ML20088A808
| ML20088A808 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 04/10/1984 |
| From: | Mills L TENNESSEE VALLEY AUTHORITY |
| To: | Adensam E Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8404130151 | |
| Download: ML20088A808 (6) | |
Text
--
TENNESSEE VALLEY AUTHORITY CH ATT ANOCGA, TENNESSEE 37401 400 Chestnut Street Tower II April 10, 1984 Director of Nuclear Reactor Regulation Attention:
Ms. E. Adensam, Chief Licensing Branch No. 4 Division of Licensing U.S. Nuclear Regulatory Commission Washington, D.C.
20555
Dear Ms. Adensam:
In the Matter of
)
Docket Nos. 50-327 Tennessee Valley Authority
)
50-328 Enclosed is our response to Question No.1 transmitted by your August 18, 1983 letter to H. G. Parris mgarding additional information on the hydrogen mitigation system for the Sequoyah Nuclear Plant. The response for the remaining questions was transmitted to NRC by my November 1,1983 letter.
If you have any questions concerning this matter, please get in touch with Jerry Wills at FTS 858-2683 l
Very truly yours, I
TENNESSEE VALLEY AUTHORITY 4
L. M. Mills, Manager Nuclear Licensing
- Sworn, nd subsor bed pefore me I
this /
. day of
_ 1984 k
Notary Publ'io My Commission Expires @-5-P4 -
Enclosure 00:
U.S. Nuclear Regulatory Commission (Enclosure)
Region II Attn: Mr. James P. O'Esilly Administrator i
101 Marietta Street, NW, Suite 2900 Atlanta, Georgia 30303 pool i l 8404130151 840410 PDR ADOCK 05000327 p
PDR 1983-TVA 50rH ANNIVERSARY An Equal Opportunity Employer
ENCLOSURE
^
^
,1.
Qu~~ti"n With regard to the CLASIX code, the Staff has previously requestedThe clarification of the structural heat sink heat transfer models.
following pertinent points have been derived from the responses:
Heat transfer is based on temperature difference determined by A.
(T ulk - Twall)*
b Heat transfer coefficients for degraded core accident analysis are B.
determined from a natural convection (stagnant) correlation j
applicable to condensation heat transfer.
l i
CLASIX does not explicitly model mass removal due to condensation C.
heat transfer.
Based on the description of the CLASIX structural heat sink model, it appears that the CLASIX model differs dramatically from generally accepted approaches and is not, as is claimed, consistent with standard methods such as those used in CONTEMPT. The differences are related to the treatment of the three items cited above. By comparison, previously accepted approaches are daaracterized by the following:
A.
Heat transfer is based on (Tsat - Twall), when the surface less temperature of the heat sink is less than Tsat (i.e., Twall than T at*
s Heat transfer coefficients are based on condensation only when B.
Twa11 less than Tsat*
Condensed mass removal is based on condensation heat transfer with C.
provisions for revaporizing a small fraction of the condensate.
A more detailed description of accepted practice is contained in NUBEG-0588 and NUREG/CH-0255.
The effect of the CLASIX models would appear to be the desuperheating of the atmosphere too rapidly thus reducing gas temperatures and possibly altering the combustion characteristics.
Based on the above discussion, provide justification for the models incorporated in CLASIX or provide the results of analyses with The analyses should encompass acceptable models as outlined above.
selected sensitivity studies to assure that the effects of the changes are determined for both containment integrity and equipment survivability considerations.
Response
The following additional information is provided concerning the method by which CLASIX models heat transfer to the passive heat sinks.
i)
A.
The previously accepted approaches for heat transfer models d ascribed in.
your request for additional information were primarily devel3 ped foe LOCA containment analysis which emphasized pressure response. Ducause the I
LOCA blowdown is introduced into the containment as saturated liquid (see
' reference'1), the containment atmosphere steam-water component is saturated, except for a brief early portion of the transient.
?5 Consequently, the maximum temperature is the saturation temperature corresponding to the steam partial pressure at the maximum total i
pre ssure. There fore, for a LOCA, the saturation temperature is equal to 7
the bulk temperature so that either could be used to determine the energy removal. For a main steam line break or for hydecgen combustion, the containment atmosphere will consist of superheated air and steam.
m Therefore, the mass and energy removal models that were developed for Z
LOCA analysis will not correctly predict the peak containment temperature
_=
for these events. Recognizing this, the NRC has sponsored research to
=
establish acceptably conservative, yet mechanistically sound, models of h
heat transfer from superheated air and steam. This research (see reference 2) has shown that the temperature difference appropriate to 1
passive heat sink heat transfer should be based upon the use of the g
atmosphere bulk temperature (as in the CLASIX code) rather than the saturation temperature.
[
as B.
Heat transfer coefficients were never determined from a natural 4
convection (stagnant) correlation applicable to condensation heat j
transfer in any CLASIX analysis wa performed. The CLASIX condensing heat B,
transfer coefficients are based on the stagnant portion of the Tagami T
heat transfer correlation when Twa11<Tsat. A comparison of the heat transfer coefficients from this correlation with the Uchida correlation used in CONTEMPT 4/ MOD 3 (reference 4) is provided in Table 1.
An option f
available in CLASIX is to compare the rate of heat transfer from the 4
Tagami coefficient to the rate of heat transfer from the natural 5
convection coefficients from Kreith (reference 3) and then use the larger
]
of the two rates. However, since we never selected this option, the heat transfer coefficients are based on condensation only for Tw311 less r
than Tsata C.
In CLA2IX, condensate mass is not explicitly removed from the atmosphere 7
by condensation at the walls but is instead based upon a mechanistic
(
evaluation of the thermodynamic state of the atmosphere using the total internal energy in the atmosphere at the end of a time step. The iterative procedure which determines the rate of condensate annu mmoval f
is described fully on pages 38-46 of reference 5.
CLASIX's condmsate model does not include any provision for revaporizing an arbitrary y
fraction of the condensate. This treatment is conservative because it N
takes no credit for an atmospheric temperature decrease resulsing from a g
reduction in specific energy due to condensate revaporization.
It is asserted in your question that the effect of the CLASIX passive heat sink model would appear to be the desuperhesting of the atmosphero
^]
too rapidly, thus mducing gas temperatures and possibly altering the combustion characteristics. The period of greatest interest in the -
7 analysis of hydrogen burns is during and imediately following a burn.
M m
$-e l
- -
- / _;,
During this p;riod, tha CLASIX condensata removal model does not cignifictntly affiot the results since the hemt sink surface tsmparature
~
in thegartments in which hydrogen burning occurs is quickly elsvated above-the saturation temperature due to energy deposition from convective
< - batransfer. Ace the fact that CLASIX does not y_
^
+-
explicity model condensate mass mmoval at the wal$*IsTEElEFa'nt since no condensate would form on the walls during this period.
More importantly, the principal cooling mechanism for the lower compartment atmosphere is the cooler air flow from the upper compartment due to operation of the air return fan. The cooling effect of the air I,
retum fan will cause the formation and subsequent deentrainment of water l
droplets from the atmosphere. The thermodynamics condensation model in CLASIX should accurately predict this phenomena.
l
'~
$$u nca c m In summary, CLASIX handles heat transfer in a manner consistent with the In addition, the physical processes occurring in the containment atmosphere.
conservatism of the CLASIX heat transfer coefficients and the technical support provided for the structural heat transfer models provide sufficient 1
~
assurance that the current CLASIX msults are conservative without any further analysis.
e 4
h a
2 9
.s P4
^
(
e-h.
. g
.p
' l T.
'~
(
.~)
REFERENCES
- 1.,,, Wheat,, L. 3L.,, Wagnar,,R. J.,.Niederauer, F. G., Obenchain, C. F.,
" CONTEMPT-LTA...A. Computer. Program for Predicting Containment Pressure-a s onsanto a. Loss qf Coolant Accident,," Aer,olet Nuclear _
T t
~-~+--m empeca :are s e p s
s 2
~
Company Report, ANCR 1219 (June 1975) and SDR-83-76~ TApril"i976)
" ~'
2.
Lamkin, D., Koestel, A., Gido, R., and Baranowsky, P., " Containment Main Steam Line Break Analysis," NUREG/CR-1511, June 1980 3
- Kreith, F., " Principles of Heat Transfer," 2nd Edition, International Textbook Company, Scranton, Pennsylvania, May 1967 4.
Cheng, T. C., Metcalfe, L., Hartman, J., Mings, W., and Crail, A.,
"COMTEMPT4/ MOD 3, A Multicompartment Containment System Analysis Program,"
NUREG/CR-2558, December 1932 5.
Fuls, Martin, G., et al., "The CLASIX Computer Program for the Analysis of Reactor Plant Containment Response to Hydrogen Release and Deflagration," Document No. OPS-07A35, Offshore Power Systems, Jacksonville, Florida.
rw e d.M w..,,v,#,4 s..
L i 1
a 3'
'o
' ~
n
.m e -
TABLE 1 JEAT, TRANSFER COEFFICIENTS
... _. _5t? ?~W~"
my:P;' =-====
~ ~ ~ ~ * *
??'
,. ~....
,_;nn.,an_n _
~
ff L G } Q Q ^ 2 7 T g } t; g, gf - - ---- W ~ ~ ~
g Mass of Noncondensible (Btu /hrft F)
(Btu /hrft F) 0.02 2.0 3.0 0.05 4.5 8.0 0.10 7.0 14.0 0.20 12.0 21.0 0.25 14.5 24.0 18.66 29.0 m.,~-, -.
23'74' ~ ~, ~ ~. ~."37.0~, ~_.
~- '
~
=a
-29.77
- 46.02-
-=
~
.=
- x..
0.77 40.46 63.0 1.25 64.5 98.0 2.0 102.0 14 0.0
'I l
l s
a:
s
~
s.
t w
1--
-