ML20196D960
| ML20196D960 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 12/08/1988 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20196D958 | List: |
| References | |
| NUDOCS 8812090240 | |
| Download: ML20196D960 (15) | |
Text
{{#Wiki_filter:., .
+ 'o UNITED STATES
~,,
.8 o NUCLEAR REGULATORY COMMISSION O E WASHING TON, D. C. 20656 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO AUXILIARY FEE 0 WATER SYSTEM RELIABILITY STUDY TOLEDO EDISON COMPANY AND THE CLEVELAND ELECTRIC ILLUMINATING COMPANY DAVIS-BESSE NUCLEAR POWER STATION, UNIT NO. 1 DOCKET NO. 50-346
- 1. 0 INTRODUCTION In response to the June 9, 1985, Loss of Main and Auxiliary Feedwater Event at Davis-Besse, Toledo Edison Company has implemented various system modifications to improve the A@iliary Feedwater System (AFWS) reliability.
A reliability analysis submitted by Toledo Edison Company of the Davis-Besse AFWS provides estimates of the overall unavailability of the system as it existed up to June 9, 1985, as subsequently modified (modified two-train configuration), and as currently planned (three-train configuration). Addi-tionally, a qualitative assessment of the currently planned modifications was provided. A reliability analysis was performed previously for the Davis-Besse AFWS. That analysis relied on assumptions and guidance provided in NUREG-0611. The present reliability study uses a plant-specific approach, and its modeling is more detailed than that used in the NUREG-0611-based analy is. Specifically, this study assesses the AFWS reliability with an intact as well as a faulted steam generator. This involves modeling faults of the safety valves, atmospheric vent valves, main steam isolation valves, and turbine bypass valves. This study also uses a detailed model of the Steam Feedwater Rupture Control System which considers sensors / input buffers, logic boards, relay drivers and output relays. In addition, estimates of component unavailabilities are derived from plant-specific data to the extent possible, and additional human errors were considered. None of these considerations were part of the NUREG-0611-based analysis. Therefore, the numerical results of this study are not comparable directly to those obtained in the previous analysis. Also, the acceptance criteria that applied to NUREG-0611-type analysis are not applicable directly to this study. However, the methodology used in this analysis and the unavail-ability estimates obtained provide a useful tool for evaluating the AFWS modifications. 8812090240831208 PDR ADOCK 05000346 P O PDC
e e 2.0 EVALUATION 2.1 AFW System Description Since the Loss of Main Feeawater event et Davis-Besse or. June 9, 1985, the AFWS has been modified and improved. Tne AFWS reliability analysis considers three different system configurations, as described below. The reliability analysis also considers the Steam Feedwater Rupture Control System (SFRCS). The analyzed configurations are:
- 1) Existing configuration (as of June 9, 1985),
- 2) Improved Two-Pump Configuration, and
- 3) Planned Three-Pump Configuration.
2.1.1 Existing Configuration This configuration represents the Davis-Besse AFWS as it existed on June 9, 1985. A simplified diagram is shown in Figure 1. The AFWS consisted of two turbine-driven pumps with three suction sources, (a) two Condensate Storage Tanks (CSTs), (b) the Service Water System (SWS), and (c) the Fire Protection System. The AFWS is normally aligned to take suction from the two CST's (250,000 gallons each). The CST's are not seismically qualified. However, on a drop in suction pressure (2 2 psig) the AFWS pump suction is automatically transferred to the SWS which is seismically qualified. The suction transfer is established by opening SWS suction valve SW1382 (SW1383 for pump 2) and then closing FW786 (FW790 for pump 2). However, since the suction transfer pressure sensors are located upstream of FW786 and FW790 and subject to the CST's static head, an inadvertent closure of either of these valves will not be sensed by these pressure sensors. In that case suction transfer would not occur, thus disabling the corresponding AFWS train. Motive power for the AFW pumps is provided by steam lines to the AFWS turbines. The steam lines provide steam from either steam generator (SG) to each turbine through a set of motor operated valves, manual valves, and a governor valve. Normally, steam from SG-1 is provided to AFWP turbine-1; SG-1 steam also will be provided to A N turbine-2 if SG-2 has generated a low pressure signal ($ 612 psia). Pressure sensors upstream of each AFW pump (see Fig. 1) will detect a pressure drop (2 1 psig) and interrupt steam flow to the corresponding AFWP turbine. This can be achieved by automatic closure of the steam admission valves MS106/MS106A (or MS107/MS107A). The Steam Feedwater Rupture Control System (SFRCS) automatically initiates AFWS under the following plant conditions:
- d. Steam Generator low Level
- b. Steam Generator High Level
- c. High steam generator to main feedwater differential pressure
- d. Low steam generator pressure.
3 After AFWS actuation, the SFRCS will control the Steam Generator level at 46 inches on the startup range. The SFRCS will isolate either or both steam generators if low pressure signals (5 612 psig) are detected. 2.1.2 Improved Two-Pump Configuration This configuration incorporates short-term modifications that were implemented during the June 9, 1985 incident outage. These modifications did not change the overall AFWS operation appreciably. However, they did provide several improvements. Some of these are: Isolation of both SGs due to low pressure condition > is precluded. Also, SFRCS signals to close valves AF599 and AF608 have been eliminated. This improves the availability of AFW as a heat sink and redutos the likelihood of inadvertent isolation. Motor Operated Valves (MOV's) FW786 and FW790 are electrically disconnected and locked in the open position to reduce inadvertent suction isolation. A time delay of 10 seconds has been added to suction transfer pressure sensors PSL4928A/B and PSL4929A/B to reduce inadvertent suction transfers due to momentary dips in suction prer,ure. The time delay for pressure sen . 9L4;'?A/B and PSL4931A/8 is extended from 2.5 to 60 seconds. This wi' .
'u - inadvertant isolation of steam flow to the AFWP turbines.
The PG-PL-type governor for the AFWP turb. e-1 is replaced by a PGG-type governor similar to that of turbine-2. The PG-PL governor exhibited more than 15% of the total industry reported failures (Licensee Event Reports 1972 through 1985, Turbine Overspeed Trips, by AE00 dated September 2, 1986). The PGG-type governor exhibited less than 1% of the failures. This modification improves the AFWP-2 reliability. The locked open manual steam admission valves MS729 and MS730 are replaced by normally-closed, fail-open, air-operated valves MS5889A/B. MOV's MS106A and M5107A have been modified to be normally open and will close only if SG-2 or SG-1, respectively, sends a low pressure (612 psig) signal. The normally-open oosition of MOV's MS106A and MS107A and the normally-closed, air-operated valves (A0V's) MS5889 A/8 are anticipated to result in improved steam admission and thus more reliable AFW turbines. This is because the A0V's are believed to be more reliable than the MOV's. However, the length of piping between the motor-operated valves and the air-operated valves is about 350 feet *. To avoid condensation of the steam trapped in the piping during standby conditions, the Licensee has a This information was provided by Toledo Edison Company in a telephone conference call between S. Diab and A. DeAgazio of the NRC and Sushil Jain, et al. of Toledo Edison Co., August 8, 1988.
installed a moisture trap upstream of each of the air-operated valves MS5889A/B. The moisture traps are sized adequately for the removal of the condensate. The combined effects of the pipe lengths, the ambient temperature around the pipe, and the moisture trap allow a small flow of steam from the SG's in the closed pipes to keep them warm. A warm steam admission pipe that is free of condensate should lead to reliable operation of the turbine-driven AFW pumps. 2.1.3 Three-Pump-Configuration The three-pump configuration is basically the same as the improved two-pump configuration except for the addition of the full flow motor-driven Feedwater Pump (HDFP). This configuration is shown in Figure 2. The MDFP is initiated manually from the control room by a momentary-contact-spring-return switch. The MDFP has two modes of operation. Mode B is for normal plant startup and shutdown and for power operation at low power levels up to about 40% power. Mode A is for power operation above 40% power. In mode B, the MDFP takes suction from a deaerator stora valves, discharges through a flow control valve (FW5867)ge and tank feedsvia thetwo manual SG's via high pressure feedwater heaters. In this mode, cooling water is supplied to MDFP lube oil and seal water coolers by the SWS. In mode A, the MDFP takes suction from the CST via a manual valve and discharges to both SG injection flow paths through the flow control valve FW5867. In this mode, the lube oil cooler is supplied from the first stage discharge of the pump, and the pump-seal coolers are not required. In all system configurations described above the SFRCS automatically initiates the AFWS under the following plant conditions:
- a. Low steam generator level.
- b. High steam generator level.
- c. Low steam generator pressure,
- d. High steam to feedwater differential pressure,
- e. Loss of four reactor coolant pumps.
The SFRCS also isolates either steam generator upon receipt of a low pressure signal (612 psig) from that generator. This function ensures isolation of any ruptured steam or feedwater line or an inadvertently depressurized steam generator. The SFRCS has been modified to avoid isolation of both steam generators. The SFRCS consists of two actuation channels. Each actuation
5-channel consists of two logic channels. An actuation channel is tripped if both of its logic channels are tripped. Each actuation channel initiates operation of its corresponding AFWS train, i.e., actuation channel-1 operates AFWS train-7 and channel-2 operates train-2. The main stkam system equipment considered in this study includes nine safety valves (MS H's), one atmospheric vent valve (AVV), one main steam isolation valve (MSIV), an( three turbine bypass valves (TBV's) for each of the two main steam lines. The MSSV's, AVV's, and TBV's open to relieve steam pressure following a plant trip which accompanies the loss of main feedwater event. If these valves fail to reseat properly, the steam generator p essure may drop below 612 psig, thus causing the SFRCS to isolate the affected SG and align both AFW pumps to feed the intact steam generator. After the AFWS reliability anlysis was completed, additional system improvements were scheduled for completion during the current refueling outage. These improvements include the following:
- 1. The AFW flow control from the two turbine-driven pumps will be achieved by the two M0V's AF360 and AF388. The AFW turbine controls will be modified to allow turbine operations only at a constant speed.
- 2. The positions for AFW discharge valves AF3870 and AF3872 will be changed to normally-open.
- 3. Two cavitating venturis will be installed, one in each SG injection line. This will limit the AFW flow to a faulted SG.
- 4. For ease of operation, all AFWS controls will be placed on one mimic panel in the control room.
- 5. Separate flow control valves will be installed to control automatically the MDFP flow to the two SG's.
- 6. Redundant AC power supplies will be provided to the MOFP lube oil i pump.
- 7. The capability will be added for the MDFP to take suction from the Seismic Category I SWS, 2.2 Mission Success The AFWS mission is considered successful if the system is actuated and continues to operate until the conditions that initiated the event are removed or the reactor has reached stable shutdown conditions.
i
4 2.2.1 Success Criterion The criterion adopted for the AFWS reliability analysis is the availability of sufficient AFW flow to at least one SG within 5 minutes following a loss of main feedwater (LMFW), a loss of offsite power (LOOP), or a loss of all AC (LOAC) event. Once-through SG's, the type used by B&W plants, have a small inventory of water which is exhausted within approximately 5 minutes after any af the above-mentioned events. Therefore, when estimating system reliability, credit is not taken for a recovery action if that action requires more than a simple straightforward manipulation by an operator. We believe that the 5-minute criterion is appropriate. 2.2.2 AFWS Mission The AFWS mission time assumed for the three initiating events is as follows: Event Mission Time (hours) INT 10 LOOP 10 LOAC 2 The 10-hour mission time for the LMFW and LOOP events is consistant with the time required for the AFWS to complete normal plant shutdown. The 2-hour mission time for the LOAC event is selected because of the high likelihood of restoring emergency diesel generator power or offsite power sources within 2 hours of the event. 2.3 Modeling Features Assumptions and Failure Data The following is a discussion of the major modeling features and assumptions used in the analysis:
- 1. The system analyzed covers the AFWS, the MDFP and its lube oil system, and the suction and discharge piping valves. The system boundaries also include suction from the CST's and the SWS. With respect to the main steam system, the MSSV's, AVV's, MSIV's, and TBV's are considered. Also considered are the SFRCS interfaces with the AFWS and main steam system, and the Integrated Control System and its interfaces with the AVV's and TBV's. The steam supply to the AFW turbines and the AC/DV power supplies to the various pumps and valves are modeled.
- 2. Isolation of both SG's due to incorrect anticipatory manual SFRCS actuation is modeled in the existirg AFWS configuration but not modeled in the other configurations.
- 3. In reference to automatic SG 1evel control systems malfunction, no credit is given for operator recovery actions because of the limited time available before SG dryout.
- 4. The Feed and Bleed mode of decay heat removal is not modeled in this analysis.
- 5. The analysis models the failure of the steam admission air-operated valves in terms of quick opening, thus leading to an overspeed trip of the associated AFW pump turbinc.
- 6. In the existing configuration, the PG-PL type governor is assumed to fail if room cooling is lost for more than 2 hours (e.g., in a LOOP event). In other system configurations, the PG-PL governor is replaced by a PGG-type. The latter does not depend on room cooling and, hence, is more reliable.
- 7. The unavailability data associated with the failure events used in this study has been derived from Davis-Besse plant-specific failure rate data. This data was generated on the basis of a review of plant operational history and maintenance, test and surveillance records. The criteria for selection of unavailabilities considered the numbers of equipment challenges, an assessment of the specific type of failures identified in the plant records, and a comparison with generic sources for similar failures. When identifying plant-specific failures, AFWS components as well as similar components of other plant systems were considered. Where plant-specific data were unavailable or considered insufficient, generic data were used.
- 8. Common mode hardware failures are not modeled in this study. For example, common mode failure of the air supply to the two air-operated valves MS5889 A/B is not modeled. However, Toledo.
Edison Company states that these two valves are designed to fail-open upon loss of air or loss of power to the solenoids that actuate the air diaphragms.
- 9. All systems and components are assumed to have regular preventive maintenance and testing activities which make such equipment unavailable. Consistent with plant operation practice, only one AFWS train (and its affiliated components) is permitted under maintenance at any time during plant operation.
- 10. Wheelever manual recovery action is involved complete dependence (coupling) of human error is assumed (i.e., if the operater fails'to actuate one train, failure to actuate the other trains is assumed as well.)
- 11. Dependent human errors are considered where applicable (e.g.,
, misalignment of more than one valve, miscalibration of more than one
- pressure switch, etc.)
- 12. The human errors considered in this study are those committed prior to as well as in response to the accident. Pre-accident human errors include valve misalignment or instrument miscalibrations.
Human errors committed in response to an acc' dent were considered
l when appropriate. These errors include those of omission (e.g., failure to manually initiate the AFWS when necessary) and those of commission (e.g., improper selection of SFRCS manual switches). Consideration was given as to the degree of stress imposed on the operator in selecting the appropriate human error probability. All human error probabilities were obtained from generic sources.
- 13. In the existing AFWS configuration, the modeling of human errors includes the incorrect anticipatory manual actuation of the low steam generator pressure SFRCS switches resulting in isolation of both SG's. This is not modeled in other configurations because of the system modification described in Section A.3 above.
- 14. The failure of a "non-dedicated" operator to initiate manually the MDFP wit,in 5 minutes of the initiating event is assigned a probability of 0.25 (Handbook of Human Reliabiliaty - NUREG/CR-1278).
We believe that the 5-minute criterion and its associated human error probability are appropriate.
- 15. For the LMFW event, offsite power is assumed to be available with a reliability of 1.0. For the LOOP event, the failure.of AC power supplies is represented by the failure probabilities of the emergency diesel generators. For the LOAC event, only the DC power supplies are assumed to be available with a reliability of 1.0.
This assumption is not realistic. 2.4 Results The AFWS reliability analysis provides estimates of overall system unavailability for each system configuration. A total of nine cases were analyzed. The unavailat;ility values are shown in Table 1. The analysis also produced the minimal cut-sets (MCS) and provided estimates of their unavailability values. The MCS are defined in terms of failure events given in the fault tree and the independent sub-trees (IST's). The IST's were generated by the SETS computer code and represent a sub-tree structure which is comprised of failure events that exist as inputs to that sub-tree only. The IST's may be considered as lumped events. It should be noted that the three-train unavailability estimates provided by this analysis are higher than what would be expected from a configuration of this type. Since previous Davis-Besse analyses of the AFWS unavailabilities ' and equipment failure data were done using different system modeling, assumptions, human error modeling and equipment failure data, it is difficult to make a direct comparison in order to explain the results. It is possible that the modeling level of detail and, to a lesser extent, the use of plant specific failure rate data, may have contributed to the estimated high unavailability. For example, if conservatively high values of failure data were used consistently at the most detailed level of system modeling, then the cumulative effect could produce an unrealistically high unavailability. l However, this does not have a significant effect on the validity of the analysis, since it is used primarily to assess the impact of system changes.
e 9 We concipde that the analysis methodology, assumptions, and system modeling provide a useful analytical tool to tvaluate the AFWS reliability and the impact of various system hardware or operational modifications. 2.5. Impact of Modifications on System Reliability 2.5.1 Pre-Study Modifications These are the system modifications as implemented or planned before the AFWS reliability study. These modifications were modeled in the analysis. The criterion for system failure is the failure of all pumps to deliver flow to at least one SG. The following is an assessment of the significant system modifications. SFPCS Logic Change and Control Room Panel Modification These system modifications virtually eliminate total system unavailability due to low pressure in both SG's or operator error. Therefore, these modifications should have a positive impact on system reliability. Replacement of the PG-PL governor on AFW puma-1 turbine with a PGG governor ( AFW pump-2 turbine is equipped wit 1 a PGG governor) Due to the superior performance of the PGG governor over that of PG-PL governor (See Section II.A.2), the above system modification also should have a positive impact on system reliability. Addition of the MDFP The addition of this full capacity pump and the emphasis placed on its operability significantly enhances the total system reliability. Replacement of the manual steam admission valves with "normally closed","fail-open", "air-operated" valves (A0V's) M55889 A/B, and changing M0V's - MS106A and M5107A to be "normally-open" The effect of the above system modifications consists of two parts: a) On system actuation, the opening of an A0V instead of an MOV (e.g., MS5889A instead of MOV MS106 or M5106A) will start a turbine-driven pump. Toledo Edison Company states that in its experience the A0V's are more , reliable than the M0V's. Therefore, this should improve the system's ' reliability, b) The crossover steam supply piping between valves MS106, MS106A and MS5889A (also, between valves MS107, MS107A and MS5889B) stretches for , about 350 feet *, Although there is a moisture trap upstream of each of valves MS-5889A/8, moisture traps are known to fail frequently and cause i water accumulation. Water accumulation, in turn, may cause AFW pump , trip on overspeed. This effect tends to reduce the system's reliability. t
- Conference call between S. Diab and A. DeAgazio of NRC and S. Jain et al. of Toledo Edison Co., August 8,1988.
ihe net impact on the system's reliability is the sum of the above two effects, namely the potential increase in overspeed trip vulnerability and the improved A0V performance over that of the MOV. Toledo Edison Company should ensure that the above system modifications do not result in a net reduction in system reliability. This may be achieved by closely monitoring the operability of the moisture traps, and the capability of the AFW turbine-driven pumps to start and continue to operate. 2.5.2 Post-Study Hodifications These system modifications are planned
- and have not been modeled in the analysis. They are described in Section II.A. Only one of these system modifications is a direct result of the reliability analysis. The other modifications came as a result of recomendations from other studies. The following is an assessment of the significant system modifications:
- 1. The AFW turbines will be modified to o)erate at a constant speed.
AFW flow control will be achieved by t1e two MOV's AF360 and AF388. The standby position of these two valves is normally-open. This modification should result in a more reliable flow control.
- 2. Discharge valves AF3870 and AF33872 will be modified to be normally-open. This modification should improve the system reliability.
Toledo Edison Company stated that this is the only system improvement that resulted directly from the AFWS reliability study insights. By keeping these two valves (AF3870 and AF3872) normally-open, the potential for steam binding of the AFW pumps may increase. However, to avoid steam binding of both the AFW pumps and the MDFP (due to hot water back-leakage and flashing) Toledo Edison Company performs the ollowing: Checks the thermocouple readings every 12 hours (from the control room) at the locked-open AFW injection valves AF599 and AF608. These thermocouples are located inside the containment and will give early warning of any hot water back-leakage. Checks the temperatures of the AFW piping and pump casings by touching them every 12 hours. Checks the thermocouple reading (from the control room) at the NDFP casing every 12 hours.
- Toledo Edison Company stated (Sushil Jain, et al., Presentation to the staff, June 9,1988) that these system modifications will be implemented in the current refueling outage.
Every refueling outage conducts a backflow test for the AFWS check valves. I i With the above precautions this system modification should improve system reliability significantly.
- 3. A cavitating venturi will be installed in each SG injection line.
This is intended to limit the AFW flow that would be wasted in case of a break in SG or attached piping. This will conserve the AFW for the intact SG and should, therefore, enhance system usefulness.
- 4. All AFWS controls will be on one mimic panel in the control room.
This should simplify the operator actions and, therefore, should enhance the likelihood of recovery in case of an AFWS malfunction.
- 5. Separate flow control valves will be installed downstream of the M0FP to control automatically SG 1evels. This may improve the reliability of the MDFP and the associated flow control.
- 6. Redundant AC power supplies will be provided to the MDFP lube oil pump. This should improve reliability of the MDFP since it can be completely powered from either Diesel Generator.
- 7. The MDFP suction will be modified to take alternate suction from the Seismic Category I SWS. Thus, the MDFP suction will not be limited to the CST's. This should improve the availability of the MOFP.
The result of the above post-study modifications is the enhancement of the AFWS reliability over that estimated by Toledo Edison Company.
3.0 CONCLUSION
S The staff has reviewed the Davis-Besse AFWS reliability analysis and has concluded: 3.1. The analysis methodology, assumptions, and system modeling form an acceptable analytical tool to evaluate the AFWS reliability and the impact of various system modifications. 3.2. The human error rate data are used as appropriate with proper emphasis on the stress levels associated with the initiating events. Operator errors prior to the event as well as recovery errors during the event have been factored into the analysis. The human error rates used in the analysis were obtained from generic sources. l 3. 3. Component failure rate data were obtained from the Davis-Besse plant t experience to the extent possible. These data were generated on the basis of plant operational history, and maintenance, test and
surveillance records. The criteria for selection of a failure rate data considered the number of equipment challenges, and a comparison with generic data sources for similar failures. Although we inspected the plant-specific failure rate data used in the analysis, we have not verified the accuracy or completoness of plant records or its operational history. However, small variations in plant-specific failure rate data are not expected to change the estimated AFWS unavailabilities significantly, and should not change the conclusions of this report. 3.4. Thr changes in valve positions and valve operators in the steam adnission lines to the AFS turbines (See Section II.E.a, above) introduce two effects: (a) a potential increase in overspeed trip vulnirability due to any accumulated condensate, and (b) an improved i A0V's cerformance over that of MOV's. Toledo Edison Company should ensure *. hat these system modifications do not result in a net reduction in system reliability. This may be achieved by losely monitoring the operability of the moisture traps, and the capability of the AFW turbine-driven pumps to start and continue to operate. Principal Contributor: S. Diab Date: I l i
sp Fipic L ;ggg Davis Besse Auxiliary and Motor Driven Feedwater Systems l Pre Modification i 8 .-. g,- B . , _ . g.. . . , ,c a.,mc o, me,mce, I I I I
,m O O
- K)e.t...
meim.OK
- q. ..,g
.,,2 g _ z.. ... - >
.. .. ,z g g_o.....
5..., 8 = ,- r, = 3 tas ,&. W.,,, on .,,,, '
- O b,
- Q q .--4,4,. i
- r:
u a,i M &.6, u,3 M
.... .w. u,,
.,,,7-g-e
.. ,- ,,8,.. g e n c.. .........
............>2.,
Ise i the $
= =.., N..
g__,.,i
-4,4, ,. .aMD As. n#
'" ^
g" g a
-+== >o-og ,~ h o ==
=] ,,,,v y ** \.,
= .= ,,,.a ." ~,,i iX
- g. ,. , ~
,., 'e @, ~'-Z Q_
~?-
x,,,,,.
%~'= , ::::
4 2"'
~ - >
< :: w I
O F icpa 2- SEP 2 31988 e Davis-Besse Auxiliary and Motor Driven Feedwater Systems Post Modification High Level /6 Pressure 3 . . . , - v.~. 9 T
,8 - , ,i.
&. . .%. , .. . ,. s c, .u ...
DC(1) DC(21 a.. .c., .. {];];{ . , A i r M, S te. c.si . . ., eO ^e V V. c.i.i. ..,
~ .i . -
AC it) . AC() ei .
. .. ...., . n.
,, DC (11 AC (3)
.aci
. u ., .. ..
3' one. vire,./' .m eg 7, ,,
,,,,,,,,,.e ,
4.,) 4 ,$ 1,,4 5,35
..,.,,,, g ) R G..,,,,
'C'83
_l/ : ,, m _q '"' l 'a'Un'i
.gr., 7 W*b,'sy,*,*:,w,,
1'W ast,s
;; ;4 $$!f %
g_a.is : os '8 28g. 9 mr,',rg-{),ao "o cig=.',',':
,4 H,, (.. .......... ,',' a',','
O
. .;:: . . . n . . . . . . . .,
s.,,,, 1s,s AC it) L' *
. nE ,
c ....e, *'Uk
='" ~
'= a X X 'T
'Mi " ge, ,,, c" **'"
c71 **"'f, 5 o.e
~
' ' "'[., s.4.e,e save e ,.
are ....%
. . . . r e, e.
,.u..I 4'an
.y g).g g ,;...".
, n w.
3< . .. n i u.. s x...... g ,.... w)p}ch
...... T.{g * '
x
'a**
oc m "
' Levi ,mg3ss ,*.=
g) W*933' q _ .19 9 9s - 4 i, .' ',. C . ..
..'*.- a
SEP 2 a 1933 TABLE 1 UNAVAILABILITY OF THE DAVIS-BESSE 1 AUXILIARY FEEDKATER SYSTEH (AFHS) Unavailability Value Initiating of the AFHS Event
.t ~ ; . . u i Ex Co M,isting L o Pump ha.ba 9uoi .cn Con $praWr. ConR prakte f
( Loss of Hain Feedwater .062 .0068 .003 (LNFW) Loss of Offsito Power .071 .015 .0092 (LOOP) Loss of All AC .35 .27 .27 (LOAC)}}