ML20034G626
| ML20034G626 | |
| Person / Time | |
|---|---|
| Site: | 05200001 |
| Issue date: | 03/04/1993 |
| From: | Fox J GENERAL ELECTRIC CO. |
| To: | Poslusny C Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 9303100254 | |
| Download: ML20034G626 (13) | |
Text
_
_ ~
I GE Nuclear Energy i
con ow:n cw;,m l
w,
%eu &c J :. %D5 i
i March 4,1993 Docket No. STN 52-001 l
Chet Poslusny, Senior Project Manager Standardization Project Directorate Associate Directorate for Advanced Reactors
]
and License Renewal Office of the Nuclear Reactor Regulation i
Subject:
Submittal Supporting Accelerated ABWR Review Schedule - ABWR Probabilistic Fkxxling Analysis l
f
Dear Chet:
Enclosed are responses to questions on the ABWR probabilistic flooding analysts i
discussed with Glenn Kelly on February 22,1993 and documented in an NRCletter dated February 25,1993.
Please provide a copy of this transmittal to Glenn Kelly.
Sincerely, u
Ja Fox Advanced Reactor Programs cc: Jack Duncan (GE)
Norman Fletcher (DOE)
Art McSherry (GE) 5 {\\
nmr47 9303100254 930304
~
.a PDR ADOCK 05200001 A
i i
2 i
j Interoffice Memo To:
J.N. Fox Art McSherry Mk_$d*
From:
i Date:
March 3,1993
Subject:
Responses to NRC Questions on ABWR Flooding CC: J.D. Duncan, S. Visweswaran l'
l The following questions on the ABWR Probabilistic Flooding analysis were discussed with the NRC (Mr. Glenn Kelly) on Febmary 22,1993. The requested information was documented by an NRC letter dated Febmary 25,1993. The following is a discussion of i
the questions from the February 25,1993 letter.
l l
Question 1 - Explain the difference in reality and in the assumptions in the ABWR PRA i
between a fire door and a watertight door.
l l
l Response 1 - From a flooding perspective, the main difference between fire doors and l
watertight doors is that fire doors may inhibit or stop flood progression but watertight doors always do (if closed). Therefore, when analyzing fire door impacts on flooding, the j
worst case is assumed (i.e., ifleaking results in water being diverted from safe shutdown l
equipment, the door is assumed to not leak; if no leaking results in protecting safe i
shutdown equipment from flooding, it is assumed to leak). A good example of this is flooding in the reactor building corridor floor B3F. If the fire doors in the corridor were to leak, the entire corridor volume would be available to contain potential flood waters.
In this case, the doors were assumed to not leak which minimized the available room volume and increased the flood height in the room.
Question 2 - When a watertight door is closed but not dogged, will it alarm in the control room?
Response 2 - The current level of detail in the ABWR reactor building design does not address the specifics of door alarms. Operating plant designs do not indicate if watertight doors are dogged or not, only that the door is closed. Administrative procedures require that a once per shift walkdown of the plant be completed and any undogged watertight I
door will be considered a reportable event. Experience at operating plants has found this to be adequate to ensure that watertight doors remain dogged when required.
l Responses to NRC Questions on ABWR Flooding 03/03/93 Page 2 l
The SSAR will be changed to reflect this requirement.
Question 3 - Explain how CCF of operators leaving all three ECCS watertight doors open was considered. Explain how operator error is considered in flooding analysis.
Response 3 - The study assumed that a watertight door to an ECCS room could fail (leak or be open) but the probability of all three watertight doors being open at the same time was negligible. Procedures will exist to instruct operators to ensure that at least one watertight door is closed at all times and if flooding were to occur in a room with a closed watertight door that the other two watertight doors would be checked closed before l
opening the door to the flooding room.
t l
l This is discussed in Section 19R.4.2.4 of the ABWR SSAR.
Question 4 - Provide a better description of the worst flood inside of containment as estimated by the flooding PRA.
Response 4 - Section 19R.5.5 describes the reactor building flooding scenario. Figure 19R.5-6 shows the reactor building flooding in secondary containment event tree. The worst case scenario is sequence number 7 which has a CDF of 2.0E-10 per reactor year.
For this sequence, a break in a fire water standpipe or line from the CST results in the sump level switches detecting the flood (DET) but the operator does not respond and manually isolate the flood (OPACT). The sump pump capacity is exceeded (SUMPP) and the overfill lines to the corridor are clogged (OFLC). This results in flooding of all three electrical rooms on floor BlF and loss of all AC powered makeup systems. Core damage results when the operator fails to depressurize and use the AC independent water addition system to inject into the RPV. The operator failure rate for this sequence is.1 due to the j
short time available (<30 minutes) to isolate the flood before damage to the electrical equipment.
The SSAR will be changed to include the above scenario description.
Question 5 - Provide explanation of flooding ET top events.
Response 5 - The ET headings are explained below.
[
Responses to NRC Questions on ABWR Flooding 03/03/93 Page 3 ET HEADING EXPLANATION 1
Turbine Building Flooding CWS Pipe Break in Break in CWS causes (Low UHS)
Turbine Building flooding in the turbine building Level Switches Detect Level switches in the Flood condenser bay detect water from the flood.
All three pumps trip (pump CWS pump breakers open breakers open) terminating the flood since the LHS is low.
All three MO Valves close All three valves in the CWS must close because the break could be in the header that all three pumps feed.
Truck entrance door opens. The tmck entrance door at grade levelis not watertight and it is expected that it 1
would leak due to the presence of the water.
NC Watertight CB Excess The door from the turbine Dr Closed.
building to the service building access tunnel (entrance to the control building)is closed to prevent control building floodmg.
l i
Responses to NRC Questions on ABWR Flooding 03/03/93 Page 4 i
t i
Reactor brought to a safe The reactoris safely shutdown.
shutdown using equipment not damaged by the flood.
This heading is essentially l
the same as a turbine trip without bypass event with the additionalloss of l
condensate. See the answer to question 6 for l
additional information.
i Turbine Building Flooding Headings are the same as (High UHS) the previous ET except that no credit is taken for pump trip due to the high UHS l
l Control Building Flooding RSW Pipe Break in Control A break in the RSW system Building piping inside the l
RSW/RCW room of the i
control building upstream l
2 of the isolation valve.
i i'
Lower level sensors at The 2/4 waterlevel sensors
.15m detect RSW room in the room detect the flood Flooding water and send a signal to the control room to alert the operator.
l I
Operator acts toisolate Operator receives the alarm flooding.
and either trips the affected RSW pump or closes appropriate valves in the loop.
t I
t l
- =,..
s
Responses to NRC Questions on ABWR Flooding 03/03/93 Page3 l
Upper level sensors at.8m The next higher level detect RSW/RCW room sensors detect the flood and 1
flooding send a signal to the control room and automatically trip i
the RSW pump and close isolation valves in the affected loop.
Level sensors in the next If the sensors in the first l
two rooms detect room fail or the operator l
RSW/RCW flooding does not respond, the first RSW room will overflow l
into the other two RSW rooms and the sensors in those rooms will detect the flood.
Operator acts to isolate Based on alarms from the flooding.
sensors in the other rooms, the operator can determine that flooding exists and isolate the flood.
Automatic flooding RSW pump trip or MOV isolation.
closure will terminate the flood. See response to question 8 below for additionalinformation.
Reactor brought to a safe Using features not damaged shutdown condition.
by the flood, the plant is successfully shutdown. See response to question 6 for additional details.
Reactor building flooding Pipe break in ECCS room.
Double ended shear ofline in ECCS room.
from CST or suppression pool upstream from the isolation valve.
(
Responses to NRC Quesdons on ABWR Flooding 03/03/93 Page 6 l
Sump level switches detect Sump water level alarms in flood.
the control room.
Water in the corridor.
Watertight doors fail to contain flood waterin the i
ECCS room and water flows into the corridor.
Water in next division Watertight door in other ECCS room.
division fails enabling water to enter from the corridor into the ECCS room and results in loss of two ECCS l
divisions.
Reactor brought to a safe Using equipment not shutdown.
damaged by the flood, the reactor is manually shutdown. The conditional probability of core damage ist.27E-6 for loss of one division and 2.68E-5 for loss of two divisions. See response to question 6 for additional infonnation.
l Reactor building flooding Pipe break in the corridor.
Break in fire water in corridor.
standpipe or line from suppression pool or CST causes flooding in the corridor.
Level switches detect Sump level switches alarm
)
flood.
in the control room.
l l
)
{
i i
i i
I
i Responses to NRC Questions on ABWR Flooding 03/03/93 Page 7 Operator acts to stop Operator closes CST or flooding.
suppression poolisolation valves from the control room.
Water enters one ECCS Watertight door to ECCS room.
room fads resulting in loss of one ECCS division.
CCF of all three doors is considered negligible. See response to question 3.
Reactor brought to a safe Reactor shutdown using condition.
equipment not damaged by the flood. The conditional CDF is 1.0E-8 for all divisions available and 1.27E-6 for loss of one division. See response to question 6 for additional details.
Reactor building flooding Pipe break outside Break in fire water in secondary containment.
standpipe orline from CST results in flooding on B1F.
Flow switches detect Sump level switches alarm flooding.
in the control room.
i Operator isolates flooding.
Operator closes appropnate valves from the control room isolating the flood.
i Sump pumps capacity Sump pumps on BlF exceeded.
cannot keep up with the flood.
1
t Responses to NRC Questions on ABWR Floodingo3/03/93 Page 8 Overfilllines clogged.
Lines to B3F from sumps in BlF are clogged resulting in flooding of all three electrical rooms.
Reactor brought to a safe All makeup assumed lost condition.
except AC independent fire water. Operator action outside control room required to implement fire water addition.
Question 6 - Provide FT for final node that considers all other ways to keep core cooled.
Explain which systems GE takes credit for on a sequence by sequence basis.
Response 6 - Attached as Figures I and 2 are the event trees that were used to determine the final nodes. Table 19Q.5-3 lists the final node probabilities. Conditional events 1,2, and 4 from Table 19Q.5-3 are all from the turbine trip without bypass event tree which is shown as Figure 1. The fault trees for RSW and power were modified to model loss of one and two divisions. 1.lE-8 assumes all divisions are available following a turbine trip without bypass. 1.27E-6 is for two divisions available and 2.68E-5 is for one division available. Figure 2 shows the event tree for loss of all RSW. The conditional probability is 6.38E-4. The fifth conditional probability in Table 19Q.5-3 is the CDF from the ABWR full power PRA turbine trip with bypass event tree assuming the initiating event frequency is 1. The last conditional probability is for loss of all three divisions of RSW and loss of the PCS. The.1 conditional probability is for the operator failure to implement the AC independent water addition system which would be the only makeup source available for this scenario.
Question 7 - Discuss credit taken for turbine building flood being mitigated by truck entrance door. Justify assuming 95% of time it will mitigate flood when curb at watertight door to support (service) building is only about four inches high.
Respnses to NRC Questions on ABWR Flexting03/03/93 Page 9 Response 7 - The 0.05 unavailability is based on engineering judgment and the fact that the truck door is of a standard design for truck access (i.e., not a fire or watertight door) and will not present significant resistance to flood waters.
Question 8 - In Figure 19R.5-3 what is the basis for the (8)E-7 unavailability of fiooding isolation.
Response 8 - The conditional probability of flooding isolation (8.0E-7) is the probability that the RSW isolation valves and pump trip logic will not function given that the water level sensors do actuate. The component unavailabilities are listed in Table 19R.5-2.
Valve isolation is 4.0E-3. Pump trip is 1.0E-3. The combined failure probability of pump trip and valve closure takes into account the fact that only one of the two valves needs to close (see Figure 19R.4-2) to terminate the flood. Common cause failure of.1 for the valves to close are also considered. RSW pump trip or antisiphon valve failure is 1.0E-3 +
l.0E-3 = 2.0E-3. MOVs failure to close is 4.0E-3 per MOV and using a CCF of.1 results in failure of two MOVs being 4.0E-4. Therefore, the total failure probability is 2.0E-3
- i 4.0E-4 = 8.0E-7.
j i
a-s-e I
1 e
l*
l l
l I
b I
i I
f l
i a a e a : a a a : a a : a a = 2 : a : s : a : a : 2 :
J c
t
.:t
- i!
g i
I,s 3
t i
i
- ~3 T
3 I
T.'
I*I i
5
- I i
~
1 1
7 1
B
+!r.
o.
2 Jg 9
- 1*
(
l I
IIoj I
g-
-11 y
i er I
i g) 8 i
s.
l a
I 28 i
l i
eil
- s i
1 a.
a 8
et i
N e,
i Ih i;:i i
- 1' a
O gs s
O ge r
s It i
$s
= iif:
0 t' w
a
.I
i!9 i*s l
i
- i. -
f r,e o
m I
S CD 7
7 CTI
(
l O
O O
O e
r I
- D 4
8 i
t O
O W
W W
W m
e C
O G
G CD 1
a 7
7 O
M O
7 7
W G
I m
y x
x 2
x O
4 m
m O
O O
w O
O w
m p
4 O
J O
u J
l M
3 0
"WC OWC U
"4>O O
v y
JdCD u
O O
W W
i w4
>4gg C
W>
W>
I O
OO k
Wg s
J U
OU z
gg W
-W m
w ejg 1
e m
C l
W w
s 4
4 e
- wyd I
wEl>
[
w e
e o
gu
?
?
C Je 3
a m4 W
W W
J
+
gy y
C,, C O
O g
w 4
OC OC O
y i
Z E2 Sk O
J
'I J
O C
+ W w
HW p
Wu4 CW3 O
g D7W t
g JZC W
O
-sw Om 4
L O>
k i
j
(
- M4 yIQ a
w O,
C HW g
3ug m
W-00 4,70 og Q
L:D
-O h
u D
W I
O Y
i
->m J
WJ3 x
mm C4 OW g
DD i C W
JZ Wa 1
m4 CO E
41 NZ W
g
.4 l
O m;g 1
O[C W
l W 4 O
N C 3 O
i.
D C 4
l J W W
- W O
4 L
- O kN inD J'
W
<W C G
U k"W O
O J
OW O > *4 g
+.g
,r O"3 3
'e.1O W
m OJ OO C0 i
O
. g r
l d
- D
\\
1 r
1 l
-.. ~
..ii