertainty which conserva-tively assumes a ' worst case' stack-up of the components can be obtained U(~'y a multiplication of the individual components as; Xi 8

DNB =

• I*I3 E R (1 + TOO) 7 i=1 1/5 and

• LPD E.T

" II (I

• Yi 100)

= 1.16 ,

i=2 ,

Notice that the core average power uncertainty is not included in the abcve l

~

c:mputations s'nce it is currently taken to be one-sided and its magnitude I

- is given in terms of rated power rather than relative power units (overpower cargin). Independent of the method used the resultant power Icvel input to the DSBR and kw/ft calculation is given by: .

A POWER = E (POWER +5) ud Ab

-POWER

= E (POWER +5)

{v') .

where POWER CALC is the CPC calibrated power (% of rated) is &c above va w aQust for m ccrta W es ,

POWER ADJ 65-4 and 65-5 results Using the above formulations with the values given in Figures in a DNB uncertainty in the range of 14*. to 19*. and a local power density un-

. certainty in the range of 17*. to 22'. at rated power. -

llaving add essed the techniquer available to acconnodate uncertainties, the following discussion addresses the application and need for " static" allowances.

"Ihese allouances are provided for two reasons: 1) to account for the effect on the mar;in to the fuel design limits of variations in parameters not monitored 'ay the CPC's , and 2) to account for the effect on the margin to the fuel desiga limits of allowed variations (action The onlythresholds parameteror that deadbands) falls into in the

- the parameters monitored by the CPC's. "Ihc mechanism for accom-first category is the azimuthal power tilt magnitude.

modating such an allowance, if necessary, is discussed in response toDeadb question 222.102. discusses the effect of small deviations on margins

~

response to question 222.109 fuel design limit and which can be seen to be insir.nifi--

to the centerline melt liowever, if either of the ab .vc allowances (9 cant for up to 12 inclics deviation,are necessary, appropriate factors will be use '

'J adjusted for CPC uncertaintics (10N1:R A DJ) discussed previously. ~1hc adjustment for these effects is of the form:

p0WEll = POWElt A OWAE FAGORS.

DJ ADJ 65-7 P00R ORIGINAL ~

n

J --

.Q .

FIGURE 65-1 DASICCPCCALCdLATI0i!ALSCllEME STATIC CALCULAT10f!S

~

, CA ATI0i S >

v ~

TRIP DECis10il O

1 .

. STATIC CALCULATI0ilS:

HIL.I ACCURACY

. .' . l

.' . STEADY STATE ,

~

. NORMAL OPERATION DYilAMIC CALCULAT10iiS:

CONSERVATIVE CALCULATI0ri

'~'

..- .'.,. .' ANTICIPATED OPERATIO!:AL OCCURRErlCES -

?

G o'

i POR DE!K

~

, 65-8 ...

. - , _ m. _ _

,4 FIGURE. 65-2

~ ~

OL nAsic cPc EnTic cAtcutnTi0ris 4

1. POWER .

2,' fiASS FLOW RATE

3. Il0T Pill /cIlAli!!EL P0llER DISTRIBUTlDil ,

. 4. STEADY-STATE Ki!/FT TRIP SETP0 lilt 5, STATlc 'DiiBR -

> 2' -III'CII) >

(1) - .

p _LI L (i n . 0 1 n >

.h (1),(ll)

O '

(1),(11 )

~

(]).. > l- ,

(I) Nil),

' fill) -

(1), (11)#,(111)

'(1) > 3. > 4. '

(1),(11),(111)

(1) flEASUREMEt1T Ut4CERTAltlTY

' (11) CALCULATIO!!AL UllCERTAltlTY (Ill) ALLO',1AriCEs (ST ATIC)

P00R GRGINAL t

~

' FIGURE 65-3 UNCERTAlt;TY EVALUAT101 PROCESS

1. CilARACTER12E AND QUANTIFY UNCERTAINTY COMPONENTS Instrument Specifications

, Type Tests Comparison of Design Codes Comparison of CPC Calcs, to "Best Estimate" Cales, ,

Previous Operating Experience .

' Calibration Procedures .

2 DETERllINE EFFECT bF COMPONENTS ON CPC CALCULATIONS Parametric Analysis Comparison to Design Codes .

l 3. DEFINE CPC UNCERTAINTY FACTORS FOR DNBR Calculations

• D1/FT Trip Setpoint Calculation

). APPLY FACTCRS TO BIAS CPC DNDR Calculation  : Adjust Power Level Input to and . Calculation to Account for

. Mf/FT Calculation CPC Uncertainty

.k

w \ ., - -

'\. -

I, .\ *

% ,i i , ,. . .,

x . .

z 4 \. ,

4 LO .u.

65-10 P00R OREL -

i e ,

, -- . ~ , _ _ , . _ _. ,, , ... , ,, .mm.

m-FIGURE 68i-4 CPC DNia RELATED UNCUtTAINTIES Approximate (v~j Equivalent Overpower Margin Units, '.

X Uncertainty Component i i _

gnitude 1

Inlet Temperature 2'F.

1) 2 Reactor Coolant System (I) 30 of best *

2) est, flow Flow Rate 1 I) l'6 of CPC

3) CPC Flow Determination Calc. Flow Uncertainty 1

4) Reactor Coolant System 22 psia .

Pressurc -

(4) 5', of rated Core Nverage Power ( )

5) 1 3's of axial

6) Ex-Core Detector-Axial (5) peak Fitting 5

'lhermal Margin Calculation (6) 5*. ,

7) 1

8) Total CPC DNBR Processing 0.05 DNBR Units

• q V

uncertainty in d'etermining best estimate flow rate exclusive of CPC technique (1)

(field and vendor tests, pump characteristic curves , etc.)

uncertainty due to CPC measurements and algorithm approximations used to in (2) flow rate EX-CORE), alg rithms, algocithm

, * (3) includes the measurement (e . g . , Ti!0T, TC01.D '

constants and calibration uncertaintics associated with the calc -

AT and neutron flux power.

<5', at 100'. power and is currently treated as a one-sided uncertainty.

(4) includes ex-core detector sub-channel measurement error, calibration uncer-(5) tainty and uncertainty in determining rod shadowing, shape annealing and bounda2y point power constants.

\ (6) includes pseudo-hot-pin synthesis error, DNBR algorithm

-2.4*. and ko 7 =

error 7.2', and uncertai in radial ;.catin;, factors and results f rom an1-P). m7 = 1

\ (sce Sectian S6.0 of CENPD-170, Supplement l 1

s

. 65-11

FIGURE 65-5 CPC kw/ft RE!ATED U:: CERTAINTIES Approximate Equivalent Overpower Margin Units,'.

y Jacertainty Component i ifagni t ude i (2)

Core Average Power ( ) S S. of rated -

))

h) Axial Ex-Core Detec 3'. on axial 3 Fitting {ag - peak ,

Fq Synthesis Calculation (4) 10'. of Fq 10 3) 7.0 of Fq 2

4) Total CPC Fq Processing (1) includes the measurements (c.?,. , T gg ,TCOLD' FE 0D0), alc rithms , algorithm constants, and calibration uncertaui[ics associafed With the calculation of AT cad neutron flux power.

'(2) <50 at 1000 power and is a one-sided uncertainty (3) includes ex-core detector sub-channel ceasurement error; calibration uncertainty d uncertainty in determining rod shadowing, shape annealing and boundary point lower cons tan ts.

(4) includes pseudo-hot-pin synthesis error and uncertainty in radial peaking factors and results from an m 3 = -30 and 3W = W. (see Section 6.0 of Nb170W e

0 e #

9 e O 9 9

0 V e

O 4

• O 9

8

.'\ \

7m i 4

, 05-12

• l

W U Question _492.66 Provide a sensitivity study of the oveipower penalty factor, BERR1, to the location of DNB relative to the axial nower peak location.

Response

A sensitivity study of the conversion factor of DNBR to OPM was requested. This study covers the LC0 operating ranges and emphasizes the use of many ASI's. 2388 sets of conditions were examined including 597 ASI's from beginning to end of cycle 2. .The cases were iterated to a 1.24 limit and to a 1.36 limit. The derivatives f.hggbare plotted versus ASI in Figure 66-1.The scatter of the derivatives at positive ASI's is due to the minimum DilBR occuring at different axial locations. For positive ASI's MDNBR may occur high in the core or relatively low (around mid core level). If the MDNBR occurs high in the core for positi'v e ASI's the local quality will be relatively high. The high quality .

results in a strong sensitivity of DiiBR to power and correspondingly a low

'O derivative of POL ta DNBR. For cases in which the MDNBR occurs low in the core the coolr.nt at the node of MDNBR is either subcooled or low quality. The DNBR then has a lower sensitivity to power and thus the derivative has a D

larger nagnitude.

Nothing unexpected is observed in this figure.

POL = Power at the Operating Limit.

O V

66-1

(?

~

Ex. I

~

s ()

o r P-OP=

e a J sr

• i ,

se u's = , s.*

ym; r

' Cy p to *e CA O

(*s %.%is* C}

O( h P. ) o

'arb*sm + e-90 tJ Os'JC4aJC4 LJ C is . i.

g

• ' Oc0 O c

O E5 o o O

. --4 V) .,

E g e.

O l 's x

w V)

~

6, o

H- c$

~

.-4 .

O

~

~

>~.---

. c h 1

, 6-4

%Q CO o

O 4- d rJ i.ij 'd '

U3 Ll.)

> J

>~4 C .4 O

! I

}- - - .y_._._______g----__ ,y g -

g g o o O o o c3 O,

O rJ v (O g . . . (D. . -

i . i -

1

$----4 i

l (U

%)'

's - l i

L  ! ..cty:

!)t!C J r,'L,b 3 J G1 '!C J .1!. 'O b 3 0 m 1/.!2U/.:b7C t i 1

Question 492.67 c In response to question 492.29, provide a quantitative assessment of specific U~ results for DNBR and power margins gained on Afi0-2 Cycle 2 as a result of proposed methodolo'gy change (e.g., use of SCU, CETOP/CE-1 vs. COSM0/W3, etc).

Also provide the overall value of DNBR limit and a corresponding BERP.1 to invalidate the margin gained by SCU in the final CPC system. In addition, please provide a proposed technical specification which limits the value of BERR1 to prevent entries inconsistent with approved DNBR limits.

_ Response Figure 67-1 shows the estimated DNBR and overpower margin changes for ANO-2 Cycle 2 as a result of methodology changes.

The overpower Margin effect of the change to TORC /CE-1 was estimated based on past COSMO studies. The overpower margin effect of the statistical combi-nation of system parameter uncertainties was estimated by considering the -

effects of the various uncertainties applied in each manner. This was explained in Figures 67-2through67-4which were presented to NRC ( L. Phillips) and PNL at the April 14, 1981 meeting in Richland, Washington.

In interpreting this data, it is important to understand' that the overpower margin gains quoted are estimates biased in the conservative direction. The DNBR margin gains are included for illustrative purposes using a typical con-version factor. The actual margin. gains depend on the specific operating conditions.

O The DNBR limit built into the CPC software is 1.24. This cannat be changed for Cycle 2 without a software modification and a repeat of the Cycle 2 safety analysis.

C-E and AP&L do not believe that a penalty should be applied to invalidate the margin gained by SCU. Itowever, any penalty required by NRC as an interim measure can be applied as a multiplier on the CPC uncertainty factor, BERRl.

The value of BERR1 will be provided as discussed in response to Question 492.16.

The discussion concerning a Tech Spec to limit the value of BERR1 will be

. addressed separately by AP&L.

C' 67 l

~~

O O FIGURE"67-1 (y,':

a Estimated Margin Gains for Methodology Changes for ANO-2 Cycle 2 (Response to Question .492.67)

OLD METHOD NEW METHOD ADNBR 4 OPM (")

' ~

1. COSM0/W-3 TORC /CE-1

2. Detailed TORC /S-TORC Detailed TORC /CETOP-D

3. COSM0/C?CTH CETOP-D/CETOP-2 ,

Statistical Combination

4. Deterministic treatment of system parameter of system parameter Uncertainties uncertainties e 5. Combination of Expanded statistical y deterministic and treatment of state N statistical treatment parameter uncertainties . .

of state parameter uncertainties NET fMRGIf1 GAIN SCU (Items 4 & 5) ..

No additional gain beyond Item 1 (C0ST10 vs. TORC)

O O O. -

FIGURE 67-2 METHODS OF ACC0m0 DATING SYSTEM PARAMETER UNCERTAINTIES IN THERMAL-HYDRAULIC MODELS THREE EQUIVALENT METHODS METHOD A - INCLUDE SYSTEM PARAMETER UNCERTAINTIES IN THERML HYDRAULIC MODEL -

METHOD B - APPLY SYSTEM PARAMETER UNCEP.TAINTIES T0 tiDNBR LII;iT .,j METHOD C - INCLUDE SYSTEM PARAMETER UNCERTAINTIES WITH OVERALL UNCERTAINTY ON POWER (s.o. BERR1)

O a ,

FOR CONVENIENCE, MOST SYSTEM PARAMETER UNCERTAINTIES HAVE BEEN APPLIED BY METHOD A,  ;

BEFORE SCU, BUT METHOD B OR C COULD HAVE BEEN USED EQUIVALENTLY.

)

SYSTEM PARAMETER UNCERTAINTIES ARE APPLIED BY METHOD B IN CEN-139 BUT METHOD C COULD ALSO BE USED.

j

-em

-~

_,.y FIGURE 67-3 a

CONSTANTS IMPACTED BY CHOICE OF SYSTEM PARAMETER UNCERTAINTY TREATMENT (Without SCU of CEN-139)

CETOP-D CETOP-2 Factor Ccnstant Method A Method B or C Constant Method A Methed 8 or Inlet Flow Factors _ _

(Cocbir.e d) Applied to FSPLIT 1.00 Applied to FSPLIT 1.00 Enthalpy Rise Factor Applied to D(I) 1.00 Applied to DH 1.00 Pitch & Clad 0.D. Applied to 0(I) 1. 00 Applied to DH 1.00 i Heat Flux Factor S KECDX 1. 00 SKECDK _ _

l.00 Rod Bow DNBR Penalty CE-1 CHF Correlation L _.

E3 Unce rta inty E -

CPC Data Base Constant Method A Method B Method C MONSR Limit ** 1.19 [ ] 1.19 PaltiplieronBERRI[ } 1.0 } ]"

' Red bow penalty (2% CNBR) is applied either to HDNBR limit, or BERRI independent of treat ent cf system parameter uncertainties.

• • CE-l ChF correlation uncertainty and NRC penalty per CENPD-li?-P-A always applied to MONBR limit.

O O O FIGURE 6714 .

CONSTANTS IMPACTED BY CHOICE OF ' SYSTEM PARAliETER UNCERTAINTY-TREATMENT i

OilTH SCU OF CEN-139) i CETOP-D CETOP-2 SAME AS METHOD B OR C SAME AS METHOD B OR C WITHOUT SCU

^

WITHOUT SCU m (SEE FIGURE 2) i 5 (SEE FIGURE 2)

CPC DATA BASE '

CONSTANT METHOD A METHOD B METHOD C MDNBR. LIMIT " N/A 1. 2I1* 1.19 MULTIPLIER ON BERR1 N/A 1.0 [ ]

R0D BOW PENALTY (2% DNBR) IS APPLIED EITHER TO MDNBR LIMIT OR BERR1 INDFPENDENT OF TREATMENT OF SYSTEM PARAMETER UNCERTAINTIES, i

^ *

" CE-1 CHF CORRELATION UNCERTAINTY AND NRC PENALTY PER CENPD-162-P-A ALWAYS APPLIED TO MDNBR LIMIT. ,

1 b i

l gttestion 492.68_ -

(

~

') . In resp'anse to question 492.7, provide additional data on CETOP-D/ TORC U comparisons of DNBR calculations. flore data is required especially for ANO-2 in the region of the DN3R design linit. The comparison should separate Cc1 vert Cliffs-1 and San Onof're 2 data.

Response

Figures 68-lcPand -3show f4DNBR comparisons between detailed TORC and CETOP-D for ANO-2 Cycle 2 and other plants.

Figure 68-1 contains 64 cases for ANO,2 Cycle 2 over the operating range includiig an additional 28 cases over the LC0 range beyond that provided in re,ponse to question 492.7.

In ai cases throughout the range of operating conditions, CETOP-D calculates a DNBR lower than detailed TORC. This indicates that CETOP-D is conservative relative to detailed TORC throughout the range of operating conditions and therefore can be used with confidence in the safety analysis and to derive and verify the CFC on-line thermal margin algorithm CETOP2.

O

( >6 68-1

! Figure 68-lComparison in. MDriBR Between CETOP-D (Tuned) and Detailed l c rTORC-for Afi0 2 Cyc1tthy-.3 dr 2:--  : r-- - ,

-

= y : - -p:q ---r-- = - '* .= ; - ,-:-hir" j = :.5 -

g-_~  ; :

.,. . .. l -  ;. , ; - __ g i Jx?2 c ! J . d - ' "g22-..

?!.':-i~::- __' :1

! -.ir . - 1 . r :- ~; . . : .1 .1 "L

, "= .=: a. .: a u=  :===:=. . = t .=-.s  := : = = = = :. = r =.-;- -- -- -- - -- - - '

( + .- i . ri" . : ...; .i .JJ_ : T_J.12is_b.s#=l=: M ~ H v y_ -

g j: = 7 <

j .

. .. ..- =,,...=.=,--z. - r =-

._...j.

- r- ..,; = - = =.'==.12-~

'..).~~' -

i i

l

.=~ .i Range of Operating Conqitions [=- ,. hR i .g'i. ~.=g ~

_ ~ .- - - - . '

1::. .. . . O tw-

!. ( Zi _2.._L] Inlet Temp.( f) 465 - 605 M N ' * '. "-' ' __-

i . i f.jr.1!2 Sysdem Pressure (psia) 1750 - 2400 Ih~"=*l:Ti B En":ji=J !-

+5- , Vessel flow (ppm) 60% - 120; of h!4-K=l=.=C=i M 6 C~= J - '

-T "==2=-~

. =., =-

322000 9pm x-~ :: = . ; : = - = :=. . . .

. _ _ . . z :..,. ;;, Axial Shape Index 60 . 60 _;,m...-.-e_--.--.

x a- =. n _ -. -

r rO =.j: & : .=L_=. .=; =.- ==

.1.:::. =u

=cl e_= -. .---j =.=m .

- - - z :_ .

i

\ ~- -

.=::-1, : :.::3.: = .

219=R:,r:Zi-;biE.:=i+D ':h ~-

g . . .j ; . - i . . . . . . c1; a_ 4_ . . _. g. _ . y 7. q_. y._gr j g .:; .g

=1g4 gzi5.. g 53 .--- .

j" .5 y i LD 9 1

• ..=.

I --

i =

=

! =E

i

j

, =

M ..

I

= l E9

.c5 i

1

, .;- =

1 l i o r i

(

6 '

1 _: l

( O (,w.

9 ,

- l U ._ {

J, 3 l u 1:

1 l

r, m

e1. l i e E: l

<  := '

t e- - - - -

j w: H  !

bt -

1-u ps

=E l

l

,' [ud =-

~T v",

w 5

p. r: --

o.

F E. '

2,: =.

! 6 .

et 1 W =

fb

..=

==

5.. i,

, = a h =E

> =

==

I

-;.i (g j 1 k i 21n. I mig _ y. .t u -: , -t. .. .

- : .. . . r- - , .. 3 t .t.:t u.: . ..: .-

. .i - ;l :q . + -j i -- - u f

~

'; h- i J 8;-

. -j-..

l

..~  : -

.:7{ . .j '-. 2:r : + j _. :c,' j _g.

-_l -='

-l _ , j! _a : j ~. = . . . '

j

.r 7 . _'

)

.j . jj - . _,

q; ---

. d - .l y 7

- t. -[ .

q . . . : .2 _; ; ; . -

j

! Detailed TORC MD:iBR l

n ___ _ _

68-P _ _ _ _ _

l - . . . -[..

, = , . . . . fo r~ Co

.u _l .ve r '. C11 f fs E.l]a nd G. - . ]. - :, .3:g,11c;= ;;=== rr.:==_= =z

~ *

= .- r-

^

1

,_-.m _ :_ i. . . : . =. . . . . :

{.

_ m .-

.. _!. .;- +. Q q =.,_=_'

. . . . ._ 1 -.- '* .. -i . '.j ..-E ' '- -

- r. .. ^ 3 :* r. ~.. g . r. . j;, 7 ; _,_. . ._ .

= -.g _; ..= . . p _ ull_ .q.t:.paiM.,}z++;gi4,.::12f acz y ::q(: ge=_H.i=r; g'1e az g _i2. '

, (, .  : a _; _n 2 2. -

_

i _ ;:_a.l 22n_ ...t ryyj;; . _.] 17 g ny.g ; _ -p ~ :1 m: --

nu -

l . = _:. a = = - - . - . , _ _ . .  !

!g

[7 ' 7-(, (~.~

Range of Operating Conditions IUNw.5M[

=g c , - n;g . , -. _ .

M ..

I

(

InletTemp.(ef)

• 465 - 580

_..s.

ir == + T V --: =

p System Pressure (psia) 1750 - 2400 fi = Ei:

~- J2? - =J  !

i

.:f :=%f Vessr:1 flow (gpm) 77% -

120; of 370,D0'03T95lj .

• I i .. _ g g Axial Shape Index .527-+.527 ji.rsa . ::ipt . . 3. . - '

-- A= -

s. = 5 mi-:+it =- + .= = = - >

! _: r j 7 . - :m y r c5{pf 3.in+u:::-t+gm; : _ -

i

T-- ; : j _ . . . _ l

'1-U-I

- - . * *.:ru .- : . . . .h. -- :. = , ~ ~ ~ - t i ~;~ V 1

n 1 60 N

= = .. s

- -i.-. .f: : - . - u- _-- ~=~ 1. ,:::a+ - - -- -l - -L "A - - - *^ * ~~ 3 . =- T_=:1 _ _--r-

--.:_.=. ;

I l

"l i '"'  : _

i C 7 --i j j

"T T-1:

t

-= J f k 9't I  ;

~

-}

, ;. J i l

a -

} -- L {

Li \

i j l

. . c1 f

4 -  :

)

I .. >

0 -

I -, ,

.i w f3

]

i tr e er i

!I 43 b,, .

, .E9 ._ - (

4 2t 1 ;* I 9

^ . :.;

3  !

i lou en .;__

1 vr e  ;; t 4

g en --. ; }- .

[

O l IE .h '

j -"

e-r,; H l

0 L13

.=

--. t

\

) e-j U - - -

ir j

~ -

. A :s

24 I

, c7 G r i pa 2T d_ . .  !

3 . = ,

Z z_. _E.

a i 2

-=-

_:- r:]

.=.

. =E b.

i In

] _ d ' ;__b -

.;2 .

~f =U

_( s =

M

}

! .: l ej ..

-*J O

._._._,__,._,_g

._; .1:_._ - ,l . , l :*. , 1-g. g . . F-- - 3_. .,

. ..lia . .

j u - 1 ., 1up ,21wi+q: - ;p - 1;.7 .]

- q: -

q .;.+

l Detailed TORC MDMBR

, 68-3

Figurc 68-3: Comparison in MDNBR Between CETOP-D (Tuned) and Detailed n-cu - v;_5. TORC

_ _ :_a _. 4 -. 2- 9 2 _;.;_ ~~fW 50';GS 243. ..7CiaisT~JR)_===1:=5i3==L: 3 ._.

=ei==m == t== .

. =.; . . . =!L. n ....E -.q ,. p:v.c-i3-1+q:==-+1:r. -

.--t-- ,

u - t : - .ar-.. - - .- r i r. : :~

=:- 1=-~i-..

r-=:~=

. .1 c =.= . .] r m= ====:r =~=---- =- .--- ..l =- ~12-

~

..:u_r =. a ::. ::_. ;: z.

- g: 3 5=.= g :=;,= - ;=g-y= __- q=.:=y=:; /..g

- , _.; .; =.

.. . , _ p_= , z. . {q:-.;.=3=- y.=.g--. . =;- 5+ j . - _,., _. .

. . , . j ;m

.m m., .

.q. _

, O. - - .

^

.: i.; b 5$Yl' .'I')'E ~}i/ S I! ~ . 'h

\~ / t vi-p.:.s. . .:.L=.M .

5 Range of Operating Conditions ri V R f=i M T : E U 4 si .--4 ~l

= : :. s 5-- n a a .i 5 =_=5 = : : : == : =::: . : : =4

-(

M...y- 5 .:dj Inlet Temp.( f)

SystemPressure(psia) 530 - 505 E@MM(iM$K J:e:=:

=t 1785 - 2415 I-iti Z== .4

JIl,

--. :=.2 r

. :.L3 ...

Vessel Flow (gpm) 75% - 120% of:_r_: M..

E.L @ F.-R

. . Mi=?=.I .i. .M' 4lM...:r==

2--. . 396000 gpm .._.

Axial Shape Index _ .g= a .. . ".

iE .W-[_=~519'i:

r h: :>.= .527 +.527 - = . + . -m

. - =r. _ . =. . == ; -- -- a

=. t. . . . . .a . .: ;. ~

- 3

.;:: s_

_ . r 2.

v1 =

s=i

=t m =:- :::=.t: -===  :; =, .;....

:- -a ,,

_Iz-r =::..-- = -----1 === 7

=:2 = -- - 4

=.

=- .;:.1. _.. :1_ . .; . ...E , c,=;  :. . . 3, . . . =.

N I.T., e-e n

(

n nd v n!

b C

{

u O rl o,_ -

e E s. w Uf 4.,,)

2 .

Ei .

. M=

d.

g.r er u5 21 l>J X

b i p) (

v t

L- .

-.n- ' ' *

... .T~~~' -'4.- 4-~~~'- - ; .- r : d

. .. .. - , g ; L : _.

_. _- ~._.*._'c _.3 . :

t' 6 -

. ...?._

- _i t. ;'". ' :,

?.-,

}

.'n

_ . - . i-

='; ~

m: -

A:

,' :- j ,.

_; - ; . ; ; _3

-y

. s~ ::: : . :.Q.;.= - :_.;:. q . ] .- u .;k , .. d :m ,3..;g ,- - q: =r:::, ..-

=

- v= =_v :s.. .-. -

.l- -

s

. x.;

- >1 -s 1

Detailed TORC NDNBR 68-4 9

Question 492.6x

3 7

s/

In the development of the CE-1 correlation with uniform axial power distri-bution (CENPD-162), nultiple data points have been used for some heater rods with quadrant instrumentation in non-matrix subchannels. Therefore, the statistical ar.alysis should be re-performed by eliminating the redundant data poin;.s. Provide, based on the reduced data, new minimum DNBR limits for 14x14 and 16x16 rod bundles, separately and jointly.

Response ,

A technically consistent response to this question would require re-correlation of the new subset of test data (i.e., the subset with " redundant" data points eliminated) to produce revised values for the constants in CE-1.

Such an analysis is not feasible in the time frame available. However, a conservative approximation to the requested information has been obtained by performing a statistical analysis of primary DNB indications only, using un-revised CE-1. This approach is conservative because the constants in CE-1, not being optimized for the data subset being examined, will introduce spurious scatter into the resulting measured / predicted CHF ratios.

In Table 69-1, means and standard deviations for the measured / predicted CHF ar.e shown using data for the primary DNB indication only in each test run. The corrasponding 95/95 MDNBR limit is given separately for 14x14 and for 16x16 fuel types, as well as for the complete primary indication data s set. For comparison, the same information is provided for the total CE-1

( ,) data base (including multiple indications).

It can be seen that the primary-indication-only statistics compare well with the total CE-1 data base statistics, and continue to support an overall 95/95 MDNBR limit of 1.13.

O 69-1

-s TABLE 69-1 .

4 COMPARISC:t OF TOTAL CE-l DATA DASE

('s 's 111T11 PRll:ARY Drib IfiDICATI0ti SUBSET 14x14 16x16 Type 21 21 25 21 21 25 fio. Rods 12.5 7 TOTAL 7 12.5 7 7 Len9th in ft.

114 73?

169 157 All 141 99 51 n

Data Ilomber of ,

70 52 55 3 31 Data Points Primar.' 72 45 37 Data

. All D ta Mean of Cl!F !!eas.

tI D red. Primary

()' Data ,

1 was o All Standard Data Deviation of Clif lieas. Primary .

'CliT~Pred . Data -

1.133 All 1.122 1.139 -

95/95 Data

~

MD!iBR 1.122 1.123 Primary 1.136 Data Ov 69-2

.Que_stion 492_.70 .

A How would the licensee handle a 1 percent uncertainty on the TORC design code?

U Pesoonse r ,

As indicated in the response to Question 9 of the PNL questions of tiarch 27, 1931, C-E does not agree that a penalty is required to accommodate this uncertainty. How-ever, should such a penalty be arbitrarily imposed, C-E would handleit by an increase in the enthalpy rise factor equivalent to 1% DNBR and would then include this ad-justed enthalpy rise factor in a revised statistical analysis.

As a result of the !!ay 7,1981 meeting the NRC requested a nunerical and physical explanation on how the 1% TORC Code uncertainty (on 14DNBR) would be acconmodated in the statistical analysis. Also, explain why the 1.24 DNBR safety limit is still valid after applying the 1% uncertainty. This explanation is given below.

The 1.24 DNBR limit as reported in Ref.1 contained conservatism (which was later identified during Quality Assurance) in the statistical data for the CE-1 CHF Correlation probability distribution. In Table 5-1, Ref.1 the "mean" of [ ,]

'tould be [ ,]and the " standard deviation at 95% confidence" of f. , ] should ue [ ] These corrected values would justify a MDNBR limit of[ J,but it was elected to retain the reported value of 1.24 since that value was conservative.

The 1% uritertainty on the enthalpy rise factor can be approximated by taking the partial derivative of the response surface (Eq. 1) with respect to the enthalpy rise factor (x5 )

/_T _

U 7 2

11DNBR = bo +I b ,I bg$ (n$ -c) 5 n$ +1=1 1=1 7 7 Eq. 1

+I I b.. nj n.

i=1 j =1 1J J i<j

~ ~ ~ ~ '

2 a!1DNBR = (b5"5) , a(b55 ("55 -c) ) _ _

a

! x5 ax5 ax5

~

9

+ a(bl5 "1 5) + a(b25 "2"5)+ a (b35 "3"5) a axf x5 3XS n405) 0 Eq. 2 .-

+ 3(b + a (U5_6 "5 6 ) + a (b57 "5"7) ,

8X ax5 5 DX5 O

70-1

E l,

l -

F l

i  :

i U b a MDNBR

= - S+ - 2 X5 55 2 "5 b 55 + ,

_ 82

!9 ax5 5 5 62 5

i b b bl5 (X1 -"l) + -25 IX2 -"2) + ._3 5 IX3 -"3) f Sj B 5 62 65 83 85 * '

j-i .

i Eq. 3 i IX 4 4) f b45 +

b56 (X6 -"6 ) +

b57 (X77) 64 65 85 06 05 B7 l 4 _

i Because the 1% DNBR increase is being analyzed at nominal values of the system parameters: ,

i D ~

i X. -ai l

=0

B;  ;

i i

l Therefore, the last six terms of equation 3 drop out so that, j i

i l

g a!1DiiBR "

b p#

5

+

Sr2 b

55 IX 5 -"5) Eq. 4 l

a #  !

x5 l l i l j where, ,

i b5" l b55 *  !

4 B5"

! t

~

X5 _  !

"5 " ,

Inserting the above values into equation 4, B_11DNBR =[ }

ax5 i= t a l

l l

70-2

i l

For a 1.0'4 t2110!iBR (i.e.1.24 Df1B points)

AX5 *b I .

This penalty is then applied to the standard deviation (c) of the enthalpy rise factor distribution. From Table 5-1 (Ref.1), 5 "l Therefore the standard deviation for the penalized case (op) would be, op = a +Ax5 t- q U

p b J Based upon this value, using the SIGf tA code (Ref.1) a 95/95 Df1BR limit of f- ]was calculated.

O .

O I

i 70-3

. - _ . _ . _ _ . _ _ _ _ _ _ _ _ _ _ . . _ _ - - - - - - - - . ~ . . . , _ _ _ _ . , _ . . , , _ . _ _ _ _ ___ __ __ _

, Question 492.71 Provide a comparison of the most limiting transient including time and value of minimum D!iBR with and without Asymetric Steam Generator

- Transient Protection trip setpoint for ANO-2 Cycle 2.

Response

The instantaneous closure of a single MSIV (Asyametric Steam Generator Transient, ASGT) results in a minimum DNBR value of > l.24, as presented

~

in the All0-2 Cycle 2 Reload Analysis Report (Section 7.1.10).

The ~value of the RCS cold leg AT trip limit (ASGT protection trip) is based on the minimum margin of 117% set aside by COLSS. This margin is required for the 4-pump Loss of Flo'.t event which experiences the most rapid margin degradation to DNBR.

The AT trip setpoint in the CPC is that value of cold leg temperature differential, with appropriate allowances for uncertainties, which results in a power tilt in the core that degrades margin by an amount equivalent to that preserved by COLSS. However, since the instantaneous closure of a single MSIV does not result in a margin degradation rate as rapid as that for the loss.,of Flow event, CPC will actually tenninate the transient before a DNBR of 1.24 is reached.

I l

l

\

+

Question 492.72

( ,) In response to the question 492.62, the initial conditions for both of the events (loss of flow and CEA withdrawal) should include the initial

, power level, ra' dial and axial peaking ifactors and peak linear heat rate of the hot pin. Please provide this information for AN0-2 Cycle 1 and Cycle 2. Briefly describe the changes in values. Is the selected case representative of worst case operating conditions for core life?

If so, describe how this case was chosen, .if not, why not?

Response ,

Information requested has been incorporated into the response to question 492.62.

s_-

4 L

kJ 72-1

, Question 492.73- ,

There is a discrepancy between the CPC algorithm and the CETOP-D topical in the

(,J - _ two-phase friction factor multiplier. For pressure below 1850 psia, f2 (P, G) 6 is used for the multiplier with mass velocity G > 0.7 x 10 lb/hr- f t2 as described in the CETOP-D topical, whereas f3 (P, g) is used in the CPC algorithm.

Please clarify the discrepancy.

Response

The discrepancy is due to a typographical error in Table 2.2 on page 2-9 of the CETOP-D topical. The correct equations for2 f and 3f are as follows:

f2 = 1.26 - 0.0004P + 0.119 ( ) + 0.00028P ( ,)

f 3 = 1.36 + 0.0005P + 10 0.1 6 ( 6 ) - 0.000714P 10 6 (0)

Correcting this error in the CETOP-D topical will make it consistent with the O CETOP2 top i cal, CEN-143(a)-P - .

Note that the units for the mass velocity (G) in the CETOP-D and CETOP2 topicals 2

are not the same. In the above mentioned equations G is in lb/hr ft whereas in 2

Section 2.13 of CEN-143(A)-P;G is in lb/sec ft .

b.

1 O

73-1 n

,, Question 492.74 v

for single-phase friction '

correlation,F=0.184Refagtor,theCPCusesBlasius

.<, whichiseems to assume smooth surface. What is the value of surface roughness for the ANO-2 fuel cladding? How much difference in 8 friction factor is it compared to the Moody friction factor using the correct relative surface roughness.

Response

Surface roughness measurements on ANO-2 fuel rods have ranged from 14 to 21 micro inches RMS. These measurements have been documented in EPRI report RP 586-1 Task B, " Fabrication and Chardcterization of Arkansas Nuclear One Unit 2 16x16 Fuel Assemblics", dated October,1978. At nominal operating

(')T m

conditions, the single-phase friction factor is F = .0132 using the Blasius correlation compared to a single-phase friction factor of F = .0134 using the Moody friction factor

, and the correct relative surface roughness.

O

(_)

74-1

\ ,

n Question 492.75_

i s_.,  !

In your response to question 492.3, the pressure transport coefficient, flp, is[ ]forCETOP-2. However, the Celi-143 report indicates the transport coefficient CN, defined as CN = (No. of gaps)/Np, equal to

[ ] Please explain as to how many channel gaps and how they are How does the CPC modeling obtained in the CPC four-channel modeling.

differ from the CETOP-D core themal design modeling?

Response _

In CETOP-D and in CETOP2 pressure transport coefficients are used in the solut In the response to question 492.3 a of the crossflows within the hot assembly.

table was given in which typical pressure transport coefficients - used in CETOP-D F -'

respectively.

and in CETOP2 were presented. These , . .

coefficients were[ ,andt However,theexactvalueswillbe(jand respectively if exact values of Ngaps are used.

Within the codes the pressure transport coefficients are combined with the number of gaps along the lumped channel boundary:

U aps i.e., U p

In CETOP-D this term is applied in the solution of the crossflows between the The number of gaps buffer channel (channel 3) and the hot assembly (channel 2). ~

(The gap between a along the boundary between channel 3 and channel 2 is[-

fuel rod and the guide tube is jofthegapbetweentwofuelrods). -

?

9

\ ~

(_.) .,

These tems with Uphave been found to have only a minor effect on the cal'culation of DN3R.

i~ ,4

-The core thermal model used in CETOP2 is identical to that used in CETOP-D in

({})

tenns of geometry. The calculation procedures are also very similar. However due to the execution time requirements irhposed on CETOP2 by the CPC's some In simplifications in the calculations are made relative to the CETOP-D code.

CETOP-D axial nodes are typically used in design calculations. CETOP-2

~'

however uses ~ laxial nodes to calculate fluid properties and then expands to Another

!-- nodes by linear interpolation. DNBR is then calculated atfL ]nodos.

-_ simplification in CETOP2 involves the channel 3 calculations. [ L

} In CETOP2 the enthalpy transport coefficient is input as a constant. In CETOP-D this transport coefficient is internally calculated at each node, Finally CETOP2 uses some polynominal fits to

~

exponentials in order to reduce execution time.

4

()

d J

f 4

O 75-2 i _ _ . . _ _ _

, question 492.76 The response to question 492.62 indicates that, using the COSM0/W-3

() methodology for the ANO-2 Cycle 2 core, the e:timated minimum DNBR for loss of flow and CEA withdrawal transients with initial power level of 103 percent rated power are 1.115 7nd 1.121, respectively. These values are well below the allowable DNBR limit of 1.3 for the W-3 correlation ard infer that Cycle 2 and later power distributions were not fully censidere 1 in the FSAR analyses of these events. Explain and justify the adequancy of your methods for selection 'of bounding cases. Are the Cycle 2 analyses representative of Cycle 2 only? If so, is there any assurance that this reactor can operate at the licensed power level without excessive DNB trips throughout future cycles without further revisions to design methodology in order to achieve more thermal margin?

Response

{: There are practical limitations on the degree to which an FSAR analysis can fully envelope future cycles. In any case it is stated in the FSAR that primarily first-core parameters were used in the analyses. Changes at the AN0-2 facility, in this case the first reload, required changes to the technical specifications and incorporated methodology changes. In accordance with 10CFR 50.59 a submittal was tendered to demonstrate continued compliance with the basic safety criteria upon which the original SER was written. The methcJology change alluded to in your question is the change to TORC and its derivative code for CPC, CETOP, and the CE-1 correlation. In the years 5etween the ANO-2 FS tR analysis and the first ANO-2 reload amendment, the development of the TORC /Cr_-1 methodology came to its conclusion and has since been used with NRC approval to update the licensing bases of several C-E plants, including Calvert Cliffs Unit 1 and Unit 2 and Millstone Foint Unit 2.

76-1

\

In comparing the COSf40/b-3 DflBR value of 1.115 or 1.121 to the CETOP-D/CE-1 value of 1.24, one must consider that these comparisons were made at the eg point of minimum DflBR during the LOF and CEA withdrawal transients. These

( i transients were terminated by CPC trips preventing violation of the DNBR limit of 1.24. The CPC DilBR-trip projection and the allowed operating space were developed based on the more rigorous thermal margin methodology TORC /CETOP/

CE-1 rather than the more conservative COSM0/U-3. Itad the analysis been done with the COStiO/W-3 methodolcgy, CPC would have been designed to disallow a W-3 DflBR below 1.30 and more restrictive operating limits may have been required. An increase in the number of unnecessary DNBR trips could then have been a possible outcome;however, the comparison of DNBR values alone is not a complete indication of power capability.

Although there has been no change from Cycle 1 in the process by which LOFA is evaluated and factored into the CPC, we have appended a review of the t

relationship between the CPC-related analysis and the case presented in the reload report. - From our discussions we believe this may clarify some apparent misconceptions about the process itself.

The Cycle 2 analyses make use of physics data that should remain applicable to later cycles. Given stability of the licensing basis for this docket and the absence of marked changes in operating requirements, we would not anticipate the need for further revisions to the methodology to achieve more thermal margin for future cycles.

O V

76-2

Appendix to Question / Answer 492.76 As explained to Gene Hsii at C-E on May 28, 1981 a large number of loss of fTow g/

n cases were run for Afl0-2 Cycle 2 to determine the maximum required mar dn (under flow fraction, UFF, for COLSS and the'DNBR limit of Tech. Spec. Figure 3.2.4) and the CPC projection constants (KPTAU and RDf!BR). These cases were chosen based on a parametric study in axial shape index and at a pressure, temperature, initial flow rate, power level and axial shape which makes the UFF and CPC projection constants most conservative.

DftBR for these loss of flow cases was calculated with CETOP-6/CE-1. The UFF and CPC projection constants, along with the CPC Df!BR trip limit of 1.24, guarantee that any loss of flow which occurs during operation starting from within the LCO's (including the COLSS power operating limit defined using the UFF or Tech Spec Figure 3.2.4), will not result in a minium CE-1 DriBR below 1.24.

O' The loss of flow case presented in Section 7.1.8 of the Reload Analysis Report is a selected case from those limiting cases which define the UFF and CPC projection constants. CPC is formulated to guarantee that no loss of flow will result in a minimum Df!BR below 1.24. The CPC analysis is parametric in axial shape index, encompassing 4080 axial shapes.

Therefore, all cases would exhibit Of;BR's which are greater than 1.24.

This process is no different than that employed for Cycle 1 except for the different DNBR trip limit.

In the response to question 492.62, we estimated that the DilBR at the conditions corresponding to the point of minimum DNBR (1.24) for CE10P-D/CE-1 would be approximately 1.115 for COSM0/W-3. The fact that this DNBR is l'elow the allouable DilBp ' ait of 1.3 for the U-3 correlation does not in itself indicate that Cycle 2 conditions would not be acceptable if C05M0/U-3 were used in the analysis instead -

of TORC /CETOP-D/CE-1. If we had used C05f'.3/W-3 for Cycle 2, we would have 76-3

t.

i i:

S performed the entire spectrum of loss of flow cases using COSM0/W-3. The re-

'i i- '

quired margin, CPC projection constants, and LC0's would have been determined l

l i Lfrom the COSM0/W-3 analysis such that any loss of flow which ,would occur during 3-s operation starting from within the LCO's would not result in a minimum W-3 DilBR t'

below 1.3. A selected case would then have been presented in the Reload Analysis Report which would have illustrated the approach to a DfiBR of 1.3.

i-t i

}

1 i

l 1 0 ,

i 3

F 1

h i

i l

i l

i

. O i

s i

76-4

!~

Question 492.77

(~) As indi-

'l Your response to question 492.60 is not complete.

cated, there are chances when minimum DNBR might occur on channels other than a gui,de-tube channel. In these cases, the question is not only whether CETOP is more conservative than TORC. Rather, the concern is whether it is legitimate to use guide-tube channel modeling to represent other types of channels. If the guide-tube channel is to be used entirely, then it is necessary to prove that minimum DNBR neverOtherwise, occurs -

in other types of channel throughout the core life.

it must be shown that the guide-tube channel modeling always predicts the same or lower DNBR than other channels. Please be specific on your proof.

. Response -

,~

,L -)

~

[ ]theMDNBRpredictedwithdetailedTORC(inthecorner guide tube channel) will still be lower than any value in y

the core since the use of the CE-1 correlation in detailed TORC has been verified to produce conservative results relative to test n'casurements (Ref. 4 ) , which included MIF occurrences in both

~

the guide tube and matrix channels. The CFTOP-D MDNBR will also be conservative relative to the actua' MDNBR in A)

(- the core since it is benchmarked to detailed TORC .esults.

77-1

I 4

i i

h .

Ilowever, the corner guide tube channel does not have to be modeled in CETOP-D as the hot channel,since the benchmarking i

forces CETOP-D to be independent of the location of the hot

! channel and other model dif ferences between TORC and CETOP-D.

i i l i .

l l

i I f i

l i G l t

I t

i

?  !

t i

.I

.I l '

' l

!O l

4

[

i l

l .

I l

l 77-2

f im

( )

_REFEREllCES_ ,

(1) "l:esponses to first Round Questions on the Statistical Combination of Uncertainties Program, CETOP-D Code Structure and flodeling tlcthods (Cell-139-( A)-P), !! arch 1981.

(2) " TORC Code, A computer Code for Determining the Thermal flargin of a Reactor Core", (CEf1PD-161-P), July,1975.

(3) " Statistical Combination of Uncertainties, Combination of Systen Paramter Uncertainties in Thermal liargin Analyses for Arkansas fluclear One Unit 2", (CEft-139-(A)-P), flovember,19 80 (4) "CE Critical lleat flux Correlation for C-E Fuel Assemblics with Standard Spacer Grids",(CEf1PD-162-P-A), September,1976.

(5) CEf1PD-170-P and Supplement 1-P, " Assessment of the Accuracy of PWR Safety System Actuation as Performed by the Core Protection Calculators", July, 1975.

(6) CEft-35(A)-P, " Proprietary Versions of Responses to Core Protection Calculator System Questions 222.59-222.80 and 222.101-222.161," July 1976.

(7) CEN-139(A)-P, " Statistical Combination of Uncertainties," flovember 1980.

')

(8) CEti-63(A), "CPC/CEAC System Start-Up Test Requirements," July 1977.

(9) CEf1PD-153-P, Revision 1-P-n, "IllCA/CECOR Power Peaking Uncertainty" flay 1980.

O i

R-1 I

1 1

ATTACHMENT B I

i h

l;.

I f

, ., i, l

l i.

I h

I li i

l I

I l

l l

y -

i -

)s NRC Question During the NRC audit of the ANO-2 Cycle 2'CPC software there.was a demonstration of a Phase II DSVT case with the FORTRAN simulation code on the comparison between the Cycle 1 and Cycle 2 software. git was noted that the hot leg temperature in the Cycle 2 case datag was 2 less than the hot leg temperature

-in the Cycle 1 case data. Is this 2 difference due to the new pressure -

. temperature curve fits for enthalpy-temperature ratios implemented for Cycle 2?

Response ,

In designing the new STATIC program for the CPC/CEAC Systems software, pressure and temperature curve fits were developed for calculating the liquid .

properties required for the TORC /CE-1 DNBR calculation. The methods used yielded coefficients which calculated properties more closely approxmating the

-1967 ASME Steam Table values. The~same methods were utilized to determine constants for new pressure-temperature curve fits for the. enthalpy-temperature ratios that are consistent with those used for the liquid properties. As a result, the enthalpy ratios based on the new curve fits (TORC /CE-1 correlation)

- generate a static thermal power value approximately 2% greater than a static thermal power value generated with enthalpy-temperature ratios based on the old curve fits (W-3 correlation), when using the same cold leg and hot leg temperatures.

In order to compensate for this increased accuracy when making a Cycle 1/ Cycle 2 steady state or transient analysis comparison, the initiaf hot leg temperature for the Cycle 2 case data.must be reduced approximately 2 F below the Cycle 1 value-in order to generate a static thermal power value for Cycle 2 equivalent to that generated for the same Cycle 1. analysis.

s S

d S

e e

r

.. - , -. _ ._ . , _ _ _ .