ML20246D685

Forwards Response to NRC Review Questions on Plant Revised Reflood Heat Transfer Model for LOCA Analysis.To Address Recent Change in App K,Blowdown Heat Transfer Logic Modified to Prevent Use of Dougall-Robsenow Correlation

ML20246D685
Person / Time
Site:	Yankee Rowe
Issue date:	05/02/1989
From:	Papanic G YANKEE ATOMIC ELECTRIC CO.
To:	Fairtile M Office of Nuclear Reactor Regulation
References
BYR-89-80, NUDOCS 8905110007
Download: ML20246D685 (15)

Text

- - - - - - -

YANKEEATOMICELECTRIC COMPANY "*fE"%"s*g];";"l" M a 580 Main Street, Bolton, Massachusetts 017404398

• Ya&xes

%. May 2, 1989 BYR-89-80 United States Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 Attention: Mr. Morton Fairtile, Project Manager Project Directorate I-3 Division of Reactor Projects I-II

References:

(a) License DPR-3 (Docket No. 50-29).

(b) YAEC letter to NRC, FYR 88-05, January 5, 1988.

(c) NRC letter to YAEC, NYR 88-163, June 29,1988.

(d) NRC letter to YAEC, NYR 89-92, March 14, 1989.

Subj ect: YAEC response to NRC review of revised reflood heat transfer model for YNPS LOCA analysis.

Gentlemen:

In Reference (b), Yankee Atomic Electric Company (YAEC) submitted modifications to the Evaluation Model for Loss-of-Coolant Accident (LOCA) analysis of Yankee Nuclear Power Station (YNPS). The submittal consisted of a new steam cooling model and a revised reflood heat transfer correlation.

In Reference (c), the NRC approved use of the new steam cooling model. In Reference (d), the NRC requested additional information on the revised reflood hest transfer model referred to as the YAEC FLECHT/SEASET correlation. In particular, YAEC was requested to provide additional benchmarking of the correlation and to demonstrate its applicability to the non-FLECHT bundic geometry of YNPS.

Attachment A to this letter describes (a) additional benchmarking of the YAEC FLECHT/SEASET correlation demonstrating its applicability over a wider range of reflood conditions, and (b) modifications to the model which justify its use for YNPS bundle geometry. The recent change in Appendix K regulations regarding continued use of the Dougall-Rohsenow post-CHF heat transfer correlation is also addressed.

We trust that the information provided in Attachment A to this letter adequately addresses the NRC review questions contained in Reference (d). If you should require additional information, please feel free to contact me.

Very truly yours, YANKEE ATOMIC ELECTRIC COMPANY p

f e^ w - _ -

/ / l i

Georg Papanic, S6nior Project Engineer l Licensing I l GP: cam cc: USNRC Region 1 i USNRC Resident Inspector, YNPS l Attachment

l

- I i.. ,

ATTACHMENT A Response to NRC Review Questions on the Yankee Nuclear Power Station Revised Reflood Heat Transfer Model i

l l

l

\ 4 l

l p \

l '. '

l l 1 Response to NRC Review Ouaatiana l On The Yankee Nuclear Power Station Revised Reflood Heat Transfer Model MCK9ROUND The NRC review (1) of the proposed Yankee Atomic Electric Company (YAEC))

FLECHT/SEASET correlation for the Yankee Nuclear Power Station (YNPS)(2 identified several issues that needed to be further addressed. These are summarized below:

1. The YAEC FLECHT/SEASET correlation for YNPS was derived using data from the FLECHT-SEASET test program (3). This data base was used because it simulated a 17 x 17 core rod geometry representative of the YNPS fuel.

The staff expressed a concern that the correlation was based on a narrow data range of reflood conditions.

YAEC is addressing this concern by providing additional benchmarking for wider ranges of initial clad temperature, flooding rate, and inlet subcooling.

2. A second concern of the staff was that the bundle geometry, i.e., hydraulic and rod diameter, on which the correlation was based differs from that of YNPS. Namely, the correlation was based on a FLECHT-SEASET hydraulic diameter of 0.465 inches and a rod diameter of 0.374 inches compared to 0.412 and 0.365 inches, respectively, for YNPS.

YAEC is addressing this concern by adapting the FLECHT/SEASET correlation to non-FLECHT bundle geometries using a scaling method which preserves total energy per unit flow area between the reactor and FLECHT rods.

This approach was adopted for the FLECHT-SEASET overlap tests (0) in differences between the order 15 x 15to FLECHT assess the loweffects flooding of rate bundle cosine geometry tests (5) and the more recent 17 x 17 tests. The results showed that the effect of bundle geometry was minimal when the tests were scaled to preserve energy per unit flow area.

3. An additional issue is the recent change in Appendix K regulations regarding continued use of the Dougall-Rohsenow correlation in Evaluation Models. Appendix K, 100FR, rart 50, now states that:

"If the applicant or licensee submits proposed changes to an approved evaluation model, or submits corrections to errors in the evaluation model which significantly reduce the overall conservatism of the model, continued use of the Dougall-Rohsenow correlation under conditions where nonconservative heat transfer coefficients result would no longer be acceptable. For this purpose, significant reduction in overall conservatism has been defined as a " net" reduction in calculated peak clad temperature of at least 50*F from that which would have been calculated using existing evaluation models."

7716R

l l

The mc'e? changes described in the YAEC submittal of January 5, 1988(2) and this response are expected to result in a net reduction of greater than 50'F l in the calculated peak clad temperature. Therefore, YAEC is conforming to the l Appendix K requirements stated above by demonstrating that use of the J Dougall-Rohsenow correlation will be restricted to conditions where  !

conservative heat transfer coefficients are calculated.

Ad.ditidnal Benchmuking_oL110.pssad FLECHT/SEASET Correlation The YAEC FLECHT/SEASET correlation (2) was justified by comparing the calculated peak clad temperature using the new correlation in the T00DEE-2 computer code, with FLECHT-SEASET tests in the range of reflood conditions expected for YNPS. Eight FLECHT-SEASET tests were simulated; five of which were used to develop the correlation. Generally, the T00DEE-2 predictions were shown to be conservative with respect to data.

Further justification of the new correlation is provided here by five additional FLECHT-SEASET test comparisons with a wider range of conditions for initial clad temperature, flooding rate, and inlet subcooling. The test matrix is shown in Table 1. The first two tests expand the range of initial clad temperature. The third test expands the range of flooding rate. The fourth test expands the range of inlet subgooling. These four tests, along with the eight previously reported tests (2), envelope the range of reflood conditions expected for YNPS. The results of these comparisons are shown in Figures 1 through 4. As with the previously reported benchmarking matrix (2), the T00DEE-2 results in Figures 1 through 4 are generally conservative with respect to data. An additional test at a very high flooding rate of 6.1 in/see was also simulated to examine correlation trends with respect to flooding rate. The results, shown in Figure 5, further demonstrate the predictive capability of the correlation.

Additional ModeLpevelopment The YAEC FLECHT/SEASET correlation (2) consists of a set of time- and elevation-dependent multipliers to the heat transfer coefficients given by the Westinghouse FLECHT correlation given in WCAP-7931(6). The correlation will be used in the T00DEE-2 computer code to calculate the LOCA reflood heatup for YNPS. However, th- .NPS hydraulic diameter and fuel rod diameter differ from that of FLECHT-SEASET. Therefore, some additional model changes are required.

The FLECHT correlation, as used in the T00DEE-2 computer code, is based on the assumption that the heat transfer coefficient at a particular elevation is dependent upon the integrated energy distribution up to that elevation. This assumption is implemented through the use of the equivalent FLECHT elevatien model(8). Currently, this model accounts for short cores and non-FLECHT axial power distributions. The model will be modified to account for non-FLECHT hydraulic and rod diameters.

The parameters to be modified in the model are the equivalent maximum power (7) (Qmax) and the equivalent elevation (8),

7716R l

l l

4

- The current Qmax defines the peak power in a reference FLECHT rod such that the power integrated over the length of the reference FLECHT rod is equal to

- the power integrated over.the length of the reactor fuel rod. The modified model uses the same energy balance except that the energy balance is performed -

on a unit channel flow area basis. In particular, the Qmax parameter is now derived from the following balance equation:

Pg dz/Af= Pr dz/A r where:

Pr = Linear power in FLECHT rod, kW/ft.

Pr = Linear power in reactor rod, kW/ft.

Lg = Length of FLECHT rod (12 f t).

Lr = Length of reactor rod, ft.

Af = Channel flow area for FLECHT rod, ft2 Ar = Channel flow area for reactor rod, ft 2, The geometry modification also impacts the equivalent elevation model. The current equivalent elevation model assumes the reference FLECHT rod has the same diameter and thermal properties as the reactor rod. To account for the effect of rod size, the model is changed so that the true reference FLECHT rod diameter and thermal properties are used. In addition, the reference FLECHT axial power shape is now taken from the FLECHT-SEASET tests (i.e.,

Pmax/P avg equal to 1.66 instead of 1.622 as in Reference 6).

Note that the model changes described above are intended for non-FLECHT bundle geometries. Hence, they do not affect the results of the FLECHT/SEASET correlation benchmarking reported in Reference 2 and in the previous section.

Impact of Model Changes on YNPS Application The reported sensitivity studies for characteristic YNPS cosine and skewed power shape cases (2) showed that use of the proposed reflood heat transfer model resulted in lower Peak Cladding Temperatures (PCTs). The same YNPS cases were reanalyzed with the modifications described in the previous section. The resulting PCTs are compared to the results reported in Reference 2 in Tables 2 and 3. As can be seen, the modification of the FLECHT/SEASET correlation to account for non-FLECHT bundle geometries has a j small impact on PCT for both YNPS cosine and skewed cases. This is generally consistent with reflood effects having less significance for shorter cores (or lower elevations).

l Ernpnsed Reflood Heat Transfer Model and the Dougall-Rohsenow correlation '

4 As shown in Tables 2 and 3, the proposed YNPS LOCA reflood heat transfer model is expected to yield a " net" reduction of greater than 50'F in the calculated PCT. Therefore, in order to conform with Appendix K requirements, YAEC will i prevent nonconservative use of the Dougall-Rohsenow post-CHF correlation.

i 7716R  ;

l 1

l

9 .

The selection of a post-CHF correlation.in the RELAP4-EM code (9) is controlled by the IMCL (or IMCR) flag on the Heat Slab Data input. If the flag is 1, the Groeneveld 5.7 correlation (Mode 5) is evaluated.

. Additionally, if pressure is less than 500 psi, the Dougall-Rohsenow correlation (Mode 9) is evaluated. The minimum of Groeneveld 5.7 and Dougall-Rohsenow is then chosen as the film boiling heat flux. For the case where IMCL equals 2, the Groeneveld 5.7 correlation is not evaluated, and Dougall-Rohsenow is the film boiling correlation.

The current LOCA evaluation model for YNPS controls the selection of a post-CHF correlation by using a value of 2 for IMCL. That is, the ,

Dougall-Rohsenow correlation is used as the film boiling correlation  !

regardless of pressure.

The selection logic in RELAP4-EM will be modified to prevent nonconservative use of the Dougall-Rohsenow correlation. The value of the IMCL flag will be set to 1. This selection assures that the Appendix K recommended Groeneveld 5.7 correlation will be used if the pressure is greater than 500 psi. Below 500 psia, the selection logic uses the minimum of the Groeneveld 5.7 and the Dougall-Rohsenow correlations to calculate the film boiling heat flux. As described in RELAP4-EM(91, the Groeneveld 5.7 correlation is prevented from misuse near its low pressure singularity. This selection logic ensures that the Dougall-Rohsenow correlation is not used nonconservatively.

The film boiling heat transfer coefficient has an impact on the blowdown phase of the LOCA. Therefore, blowdown sensitivity calculations were performed for top skew and cosine power shape cases. The impact on cladding temperature varied from node to node with a maximum increase in cladding temperature of 50*F. The temperature at the end of blowdown is carried into the reflood phase. Therefore, the reduction in peak cladding temperature obtained from the revised reflood heat transfer models (Tables 2 and 3) will likely be offset by the film boiling heat transfer modification.

SUMHARY This response provides additional benchmarking of the YAEC FLECHT/SEASET correlation and additional model development to account for the non-FLECHT YNPS bundle geometry. Thus, the NRC concerns regarding the proposed YAEC FLECHT/SEASET correlation for YNPS have been addressed. To address the recent change in Appendix K regulations, the blowdown heat transf ar logic has been modified to prevent the nonconservative use of the Dougall-Rohsenow correlation.

REEEEENCES

1. Memo, M. B. Fairtile (USNRC) to YAEC, " Meeting Summary (TAC No. 66993),"

NYR 89-92, dated March 14, 1989.

2. Letter, G. Papanic, Jr., to M. B. Fairtile, "LOCA Reflood Heat Transfer Models," FYR 88-05, dated January 5, 1988.

7716R

4 -

3. M. Ji Loftus, et al., "PWR FLECHT-SEASET Unblocked Bundle, Forced and Gravity Reflood Task Data Report," Vo ume 1 2, Appendix C, NRC/EPRI/ Westinghouse Report No. 7, kcAP-9699, September 1981.

4. N. Lee, et al., "PWR FLECHT-SEASET Unblocked Bundle, Forced Gravity Reflood Task Data Evaluation and Analysis Report," NRC/EPRI/ Westinghouse Report No. 10. WCAP-9891, February 1982.

5. E. R. Rosal, et al., "FLECHT Low Flooding Rate Cosine Test Series Data Report," WCAP-8651, December 1975.

6. F. F. Cadek, et al., "PWR FLECHT Final Report Supplement," WCAP-7931, 0ctober 1972.

7. Exxon Nuclear Company WREM-Based Generic PWR ECCS Evaluation Model:

Further Definitions and Justifications to Reflood Heat Transfer Models,

'XN-75-41 (Supplement 1), dated August 14, 1975.

8. Exxon Nuclear Company WREM-Based Generic PWR ECCS Evaluation Model:

Supplementary Information, XN-75-41 (Supplement 7), dated November 9, 1975.

9. WREM: Water Reactor Evaluation Model (Revision 1), May 1975, Division of Technical Review, U.S. Nuclear Regulatory Commission, NUREG-75-056.

I 7716R

. l IABLE 1 ELECllI-SEASET Test Matrix for Additional Benchmarking _Qf XAEC FLECllT/SEASET Correlation Initial Clad Inlet Pressure Flood Rate Temperature Peak Power Subcooling leal 1931a1 (in/sec) (*F) (kW/ft) (*F) 34420 39 1.53 2045 0.74 142 30817 39 1.53 987 0.70 138 31302 40 3.01 1597 0.69 141 35114 40 0.98 1638 0.74 14 31701 40 6.10 1601 0.70 140 i

7716R

L' TABLE 2 Reflood Heat Transfer Model CDmparissas - Peak Cladding Temperature (PCT) for YNPS Cosine PCT With Existing Reflood Heat Transfer Model(a) = 2149.7'F PCT With New Steam Cooling Model and FLECHT/SEASET Correlation (a) = 2022.1*F PCT With New Steam Cooling Model and Modified FLECHT/SEASET Correlation = 2033.2*F i

(a) PCT result reported in Reference 2.

7716R

IM LE._3 Reflood Heat Transfer Model Comparisons - Peak Cladding Temperature (PCT) for YNPS Skewed PCT With Existing Reflood Heat Transfer Model(a) 1951.2'F PCT With New Steam Cooling Model and FLECHT/SEASET Correlation (a) = 1816.5'F PCT With New Steam Cooling Model and Modified FLECHT/SEASET Correlation = 1804.9'F i

\

l (a) PCT result reported in Reference 2. i

_e ..

7716R I

i l

1

8 g i i . . .

O e

O h e.

e -

a ,

O g O

Q O- O O

W 0 -

w$ @

S W e 4

1 o

8

~

0-s CALCULATED VS MEASURED o i i i i i 00 2 4 6 8 10 12 ELEVATION (FT)

FIGURE 1 TOODEE FLECHT SEASET SIMULATION, FLECHT/SEASET TEST 34420

h O e

8 w , . , , ,

o -

0 4

9 O@e g -

e ee d

12 8

' e g O

9 E e R0

@8 k"

8 *! -

9 LL) k 2o 8

Oe O O 8

~

CALCULATED VS MEASURED O -

O ,

O

, , , i 2 4 6 8 10 12 ELEVATION (FT)

FIGURE 2 TOODEE FLECHT SEASET SIMULATION, FLECHT/SEASET TEST 30817 l

+. .- , ,

t 8

0 N

O ~

- Og f

O b G gg O

C hO -

O _

O W4 2

• 5 Q.

O

• O .

WO -

e 4

' s <

g _

m CALCULATED VS MEASURED 8eo O , , i 6 8 10 12 00 2 '

4 ELEVATION (FT)

FIGURE 3 TOODEE FLECHT SEASET SIMULATION, FLECHT/SEASET TEST 31302

S '

4 o '

O i i i i i N

8

c. .

n O g

b @ e @

$ e ee a

t- o e g -

w-@ e e s @

w .

O H

e 20 -

.e ~

ge CALCULATED VS MEASURED I l I I I 00 2 4 6 8 10 12 ELEVATION [FT)

FIGURE 4 T00DEE FLECHT SEASET SIMULATION, FLECHT/SEASET TEST 35114

.6.. , .

1 O.

O , , , ,

g, G g ee e '

e e

e e CO e

3 x v o e T e x

W $

o_.

2 YO xo e

<m CALCULATED VS MEASURED 8

0 i i i i i 0-0 2 4 6 8 10 12 ELEVATION [FT) .

FIGURE 5 T00DEE FLECHT SEASET SIMULATION, l

L FLECHT/SEASET TEST 31701

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ - _ - _ _ _ _ _ _ _ - _ _ _ -_-