ML20050E280
| ML20050E280 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 04/06/1982 |
| From: | Mills L TENNESSEE VALLEY AUTHORITY |
| To: | Adensam E Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8204130191 | |
| Download: ML20050E280 (10) | |
Text
ar TENNESSEE VALLEY AUTHORITY CHATTANOOGA. TENNESSEE 374ot N
400 Chestnut Street Tower II D
s April 6, 1982 RECEIVED.
~
t APR1219825 x)
Director of Nuclear Reactor Regulation
- ssutunamuemmu Attention:
Ms. E. Adensam, Chief g
8M(MR 84 Licensing Branch No. 4
//
Division of Licensing g
4 U.S. Nuclear Regulatory Commission Washington, DC 20555
Dear Ms. Adensam:
In the Matter of
)
Docket Nos. 50-327 Tennessee Valley Authority
)
50-328 Enclosed is our response to your letter to H. G. Parris dated February 12, 1982 regarding the request for additional information on hydrogen control for our Sequoyah Nuclear Plant.
If you have any questions concerning this matter, please get in to Ich with J. E. Wills at FTS 858-2683 Very truly yours, TENNESSEE VALLEY AUTHORITY
. M. Mills, Manager Nuclear Regulation and Safety Sworn t
- d. subscr be<'. before me this day of 1982
(
Notary Public
- b My Commission Expires Enclosure cc:
U.S. Nuclear Regulatory Commission Region II Attn:
Mr. James P. O'Reilly, Regional Administrator 101 Marietta Street, Suite 3100 Atlanta, Georgia 30303 l
I 8204130191 820406 q
l PDR ADOCK 05000327
\\
b' \\
PDR P
An Equal Opportunity Employer
ENCLOSURE RESPONSE TO NRC REQUEST FOR ADDITIONAL INFORMATION REGARDING HYDROGEN CONTROL SEQUOYAH NUCLEAR PLANT Question 1.
Provide a description of the HEATING 5 computer code you used in calculating thermal response of the equipment during a hydrogen burn.
Response
HEATING 5 is a heat conduction computer code written for the NRC-NRR by Oak Ridge National Laboratory (ORNL). The following description of the code is a copy of the abstract from the HEATING 5 manual.
"HEATINGS, a modification of the generalized heat conduction code HEATING 3, is designed to solve steady-state and/or transient heat conduction problems in one, two, or three-dimensional Cartesian or cylindrical coordinates or one-dimensional spherical coordinates.
The thermal conductivity, density, and specific heat may be both spatial and temperature-dependent. The thermal conductivity may be anisotropic. Materials may undergo a change of phase. Heat generation rates may be dependent on time, temperature and position, and boundary temperatures may be time-dependent. The boundary conditions, which may be surface-to-boundary or surface-to-surface, may be fixed temperatures or any combination of prescribed heat flux, forced convection, natural convection, and radiation. The boundary condition parameters may be time-and/or temperature-dependent. The mesh spacing can be variable along each axis. The code is designed to allow a maximum of 100 regions, 50 materials, and 50 boundary conditions. The maximum number of lattice points can be easily adjusted to fit the problem and the computer storge requirements.
The storage requirements on an IBM 360 machine range from approximately 250K bytes for one lattice point to 1256K bytes for 6000 lattice points.
The point successive overrelaxation iterative method and a modification of the " Aitken delta squared extrapolation process" are used to solve the finite difference equations which approximate the partial differential equations for a steady-state problem.
The transient problem may be solved using any one of several finite difference schemes. These include an implicit technique which can range from Crank-Nicolson to the Classical Implicit Procedure, an explicit method which is stable for a time step of any size, and the Classical Explicit Procedure which involves the first forward time difference. The solution of the system of equations arising from the implicit technique is
Page 2 accomplished by point successive overrelaxation iteration, and includes procedures to estimate the optimum acceleration parameter. The time step size for implicit transient calculations may be varied as a function of the maximum temperature change at a node.
Transient problems involving materials with change-of-phase capabilities cannot be solved using the implicit technique with this version of HEATING 5."
Question 2.
In your response to item A.1.C in your letten' dated August 27, 1981, you indicate that the temperature reached by the equipment during a hydrogen burn is below the qualification temperature for that equipment. Since the actual temperature reached by the equipment during qualification tests is not measured, demonstrate by analytical or experimental means that this temperature will exceed the temperature reached by the equipment during a hydrogen burn.
Response
Our submittal of December 1, 1981, included a section (2.2) on analytical model verification by comparison with results from another accepted computer program. Results of analyses were provided comparing the Westinghouse transmitter equipment qualification analytical model with the TVA transmitter model for MSLB conditions. The Westinghouse models are discussed in WCAP-8936. These analyses showed excellent agreement between the Westinghouse and TVA models (which were developed independently). The TVA analyses using the MSLB qualification heat flux resulted in higher transmitter temperatures than were produced in the analysis of hydrogen burns.
Question 3 The CLASIX analysis in your submittal is based on four arbitrarily chosen flame velocities of 1, 3, 6, and 12 fps. The 1 fps velocity case was chosen as a basis for the equipment survivability analysis. Although the 1 fps velocity flame would emit the largest amount of thermal energy to the equipment, it corresponds to only 6 burns in the lower compartment and 26 burns in the Upper Ice Condenser Plenum. In contrast, in some other cases analyzed more burns and higher containment temperature were predicted. Also, in some cases hydrogen burn in the Upper Containment was foreseen.
For example:
Case 1C (Table 18 in your submittal) predicts 19 burns in the Lower Compartment and 36 burns in the Upper Ice Condenser Plenum; Case 1F (Table 18) predicts 8 burns in the Upper Compartmentandthebasecaseinyoursensitivityagudies (Table 21) predicts containment temperature of 1578 F.
In view of the fact that these parameters could have significant effect on the temperature reached by the equipment, show that considering the 1 fps flame velocity case (1-G) as a basis for your equipment survivability I
analysis is a conservative assumption.
Page 3
Response
In May 1981, TVA submitted equipment survivability analyses. As a result of those submittals, a request for i
additional information from the NRC was sent to TVA. The request specifically stated that the " minimum flame front vislocity" should be used.
The survivability analyses performed and submitted on December.1, 1981, were based on an 8 v/o ignition criteria. The adiabatic flgme temperature of an 8 v/o mixture of hydrogen is 1400 F.
Case 1C, discussed in your question, used a 6 v/o ignition criteria. The adiabatic flame temperature of a 6 v/o mixture of
^'n hydrogen F, which is much lower than the case used to anal:
equipment.
In addition, the burns in case 1C are oi much shorter duration, and the ambient S
atmospheric temperature is lower.
Case 1F, which showed l-burns in the upper compartment, was based on an ignition criteria in the upper compartment of 4 v/o. Again, the associated adiabatic flame temperature is lower and the environment is less severe than the case used for equipment survivability analysis..Besides, there is no small critical equipment in the upper compartment other than ignitors and cable in conduit. These have already been analyzed for multiple burns based on ignition at 8 v/o, which produces a much more severe environment than 4 v/o, and shown to be acceptable.
The S D event used to evaluate equipment used a maximum 2
hydrogen release rate of 1.1 pounds / minute. As can be i seen from the table in section C on scenarios, this mass flow rate is appropriate 'for real plant transients.
)
The results presented in table 21 are for base case
. assumptions except that the mass flow rates for the blowdown are over three times the rate expected in a degraded core event. Therefore, this is another sensitivity study and should not be used for equipment survivability calculations. The containment transient based on the S D release rates ~provides a conservative 2
environmental profile.
Thus, the case chosen as a basis for equipment survivability analysis is appropriate due to the substantially higher flame temperature and longer burn dura,tions than would be found in other cases. We believe-s s
that it directly addresses your concerns with analyses 3'
submitted previously.
Question 4.
In your submittal you describe analytical and experimental evaluations of a selected number of safety-related equipment which has to survive hydrogen burn environment. Provide a detailed justification of 4
g
\\
i
N s
Page 4 your choice of equipment for these investigations. Show also that the results of the' investigations will provide bounding information for all the equiprent needed for safe shutdown of the plant.
Response
Our December 1, 1981, submittal includes a section
" Equipment Selection" which provides our justification of the choice of equipment for evaluation.
Lists of equipment required for safe shutdown have been provided in this submittal and justification provided for the acceptability'or this equipment.
Question 5.
List and discuss the assumptions (e.g. shape factors,'
emissivities, etc.) used in calculating the heat transfer from the stationary flame in the ice condenser. Provide a justification for your assumption that the temperatureg of hot gases below the fleme front drops abruptly to 250 F (see figure 12-1 in the sucmittal).
Response
The following response was provided by Duke Power Company:
"The computer program HEATING 5 was used to calcu' ate heat l
transfer to the ice condenser wall. The capabilities of HEATING 5 allow for heat transfer by conduction, convection, and radiation. This capability was utilized as follows:
1.
For conduction, the necessary thermal conductivities were obtained for all materials and input to the code. No further programming work is required for the conducting mode.
2.
For radiation, the geometry that was modeled was
]
radiation between parallel planes. Even though the nearest ice basket is 0.5 feet from the wall, the flame and hot gas layer radiating to the wall was assumed to be one foot thick. Calculation of the necessary shape factors was then performed by HEATING 5.
The emissivity of the flame and hot gasses was 0 3, in accordance with information from Dr.
Bernard Lewis.
3 For convection, a convection boundary condition was imposed on the wall as represented by figure 12-1.
The temperature of the convectivg boundary was conservatively assumed to be 250 F for the following reason. Recall that the ice condenser is an area of constant upward flow. A mixture of steam and noncondensible gasses enters the ice bed at the bottom through the lower inlet doors and passes up
Page 5 through the ice at 1-2 feet /second. The ice bed lowers the temperature of this mixture and condenses some of the steam. The maximum temperature in this region of thg ice. condenser, as predigted by CLASIX, is about 200 F.
A temperature of 250 F was selected as a worst case value to reduce the cooling effect of this flow on the wall and generate higher wall temperatures. The actual temperatures on the face of the duct immedigtely below the flame are, of course, higher than 250 F due to conduction and radiation from the hotter regions above, but this wall area is continuously cooled by the flow of the steam / gasses mixture from below.
It is our judgement that figure 12-1 represents, therefore, a conservative convective boundary condition for use in HEATING 5."
Question 6.
In your submittal, you do not discuss the effect of pressure differentials developed during a hydrogen burn across the air return and hydrogen skimmer fans. What are the values of these pressure differentials calculated by the modified CLASIX Code? Show that they do not exceed the limiting pressure differentials which the fans can withstand.
J I
Response
The submittal did not discuss the effect of pressure differentials developed during a hydrogen burn across the 2
i air return fans for two reasons. First, differential pressures across the fans from the upper compartment to the dead-ended compartments would only result from burns in the upper compartment which are not predicted by CLASIX to occur for base case or best estimate assumptions. Second, any differential pressures across the fans from the dead-ended compartments to the upper compartment would be prevented by the backdraft dampers below the fan assembly. The original design basis accident subcompartment load for the dampers bounds any differential pressures calculated during hydrogen burning.
s TVA requested the air return fan supplier, Joy Manufacturing Company, to analyze the fans for overpressure in the direction of fan flow. Their analysis bounded any transient loading that would occur during a hydrogep burn by assuming an equivalent static load of 50 lb/in g imposed on the fan assembly in addition to the normal fan operating loads. For simplicity, any overspeeding of the rotor due to a turbine effect was neglected. Critical areas of the fan assembly were examined and the calculated stresses were not above yield. Therefore, the air return fans are judged to be structurally capable of withstanding any differential pressure calculated'to occur during a hydrogen burn transient.
w e-T 7
Page 6 Question 7.
Provide Table 2.1 and Figure 2 3-1 missing in your submittal.
Response
The requested information is provided in attachment 1.
Question 8.
In the T B scenario analyzed by TVA, it is not clear that B2 the restoration of ac power and actuation of glow plugs (or random electrical ignition sources) would not come at a time when excessive amounts of hydrogen are present in the containment.
In this connection, discuss for your particular utility grid whether the failure of all ao power to the plant is an event of such low probability that this type of scenario can be justifiably excluded from consideration. If it cannot be excluded, discuss the appropriateness of providing an additional reliable backup electrical supply for the igniters.
Response
MARCH runs of the T B scenario were only performed as B2 part of an analytical effort to scope hydrogen release rates to compare with the values chosen for the TVA blowdown sensitivity studies and not because we believe T B to be a valid scenario for the Sequoyah Bp plant. In this regard, it can be seen from the data in our December 1, 1981, submittal that the peak T B hydrogen release rate is about two orders of magn $tude B
lower than the rates used in our sensitivity studies.
Further, only 35 pounds of hydrogen is released before core slump in the T B scenario, which would be B2 insufficient to allow a burn in the containment.
However, it should be emphasized that we consider the loss of all ac power at Sequoyah to be an extremely unlikely event.
In order to have a loss of all ac power at the plant, both onsite switchyards or all 13 transmission lines to the plant must be disabled followed by a failure of both trains of the diesel generators.
Because of this redundancy, the failure of all ac power to the plant is an event of such low probability that it may be excluded from consideration.
.r.-.4
.u_.,__,.a L4 W
..wh_.ba w-A--
_._ c M
F.
_--a__
--a 4
.1w h-ma-w a
m.
m
+
r h
e e
t d
1 ATTACHIStri ;
4 i
1 I,
4 I
a i
4 I
i l
I l
f I
t i
6
. -.+
w
,-re.
,_.-,,-e x-7--
r.,
d i
TABLE 2-1 SEQUOYAH CLASIX S D EQUIPMENT SURVIVABILITY CONTAINMENT RESPONSE
SUMMARY
2 FLAME SPEED = 1 ft/s I
Number of Burns LC 6
UP 26 l
UC 0
Magnitude of Burns (1bm) LC 101 UP 33 UC
]
Total H Burned (lbm) 1086 2
H Remaining (lbm) 451 2
Peak Temperature ( F)
LC 884 j '
UP 1114 UC 150 DE 170 Peak Pressure (1b/in a)
LC 24.8 UP 24.5 UC 24.3 DE 24.8 5
Ice Remaining (lbm) 7 71 x 10 1
4 l
f i
i I
l' w-mm m
w
-r 4
+-ee--
--8 ss+,-
g..
,v.-.-
-r p.
-_4-,
p u
T-Md
'P";'"
- M--*-
e e
q e
p 3
oc w
6 m
t CW D
a IQ
.Ue
+
a C
M C
ha E
O a
W B
be
+
te C
M p.
S in O
U U la
~>
W c
U M
a 3 4 9 u O
c me O Oc
=
r
.to C
O U
k.
U O
l n
I I
O l
~~~~
-.,~--
om
-_~
- -==~-
"" A b3 s
s e
D2 M g
3 4.
l n
% O C U n.
D M N c f
N
._._-_____.____"M l
k
- ., e V
W
". M ll
.= W 0
W 00 W ta g
=
e Es.
ta lC P
O OH
= W
=<..Jc O
g,=
.4 I
Ed to I
CO
-- O*
I
~
1 I
i I
i L--
I I
I C
e I
e l
l t
i L
l i
l i
I l
l l
I #
C C
O O
O C
C C
C C
C O
C O
O C
O O
O O
C O
C
.O O
O O
.O O
O CC N
c c
M N
M.
.N l
=
(
I i
(,I ) a:nse:ad:e,1
,/ ~
,