ML20033E846
| ML20033E846 | |
| Person / Time | |
|---|---|
| Site: | Zion File:ZionSolutions icon.png |
| Issue date: | 03/02/1990 |
| From: | Chranowski R COMMONWEALTH EDISON CO. |
| To: | Davis A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| References | |
| NUDOCS 9003140371 | |
| Download: ML20033E846 (29) | |
Text
o 1,7 Sc0 7
i Commonwealth Edloon o
h YNNisan
+
i o
phnois 60099 onophone 708 / 746 2084 March 2, 1990 l
l i
l Mr. A. Bert Davis Regional Administrator j
U. S. Nuclear Regulator Commission Region III_
i 799 Roosevelt Road
-l Glen Ellyn, IL 60137 i
Subject:
Zion Station Unit 1 l
License Number PPR-48 And DPR-39 i
6 Request for Temporary Walver of Compliance Oletel Generator Operability Roqdirements BRLDocket im_52-m.aa:L10.304 i
a The purpose of this letter is to document the basis for the Halver of j
Compl hnte issued verbally on March 2, 1990 for Zion Station.
(
1.
Roson for the Reouest:
Zion Station is requesting a Temporary W1ver of Compliance to the 1
Confirmatori Order of February 29, 1980 in order to avoid putting the two
{
Zion Units through a thermal transient (i.e. a cooldown of the primary coolant system from 547 degrees F to less than 200 degrees F).
The integrity of the. reactor vessel and other components of the primary system of a nuclear plant can be adversely affected by the number.of thermal transients that they are subjected to during their lifetime. As each l
additional thermal transient can affect this integrity, it is prudent to i
avoid such transients as long as the health and safety of the public is i
preserved.
l t
Zion Station is requesting relief from the February 29, 1980 Confirmatory Order for Zion Station, Appendix A Item B.6 which stipulates allowable outage times dependent on the number of Diesel Generator 1
failures in the prior 100 tests.
Specifically,-Zion Station is requesting l
L that the allowed time to remain in Hot Shutdown for both Zion Units be I
extended 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br />, to allow for continued troubleshooting and repair of l'
the "0" Diesel Generator.
If the problem with the "0" Diesel Generator is not identified and corrected, and the diesel generator does not successfully pass its operational surveillance test by 1200 on March 6, 1990, a cooldown will immediately commente so as to place both Zion Units i
in Cold Shutdown by 0600 on March 7, 1990.
1 90031%h e
4(
nm P
l 2183o(1) d i
j 1
- a.
I y
2.
Circumstances Leading to the Request:
. On Thursday,. March 1, 1990, at 0400 the "0" Diesel Generator was
.taken out of service to replace certain components in the control system that had been suspected to be intermittently causing a premature shutdown of the Diesel Generator.
Based on the existing number of failures of the o
Zion Diesel Generators (5 in the last 100 starts for each unit) the Confirmatory Order allows 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> to repair the affected diesel and 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> to attain Hot Shutdown if the repairs are unsuccessful.
Upon performing the operability surveillance test after replacement of the suspected faulty parts the "0" Diesel Generator again tripped prematurely. As a result of this failure the two Zion Units, which had i
been near 1001. power at the beginning of the time clock, were brought to Hot Shutdown by 2400 on March 1st, within the required time of the Confirmatory Order. Upon achieving Hot Shutdown the Confirmatory Order then requires the Units to be cooled down to Cold Shutdown within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> (by 0600 on March 3, 1990) if the diesel generator continues to remain inoperable, Due to the intermittent nature of the problem, the identification and repair of the specific faulty component has_been extremely time consuming.
Zion Station, Commonwealth Edison Corporate and Cooper Bessemer experts have been working around the clock attempting to solve this problem.
The symptoms of this failure are not similar to recent u
problems experienced with the 21on Diesel Generators so this current problem is mt indicative of ineffective corrective action to previous failures..
addition, based on our knowledge of industry experience, there have been no vendor correspondence which might have pointed to similar problems for which we might have taken preventative actions.
3.
Compensatory measures:
During the period of time that this Halver of Compliance is in effect, Zion Station will take the following compensatory actions to ensure that both Unit I and 2 are maintained in a safe and stable r
condition:
1 i
a.
Normal temperature and pressure will be maintained for Hot Shutdown conditions, b.
Other.than required Technical Specification or Confirmatory Order i
L surveillance test requirements, additional testing or preventative l
maintenance will not be performed on the remaining operable Emergency L
Diesel Generators 1A, 18, 2A and 2B.
1 i
2183o(2) l
- l3 3.
(Continued) c.
Electrical distribution equipment, either normal, standby or reserve, will not be taken out-of-service on either unit.
d.
The Northern Division Load Dispatcher has been notified to not take out of service any incoming transmission lines to Zion Station except for emergency requirements, t
e.
To preclude unnecessary challenges being made to safety systems, other safety related equipment will not be taken out-of-service unless required by Technical Specifications or the Confirmatory Order, f.
All operating personnel will be notified via Standing Orders to ensure these actions are maintained.
4.
Preliminary evaluation of safety significance and potential consequences:
A.
Electrical System Design and Minimal Safety Requirements:
Zion Station's ESF electrical distribution system consists of three ESF buses per unit.
ESF buses 147, 148, and 149 (247, 248,'and 249) provide power.to three divisions of ESF equipment.
Attachment A to this letter provides a listing of the components supplied by these buses, as well as the minimum number components necessary-to meet licensing bases.
As can be seen in Attachment A a minimum of two of the three ESF divisions are necessary to meet the licensing bases.
The current Zion Station Technical Specifications require all required redundant components to be operable when a Diesel Generator is inoperable.
For the purpose of the extension granted, all required redundant components are operable.
A postulated loss of an ESF bus results in the inoperability of the components supplied by the bus. As long as the equipment supplied by the other two buses remains operable, sufficient components are available for safe-shutdown and decay heat removal following a loss of coolant accident coincident with a loss of offsite power.
- Thus, while this situation would represent a loss of redundancy, the performance requirements presumed in the UFSAR Chapter 14 analysis i
would remain available.
Each ESF bus is capable of being powered from several diverse sources.
These sources include; the System Auxiliary Transformer being supplied from the offsite distribution system, the Unit Auxillary Transformer powered from tne Main Generator (not available in current plant conditions) or offsite power.when in a backfeed configuration, reserve feed from the opposite Unit's 4 kv bus 141, (241), or the Diesel Generator associated with that ESF bus.
l As additional information regarding the reliability of Zion's electrical power system, recent diesel reliability data has been compiled.
The other four diesels have a high confidence factor for i
starting. The latest reliability figures showed a failure rate over the last-100 starts of 0.02 to 0.03.
f 2183o(3)
e 1
-jg :: s
'4; - (Continued)
B.
Estimation of Safety Significance:
The effect of the 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> extension on core melt frequency has been analyzed in Attachment E.
This analysis considered a number of
- cases to establish the potential effects.
The conclusion reached by that analysis is that the 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> time extension for the "0" diesel generator will not have any adverse impact on core melt frequency associated with LOCA sequences.
5.
Justify the duration of the request including the expected duration of
- i testing to be performed:
Zion Station feels an additional 90 hours0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> in Hot Shutdown is justified due to the nature of the failure and the extent of the troubleshooting and testing that may be necessary to positively find and correct the problem.
The troubleshooting process has been time consuming due to the intermittent nature of the problem.
For example, during the six actual starts on March I and 2 and approximately 20 simulated starts the failure was only observed twice, i
As a number of components could be the cause of the problem a systematic plan (see Attachment B) 1s necessary so that the root cause is positively identified.
This time consuming approach which considers one component at a time is needed to resolve the cause of the failure.
Once the systemmatic troubleshooting plan has identified the root cause, Zion will begin a testing program to conclusively demonstrate the required operability.
This testing program will be comprised of more diverse testing than normally required to demonstrate operability. The systemmatic troubleshooting plan and testing described above forms the basis for the additional time requested.
l p-L l
1 i_
2183o(4)
I
g 6.
Basis for' concluding that the change does not involve any significant hazards:
Based on Attachment C, Co:nmonwealth Edison has concluded that the probability and consequences of accidents previously evaluated are not significantly increased, the possibility of a new accident is not created, and the margin of safety is not significantly reduced.
He believe this request meets the criteria for no significant hazards consideration.
i 7.
Basis to show that this request does not involve an irreversible environmental consequence:
This request does not involve a change in the installation or use of the facilities or components located within the restricted areas as defined in 10CFR20. Commonwealth Edison has determined that this Temporary Halver of Compliance involves no significant increase in the amounts, and no significant change in the types, of any effluents that may be released offsite and that there is no significant increase in individual or cumulative occupational radiation exposure, Accordingly, this Temporary Halver of Compliance meets the eligibility criteria for categorical exclusion set forth in 10CFR Section 51.22(c)(9).
Pursuant to 10CFR51.22(b) no environmental impact statement or environmental assessment need be prepared in connection with the granting of the temporary waiver of compliance.
In conclusion, if the "0" Diesel Generator is successfully returned to operable status, the results of our investigation will be discussed with Region III personnel and concurrence received as to our conclusions prior to taking either unit to a mode above Hot Shutdown.
r f
l L
l 2183o(5)
r-
- g
~
l t
The course of action described in this request has been reviewed and
[
approved by the Zion Station On-Site Review Committee.
N If.you have any-further questions regarding-this matter, please contact
. his office.
t Very truly yours, NNb R. A.
- hrzanodi, Nucle r Licensing Adminstrator 4
t RAC/rmd p
cc: C.-Patel - NRR Senior Resident Inspector.
L 1
I l
4 I'
L i s l'-
2183o(6)
L
~
x,-
.'^..
e ATTACHMENT A~
'.1 4KV Bus Components 147/247 148/248 149/249 Containment Spray pumps A
B C*
Safety Injection Pumps A
B Centrifugal Charging Pumps B
A Auv111ary Feedwater Pumps B
C Residual Heat Removal Pumps B
A Component Cooling Pumps OE(147) 00(148) or OB(248)
OC(149) or OA(249)
Service Water Pumps A
B C
Containment Fan Coolers C
A, D B, E
- The "C" Containment Spray Pump is diesel driven. However, its operability requires operability of motor driven valve 1(2)MOV-CS0006, which is a normally closed containment isolation valve powered by bus 149(249).
l l
l l
t l
l i
i-l 2183o(7) a
ATTACHMENT A (Continued)
Number Available Number Required With No Power Components in Licensina Bases to Buses 247/147 2
Containment Spray Pumps
.1 (Iodine. removal) 2 (Pressure suppression)
Safety Injection Pumps i
1 Centrifugal Charging Pumps i
1-
)
Auxiliary Feedwater Pumps 1
3 (including steam driven pump)
Residual Heat Removal Pumps 1
2 Component Cooling Pumps 2*
4*
Service Water Pumps 2*
4*
Containment Fan Coolers 3
4
- Total for both units 2183o(8)
=.,,
-l..
l.
j
-y ATTACHMENT B ZION STATION "0" D/G TEST PLAN The following is Zion Station's remaining test plan to determine the probable root cause of the Unit "0" D/G random failures to start during the period of the Halver of Compliance from the Confirmatory Order issued to Zion Station on February 29, 1980.
1 The following items will be individually inspected and tested prior to any attempts to start the "0" D/G:
1)
Lube oil setpoint 2)
Turbo oil setpoint 3)
Vibration trip 4)
Check valve in first out panel 5)
. Filters in regulators After verification of these items, a' Zion Station Test Procedure (TSGP-38 D/G Control System Test) will be performed to individually test each pneumatic trip that has indication on the first out panel.
After successful completion of TSGP-38, simulated and actual starts will be performed.
During each of-i these starts the following parameters will be monitored:
1) 205 solenoid venting 2)-
4MX1 relay 3)
Lube oil pressures and time to setpoint 4)
First out trip switch and reset pressure gauge 5)
Over-speed governor-6)
Vibration switch 7)
Lube oil pressure / vibration trip block valve (its response will also be timed)
Data obtained from Zion Units 1 and 2 D/G starts will be compared to the parameters observed on the Unit 0 D/G and evaluated for any correlation.
Independent of any work being performed on the engine, bench tests will be conducted on the following items:
1)
Removed relay i
2)
Removed socket 3)-
Removed solenoid These parts were removed during previous troubleshooting activities due to their suspicious behavior.
After a root cause has been determined the following tests will be performed prior to declaring the "0" D/G operable:
1)
Minimum of 5 maintenance runs 2)
A start from each train of safeguards from the safeguards test panel 3)
A normal operability test (PT-11) 2183o(1)
n
.l.
l AITACHMENT C EVALUATION OF SIGNIFICANT HAZARDS _CONSIDERATIQH PROPOSED CHANGES TO ZION _IECHNICAL SPECIFICATION APPENDIX A - SECTION 3.15 AND 4.15 AUXILIARY ELECTRICAL POWER SYSTEM DESCRIPTION CF THE TEMPORARY HAIVER OF COMPLIANCE REQUEST A Temporary Halver of Compliance to the Zion Facility Operating License is proposed to allow a.96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> extension outage for DG "0" outage..
BACKGROUND 10 CFR 50.92 states that the Temporary Halver of Compliance will involve a no significant hazards consideration if the Temporary Halver of Compliance does not:
1.
Involve a significant increase in the probability or consequences of an accident previously evaluated; or 2.s Create the possibility of a new or different kind of accident from any accident previously evaluated; or 1
3.
Involve a significant reduction in a margin of safety.
1 The discussion below addresses each of these three criteria and demonstrates that the Temporary Halver of Compliance involves a no significant j
hazards consideration.
i
=ll l
I 2185o(18)
l t
.p.
s.
ATTACHMENT C (Continued)
Basis'for no significant hazards consideration:
The justification for no significant hazards consideration is based on the following factors:
1.
The high reliability of the Commonwealth Edison electric system.
'11.
The small effect of the proposed extension on predicted core melt frequency, as demonstrated by a probabilistic risk assessment (PRA) based analysis.
111. The surveillance frequency of the other four station diesel generators will ensure that all ESF components are operable with both normal and alternate AC power at all times during the O diesel generator time extension.
iv. Consideration in the Zion Emergency Operating Procedures (EOP's) of proper operator action in the unlikely event that only one of three 4160 volt ESF busses is available during an accident.
Attachment D summarizes the conclusion of the Zion Probabilistic Safety Study regarding Loss of Offsite Power (LOOP).
Due to the reliable design of the switchyard ring bus (featuring six incoming lines) and the historical stability of the Edison system, the probability of total loss of offsite power l
at Zion Station is considerably less than the industry average.
. Table 1 to this Attachment provides a list of ESF equipment powered from each ESF bus and Table 2 identifies the minimum number of components required to meet the accident analyses for each unit. Table 3 provides references identifying the bases for minimum component requirements.
E The Commonwealth Edison PRA group has analyzed the probability and L
consequences of a LOCA coincident with total loss of offsite power, for both the normal case and the case of a 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> 0 D/G outage extension.
The'results are presented-in Attachment E.
The largest contributor to risk was found to be a hypothetical event involving the small break LOCA with simultaneous loss of all offsite power and failure of ESF bus 248 or-249 to energize.
The analysis shows that a 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> diesel outage extension does increase the probability that charging will not be available during a LOCA.
However, the PRA has used deterministic methods (also discussed in Attachment E) to analyze
, the probability of core melt with one of four charging or SI pumps available.
L
~ The analysis shows that the risk of core melt is not significantly increased L,
by the 96' hour outage extension outage, because coolant is supplied by the available SI pump.
This conclusion is consistent with the Zion Probabilistic Safety Study (ZPSS), which requires only one of four charging or SI pumps for safe shutdown from a LOCA. A description of the methods and computer codes i
L used for this analysis is also incorporated into Attachment E.
l l
2185o(19) 1
7_
,'i ATTACHMSLG (Continued)
The Commonwealth Edison Nuclear Fuel Services department has analyzed the effects of the 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> 0 diesel generator outage time extension on containment pressure reduction capability during a postulated accident.
The accident most sensitive to containment pressure reduction concerns is the large break LOCA.
Using NRC approved methodology, a containment integrity analysis was run assuming a large break LOCA with only one containment spray pump and two Reactor Containment Fan Cooler's (RCFC's) operable.
The analysis showed that peak containment pressure was still within design pressure under this scenario. Additionally, the PRA analysis presented in Attachment E shows that r
the-large break LOCA with only one of the three ESF busses available is not a significant contributor to risk.
The Zion E0P's have been reviewed for applicability during a LOCA with only one ESF bus available.
The review showed that the possibility of having either one available charging pump or one available SI pump is accounted for in the E0P's, and that the procedure allows for unit shutdown following a LOCA using either a single charging pump or a single SI pump.
The diesel generator test circuitry is designed so that ESF component actuation is not interfered with in the. event of a safeguards actuation during a diesel test.
This is confirmed every outage during performance of procedure T.S.S. 35, which simulates loss of offsite power and Safeguards Actuation and verifies proper operation of the circuits which sequence ESF loads onto the L
diesels.
1 Based on these analyses, Commonwealth Edison has concluded that the.
probability.and consequences of accidents previously evaluated are not significantly increased, the possibility.of a new accident is not created, and the margin of safety is not significantly reduced.
He believe this request meets the criteria for no significant hazards consideration.
l^
L i
i.w 2185o(20)
l l.;..
ATTACHMENT C (Continued)
TABLE 1 I
4KV Bus Components 147/247 148/248 149/249' Containment Spray Pumps A.,
B C*
Safety Injection Pumps A
B Centrifugal Charging Pumps B
A Auxiliary Feedwater Pumps B
C Residual Heat Removal Pumps B
A Component Cooling Pumps OE(147)-
0D(148) or OB(248) OC(149) or 0A(249)
Service Water Pumps A
B C
Containment Fan Coolers C
A, D B, E
- The "C" Containment Spray Pump is diesel driven. However, its operability requires operability of motor driven valve 2MOV-CS0006, which is a normally closed containment isolation valve powered by bus 249.
/scl:0114T:21 2185o(21)
. :^
ATTACHMENT C'(Continued)
TABLE 2 NUMBERS ARE PER UNIT UNLESS DESIGNATEC Number Available Number Required Number' Required With No Power Comoonents
_in Licensing Bases in ZPSS
. to Buses 247/147 Containment Spray Pumps 1 (Iodine removal) 1 2
2 (Pressure suppression)
Safety Injection Pumps 1.
Any 1 of 4 SI
'I and/or charging pumps Centrifugal Charging Pumps 1
1 Auxiliary Feedwater Pumps 1
1 3 (including steam driven pump).
Residual Heat Removal Pumps 1
1 2
Component Cooling Pumps 2*
1* (Loss of offsite 4*
power) 2* (LOCA)
Service Water Pumps 2*
3*
4*
Containment Fan Coolers 3.
3 4
- Total for both units 218So(22)
~, - _ _- __ _,_..=
=:-
. _ =
.=
L..-
ATTACHMENT C (Continued)
.TA8LE'3 References for Table 2 1
Components Zion Tech. Spec. Bases Zion Prob. Safety Study Containment Spray Pumps
- p. 154 Vol. 3, p. 1.5-545 Safety Injection Pumps
- p. 193 Vol.-3, p. 1.5-361 Centrifugal Charging Pumps
- p. 193 Vol. 3, p. 1.5-361 Auxiliary Feedwater Pumps
- p. 162 Vol. 3, p. 1.5-683 Residual Heat Removal Pumps
- p. 193 Vol. 3,-p. 1.5-400 Component Cooling Pumps
_ p. 194 9/12/83 F. G. Lentine to H. R. Denton; Section 3,
- p. 1.2-I.5 Service Water Pumps pp. 194, 195 Ibid., p. I.6, I.7 Containment Fan Coolers
- p. 148, 154 Vol. 3, p. 1.5-580 Additional references:
Amendment 68 to DPR-39 (9/25/81) and SER Amendment 72/66 to DPR-39 and DPR-48 (12/31/81) and SER 2185o(23)
- j., n -.
f.,
h
>k ATTACHMENT D
- ZION NUCLE 6fLEQWER STATION 1
t:
EXCERPT FROM ZION PROBABILISTIC.
SAFETY STUDY. VOLUME 3 r.-
e l
.i l'
I i
i 1-i l
l-2185o(24)-
1 i
ATTACHMENT D frerot from Zion Probabilistic Safety Study. Vol. 3 l
It is assumed in this study that all initiating events ultimately result in a trip of the main generator on the affected unit.
Since each of the 7,lon units is rated at.1085 MH(e), the instantaneous' loss of this' input could have a significant effect upon the stability of the offsite power supply network due to reduced transmission voltages, frequency' fluctuations, or power flow imbalances as the grid recovers from the transient.
The Commonwealth Edison transmission network has been designed to provide a stable power supply grid under conditions of multiple, large generating unit and major transmission line outages. Utility interconnections to the north, west, south, and southeast provide excellent geographical stability.
Two of these tie lines (to the Hisconsin Electric Power Company) directly supply the Zion 345 kV switchyard, providing increased diversity in the power sources for this specific site. Detailed guidelines have been established for the entire Commonwealth Edison power supply network which define the basis for system operation under a wide variety of steady-state and transient conditions. A prime consideration in the establishment of these guidelines is the requirement that no single loss of a generating unit or transmission facility should result in an unacceptable condition of degraded system operation.
System operating contingencies are defined by these guidelines and specify the need to provide additional. generating capacity from Commonwealth Edison's own facilities or to provide power from network inter ties long before critical operating stability limits are approached. A detailed voltage reduction and selective load shedding program is also specified in order to maintain grid stability with adequate margins under the most severe conditions.
Detailed system stability studies have been performed to verify the efficacy of these operating guidelines under a wide range of scenarios. The combined effects of these guidelines and the overall design stability of the Commonwealth Edison
~
transmission network are demonstrated by the fact that there has never been a major grid failure or a local failure of the offsite power supply to Zion as a result of the loss of a single _ generating unit.
(In fact, there has never been a loss of all offsite power at Zion from any cause.)
The assignment of a distribution for the probability of losing all off-site power to the Zion switchyard as a result of the trip of one of the Zion units is an extremely difficult task.
Factors affecting this condition are total system load,.available spinning reserve capacity, the fraction of the-load being supplied from the Zion units, the status of neighboring utilities' networks, scheduled and unscheduled outages of specific generating units and' transmission lines, etc.
The analysis of this problem presented in HASH-1400 applies a median value of 10-3 for the conditional loss of offsite power as a result of a unit trip.*
L
- WASH-1400, Appendix II, page 34.
{
I 2185o(25)
^
.. '.,l-1.
A.TTACHMENT D (Continued)
Several factors limit the applicability of this value to the specific problem faced in this study.
The HASH-1400 distribution was developed from a
-review of Federal Power Commission studies of power supply networks in a wide variety of locations east of the Rocky Mountains.
While it may be applicable to the composite site studied in NASH-1400, it is certainly not directly applicable to the Zion site.
It must also be recognized that the reference study is now several years outdated and that significant advances in the design, operation and overall stability of virtually all transmission networks in the United States have been made during the intervening years.
Unfortunately, no specific studies have been conducted to address the precise-problem posed at Zion.
However, a review of the available Commonwealth Edison system design criteria and operating Ouidelines, several.
grid stability studies focusing specifically upon'the Zion site, and conversations with Commonwealth Edison engineers have provided a significant amount of applicable information. As a result of these inputs, we have assigned the following distribution for the probability of losing offsite power to the Zion switchyard-as a result of a trip of either of the Zion units:
Median:
2 x 10-5 failure / unit trip 5th Percentile:
4 x 10-7 failure / unit tr.ip 95th Percentile:
1 x 10-3 failure / unit trip The mean and variance of this assumed log normal distribution are Mean:
3.38 x 10-4 failure / unit trip i
Variance:
3.26 x 10-5, A-few observations must be made in order to place this distribution in a proper perspective.
It is our best estimate of a conservative distribution to be applied to this analysis only. Although it is broadly _ based upon the i
results of Commonwealth Edison grid stability studies developed for the Zion L
site, none of the studies reviewed was directly applicable to the problem at hand. 'Most of the studies address a condition less severe than the total loss of offsite power, and those reports developed for the loss of power do not address it from the cause-effect relationship presented in this analysis.
For i
these reasons, we feel that the median value of our distribution is a very
-conservative estimate for the frequency of this event.
However, we also feel that the assigned broad distribution adequately accounts for our uncertainty in this value.
The given distribution is thus considered to represent a conservatively bounding estimate for the conditional failure of offsite power, which is as specialized to the Zion site as is possible with the existing information base.
The mean frequency for the loss of offsite power at a generic plant site from the given population, excluding the Zion evidence, was determined to be 0.194 events per site calendar yen.
The updated mean frequency for the loss
'of offsite power at the Zion site is 0.068 events per site calendar yRAr.
-2185o(26)
L.* (
i AIIACliMENT D (Continued) i Excerpt from " Response to the BNL Peer Review of the Zion Probabilistic Study," F. G. Lentine to H. R. Denton, dated 9/6/82:
A'two-stage Bayesian-update was performed using the data from Table 1.
As suggested by the BNL review comments, the Zion data was excluded from the plant population for the first step of the analysis.
The plant population data was applied using the total number of offsite' power failure events.for each site'and the total number of site years listed in Table 1 (i.e., not
- accounting for the effects of unit availability).
The generic data was updated using the-Zion site specific evidence of no failures in 9 site years.
The resulting distribution provides the specialized calendar y nt frequency of loss of offsite power events at the Zion site, regardless of unit operating
' conditions.
This distribution was then multiplied by the average Zion unit availability (0.71) to obtain the frequency of losses of offsite power to a Zion unit dating Rower operation. The parameters of this updated and scaled distribution are 5th Percentile:
1.04 x 10-2 failure / unit operating year Median:
3.63 x 10-2 failure / unit operating year 95th Percentile:
1.27 x 10-1 failure / unit operating year Mean:
4.85 x 10-2 failure / unit operating year The mean frequency for the loss of offsite power at a generic plant site from the given population, excluding the Zion evidence, was determined to be 0.194 events per site salendar y ur.
The updated mean frequency for the loss of offsite power at the Zion site is 0.068 events per site calendar y n r.
i L-l l
2185o(27)
i
(- --
r t :-:,,..
?1-p
\\; a ATTACHMENT E 110B_NUELEAR POWER STATION PRA EVALUATION OF A 96 HOUR _0UTAGE TIME EXTENSION OF "0" DIESEL GENERATOR
~2185o(28)
- i l.
ATTACHMENT E PRA Evaluation of_96 hour outage line extension "0" D/G extension with two units in hot shutdown.
He have examined a number of cases to establish the potential effects of extending the allowed outage time for the diesel.
First of all, we consider the large LOCA's as initial events.
In terms of the FSAR analyses, the "0" diesel is immaterial since the FSAR requires one out of two (1/2) RHR trains and 3/4 accumulators for.large LOCA's.
The RHR trains operate from emergency buses served by the non-shared diesel.
He have used the MAAP computer code to examine the response of the ECCS to large LOCA events on a realistic basis and.
find that, for breaks greater than.06 square feet and up to 2.0 square feet, any one of the six ECCS pumps will prevent core melt. Our study did not go beyond breaks of that size since it is judged that the likelihood of even that sized break is extremely small.
(Please note that the MAAP study discussed herein covers a wide variety of cases and is attached as Appendix A for your information.) Given the fact that, on a realistic basis, our success criteria is viewed as 1/6 pumps, we could argue that increasing the "0" diesel outage time increases the probability of core melt. He now have to determine if the increase is, in any way, significant.
He center our evaluatton around the probabilities associated with large LOCA and no AC power. Our base point for these derivations is the Zion Probabilistic Safety Study (ZPSS).
For the base case evaluation, we look at a LOCA frequency of 9.4E-4 and a total loss of AC to the emergency buses.
occurring with a conditional frequency of 1.85E-7 (table 1.3.4.1-1).
The associated base case core melt probability is the product of these values or
- 1. 74 E-.10.
Now, how much does this increase if we extend the allowed outage I
time on the "0" diesel to 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br />? By using the same table in ZPSS, we find l
that the frequency of losing buses 148 and 149 is 2.28E-6.
He can also L
establish a new frequency for the loss of bus 147, given the extended outage, of 2.3E-1.
The new associated core melt frequency is-determined to be less than 4.93E-10. The extended outage time does have an effect on the core melt frequency but, given the overall large LOCA core melt frequency, from all l.
causes, of about 9.4E-4 times 6.6E-3, or about 6.4E-6, the extended outage has l
an insignificant effect on large LOCA core melt sequence probability.
The next step is to bound the issue at the lower end by examining small LOCA events.
L l
2185o(29)
1 l.
e ATTACHMENT E (Continued) i To put the situation in perspective, we have to consider the impact l
of the 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> outer time extension on the plant using realistic success criteria such as those in the ZPSS.
The ZPSS requires 1/4 charging or hi-head j
SI pumps be available to prevent core melt.
(Our MAAP study confirms this and, moreover, shows that 1/6 of ECCS pumps will prevent core melt for all break sizes between.06 square feet and 2.0 square feet.)
Performing the evaluation in a manner similar to that discussed above, we note that the conditional frequency of core melt from all small LOCA events is about 6.33E-4.
The frequency of the small LOCA is 3.54E-2.
This gives us a total probability of core melt from small LOCA events of 2.2E-5.
Now
-evaluating the effect of a 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> outage extension on the "0" diesel, we find a new core melt frequency of 2.2E-5.
In short, the_ extension has no effect given realistic success criteria.
The conclusion we draw from this evaluation is that a 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> time extension outage for the "0" diesel will not have any adverse impact on core melt frequency associated with LOCA sequences,.if realistic success criteria are employed.
G. Klopp PRA Group Commonwealth Edison l
2185o(30) l l
ATTACHMENT E j
Appendix A DISCUSSION OF THE "MAAP" COMPUTER CODE METHODOLOGY USED TO PERFORM DETERMINISTIC ASSESSMENT OF EFFECT OF 96 HOUR 0 D/G OUTAGE TIME EXTENSION ON PLANT RISK ASSUMING LOCA WITHOUT OFFSITE POWER j
The following is a discussion of the methods used by the Commonwealth Edison PRA/IPE group in assessing Zion's capability to shutdown safely following a postulated small break LOCA coincident with Loss of All Offsite Power (LOOP) during the proposed 96 hour0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> 0 diesel generator outage time extension.
It was assumed that one other diesel also falls during the event, requiring unit recovery from the accident with no charging pump available, The small break LOCA was identified as the accident most sensitive to charging pump inoperability, because for large breaks the RCS pressure decreases ouickly to below the shutoff head of the other ECCS pumps.
The basic analysis tool used for this evaluation was the MAAP computer code and its associated Zion parameter file.
The code was run on a VAX computer owned and operated by Edison.
The results of the NAAP runs were consolidated and plotted using the MATCHAD code (2.0) on an IBM AT.
The MAAP code, while not an NRC licensed code for accident analysis, was developed by the industry as part of the Industry Degraded Core Program to provide a realistic basis for transient evaluation in the area of severe accidents.
The code has been benchmarked against established codes and experimental data.
EPRI reports documenting the benchmarking of this code are
. listed as references.
The LOCA events postulated were all hot leg breaks.
No major impact, if any, would result from using cold leg breaks.
Zion recognized that the cold leg small break LOCA is normally the limiting case in FSAR Chapter 14 analysis. 'This is largely because the codes employed are focusing on Peak Clad Temperature (PCT) as a key variable, and because of the inherent conservatisms in normal FSAR analysis methods required by the Appendix K criteria. Using "best estimate" methods and focusing on fuel melt temperature rather than PCT as a key variable, it was found that the difference between hot leg and cold leg break is not as significant. The time to reach fuel melt temperature was found to be much less sensitive to hot leg versus cold leg breaks when using a "best estimate" thermal-hydraulic methodology.
This has been confirmed by running the code both ways in previous analyses.
Break l
sizes ranged from 0.005 square feet up to 2.0 square feet.
The upper. limit of i
2.0 square feet was selected subjectively. Clearly, for risk assessments, the design basis break has a vanishingly low importance.
When engineered safeguards were operational, only the minimum number of trains were allowed to run.
For example, only one RHR train out of two would run. This applied to containment safeguards as well.
2185o(31) l
.1
~.l.
ATTACHMENT E Appendix A Study _Beihodology g
The conditions to be analyzed were established in the form of a matrix with
~
.the columns consisting of break sizes and the rows consisting of " cases"
-showing various combinations of ECCS availability. A total of nine break sizes was employed and six cases were developed.
This led to 54 runs for the MAAP code.
Each run'was allowed to proceed through 50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br /> of accident time on the VAX.
In each case, the operation of ECCS in the recirculation mode was not permitted. A total of six key variables was plotted.for each run as a-function of time.. In addition, the time to empty the RHST (eg the
-termination of ECCS operation) was extracted from the data file for each run.
The break sizes evaluated were:
0.005 sq. ft.
i 0.01 sq. ft.
0.025 sq. ft.
0.05 sq. ft.
0.075 sq. ft.
0.1 sq. ft.
0.5 sq. ft.
j 1.0 sq. ft, j
2.0 sq. ft, d
The ECCS system combinations employed were:
Case 3, RHR and Charging /SI only Case 4, SI only Case 5, SI and Charging /SI only j
Case 6. Charging /SI only l
The six data plots made from each of the resulting 54 MAAP runs were time dependent plots of; j
MCR, the-mass of intact core material l
PPS,-the pressure in the primary system i
MH2CRT, the mass of hydrogen produced in the core l
PB, the containment pressure MCMC, the mass of molten core material in the vessel cavity MCMB, the mass of molten core material on the containment basemat l
The variable of major import is the very first one, MCR.
It allows one to l
L establish the onset of core melt in an unambiguous manner. Also, for analysts l
L familiar with severe accident work, it allows one to establish the onset of I
l.
metal water reactions.
The other five variables were selected to provide I
clarifications, confirmations, and perspectives on the accident sequences i
being considered.
l
-2185o(32)
L :_
l
[
-i
.l, i
i ATTACHMENT E Appendix A Once the 324 plots were in hand and had been reviewed for consistency, tabulations ~were made for each of the six cases of the time to onset of core melt as a function of break size and of the time to empty the RHST as function-of break size It quickly became apparent that only three cases (1, 4 & 6) were of interest.
These were plotted using MATHCAD.- Another case, case 3, was also tabulated and plotted to verify the consistency of the results, fLt.SMlh The results for each of the four cases are depicted in figures 1 through 4.
Figure 1 shows the study result for case 1, operating with only the RHR system. Clearly, the RHR system protects the core for the larger breaks (down to about 0.06 sq. ft.) as expected.
For the smaller breaks, the-primary-system remains pressurized above the level at which the RHR pumps can deliver enough water to cool the core.
In these circumstances, core melt occurs before the RHST is depleted.
1 Figures 2 and 3 show the results for cases 4 and 6 respectively.
In both cases, the core is protected throughout the injection phase of.ECCS operation..This indicates that both the charging /SI system and the SI system will-provide adequate core cooling with one pump from either system for the spectrum of breaks considered.
Herein lies the merit of this analysis approach.
Even the most optimistic subjective evaluations to date assumed that 2 out of the 4 pumps in these systems would be required.- By showing-that only one' pump is-required, we have substantially reduced the success criteria burden for risk assessment work.
He have also provided a fresh perspective on
.this matter for other purposes.
Figure 4 shows the results for case 3 and simply reinforces the previous-results.
The figures provide a great deal of other information relative to the timing of events.
For example, by looking at the figures, it becomes clear.that the shortest time to a core melt occurs for case 1 and is about 2-hours after the start of the accident. More typically, the earliest times are
.on--the. order of 4 to 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> and range up to a full day.
These types of perspectives may well prove useful for a variety of purposes.
L 1
L REFERENCES 1.
EPRI Report NP-6178-L "MAAP 3.0 Simulation of OECD Loft Experiment LP-FP-2" 2.
EPRI Report NP-6206 "MAAP 3.0 B Rev. 12 Modeling of the TMI Accident" l
l 2185o(33)
29 l
s
- ps
,1
'g i
\\
l h
l l
I
/
T.,M
\\
N i
i
'l I
\\
s
\\
\\
3 x
r i
N x
\\ k-/
k s,
i
/
I
\\,/
'\\
N
\\
-...N N
T i
0
.001 B
10 i
CASE 1, LPI ONLY, where T is time to empty the RWST, M is time to onset of core melt, and B is break size in square feet.
FIGURE 1 i
l l
I l
. l.
. f.
4
=
~
a
35
~'
e
\\
\\
\\
N
\\
T,M i
i p
\\
\\
\\
s N
s N
M N
. N i
T
~
0
.001 B
10 i
CASE 4, SI ONLY, T is time to empty RWST, M is time to onset of core melt, and B is break size in square feet FIGURE 2 t
1 i
l l
l
lt
(*,y J '*.
4:
4
\\
\\
h h
\\
\\
k N.
s
.T',M
'i i
\\
\\
\\
s \\-
M i
3
.i
\\-
T
(
i 0
.001 B
10 i
CASE 6', C.%RGING/SI ONLY, T is tima to empty RWST, M is time to onset of core malt, B is break size in square feet FIGURE 3
[
- 4l;$3 R ij.;;l.,;.'. -
l l g;c
]
-g.
N g
\\
1 k.
l l-N h
\\
.T,M
- i. i i
\\[
L
\\
i l
\\
-1
\\
\\ N
'N N
N i
s M
s g
s '
N T
i
.001 B
10 CASE-3, RHR AND CHARGING /SI ONLY, T is time to empty RWST, M is time to onset of core melt, and
,B is break size in square feet.
m l
r.
~ ^ ^
- - - - -