ML20211E091
| ML20211E091 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 02/11/1987 |
| From: | Mroczka E NORTHEAST NUCLEAR ENERGY CO., NORTHEAST UTILITIES |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| B12393, NUDOCS 8702240230 | |
| Download: ML20211E091 (13) | |
Text
~
O General Offices
- Selden Street, Berlin, Cowecticut Ni'E.
rn$rNcow P.O. BOX 270
- n os..na.o ra ca==
HARTFORD. CONNECTICUT 06141-0270 k
C J 'j",[^',' [,*1*%' *,**"]
(203) 665-5000 February 11,1987 Docket No. 50-245 B12393 Re: Integrated Safety Assessment Program U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, D.C. 20555 Gentlemen:
Millstone Nuclear Power Station, Unit No.1 Probabilistic Safety Study Update In a letter dated >uly 10,1985,(I) Northeast Nuclear Energy Company (NNECO) docketed with the Staff a summary report of the Millstone Unit No. 1 Probabilistic Safety Study (PSS).
This submittal described in detail the background of the PSS, and focused in part on the ongoing implementation of a Living PRA program for Millstone Unit No.1. The major element of this Living PRA program is the development, maintenance and use of PRA models for assistance in evaluating potential plant backfits and operating procedures modifications. The Living PRA program affords us the flexibility to quickly and accurately analyze the impact on plant safety of changes to the plant's design configuration.
As such, the PRA models and methodology supporting the Living PRA program must be periodically updated to incorporate plant design changes, significant operational changes, and relevant updated equipment performance data. This letter transmits new insights gained from the first such update which has recently been completed. These insights, which result in a net reduction of the core melt frequency due to internally-initiated events from 8.07 X 10-4/yr to 5.18 X 10-4/yr, are presented in Attachment 1. contains a detailed analysis of our continued efforts to decrease the core melt frequency by improving long-term cooling capability. Attachment 3 forwards the amended text pages. NNECO believes that the Millstone Unit No.1 PSS, as amended, provides an accurate analysis of the current risk associated with operation of the plant.
4 (1> >.
. e-a,e te, to,. A. zwo,1nm,.__ en,t No.1 P_st1c Safety Study - Results and Summary Report," dated >uly 10,1985.
I
/e B702240230 870211 ADOCK0500g5 PDH P
J
We have forwarded twenty (20) copies of the update directly to the ISAP Project Directorate for distribution within the NRC.
If you have any questions on this material, please feel free to contact my staff.
Very truly yours, NORTHEAST NUCLEAR ENERGY COMPANY E. 3. Mrppika p'
j Senior V/ce President cc:
Dr. T. E. Murley, Region I Administrator
- 3. Shea, Millstone Unit No.1 Project Manager T. Rebelowski, Millstone Unit No.1 Resident Inspector e
i
(
1 i
I j
Docket No. 50-245 B12393 ATTACHMENT 1 Northeast Nuclear Energy Company Millstone Unit No.1 PSS Update Summary of Results I
i February 1987
~
..,.~%-,.,.m_.,
v.-y,,,,
__my.,,,,_,,,m..-
,_,.,_.,,y.
,._...,m,.-,_
_.,_. _.e m mee.,,_,, m
,,,.w,,,.,,,_.,,_,,...m,_,.
e PSS Update Results The following revisions have been made to the Millstone Unit No.1 PSS as a result of plant modifications and additional analyses completed prior to the update.
1.
The unavailability of the Shutdown Cooling System (SDCS) was revised:
a.
The SDCS unavailability for the case in which no fuel failure has occurred was revised based on re-evaluation of the number of system failures and the number of system demands.
b.
The unavailability of the SDCS was changed to 1.0 (i.e., SDC not available) for cases in which fuel failure has occurred and the reactor building is not accessible. New information showed that placing the SDCS in service always requires local operation of certain equipment in the reactor building.
c.
The unavailability of Reactor Building Closed Cooling Water (RBCCW) system was revised due to a change in the test interval of MOV RC-39 based on plant experience. (RBCCW is a support system for SDC.)
2.
The unavailability af Emergency Service Water (ESW) was revised as a result of changes in pump test intervals and control room annunciation for failures of critical back-up power sources.
3.
The unavailability of Alternate Shutdown Cooling (ASDC) was revised:
a.
To include the revised success criteria, 2 LPCI heat exchangers 1 ESW pump per heat exchanger 1 LPCI pump per heat exchanger b.
To include changes to the ESW system (see #2), and c.
To include changes to the LPCI system.
LPCI pump motor bearing tube oil cooling water flow, which was previously not specifically verified, is now verified monthly.
i 4.
The end states for sequences 1 and 2 in the ATWS - 1 tree were changed from " Damage" to " Success" to remove a conservatism in the original analysis.
5.
The initiating event frequency for loss of 120V Vital AC power was l
revised based on newly-installed control room annunciation of failures of the back-up power supply and revised unavailabilities for critical
(
components based on plant operating history.
6.
The loss of 120V Vital AC event tree was quantified using the new l
initiating event frequency and incorporated into the total Core Melt l
Frequency (CMF). (Loss of Vital AC was not incorporated into the core I
melt frequency prior to this amendment.)
I i
l i
7.
All of the changes listed above were incorporated into the appropriate event trees and the event trees were requantified to produce a new core melt frequency.
8.
The text of the PSS was revised to reflect all of these changes.
A's a result of these changes, the total core' melt frequency due to internally initiated events was reduced from 8.07 X 10-4 per reactor-year to 5.18 X 10-4 per reactor-year.
This CMF reduction is due entirely to the reduction of the unavailabilities of the following systems:
o o
Reactor Building Closed Cooling Water, and o
Alternate Shutdown Cooling.(due to the revised success criteria and lower unavailability of ESW).
These higher system availabilities were a result of recent system modifications, and insights gained from further analysis. The reduction in the CMF would have been larger, but two changes were made that increased the CMF:
SDC was modeled as being totally unavailable for those sequences o
resulting in fuel failure.
o The contribution of loss of 120V Vital AC was included in the total-CMF.
Some significant results of the requantification of the CMF include:
.o
.The percentage of the total CMF related to failure to maintain adequate long-term cooling has dropped from 64% to 42%
The CMF due to loss of 120V Vital AC is 1.65 X 10-5/ reactor-year.
o
- All of the dominant sequences resulting in early core melt remain the.
o same (although their percent contribution to the total CMF changes).
However, one additional early core melt sequence was added by the loss of 120V Vital AC initiator.
As a result of the total unavailability of SDC following fuel failure, the o
frequencies of some of the original intermediate core melt sequences increased. A new dominant intermediate core melt sequence has been added due to the inclusion of loss of Vital AC. This sequence includes the failure of SDC as a result of fuel failure.
See Table i for a summary of the revised core melt frequency contributions by initiators.
Conclusions
~
NNECO concludes that the PSS, as revised and requantified, accurately reflects
,the plant design and procedural modifications that have been undertaken voluntarily by NNECO to enhance the safety of Millstone Unit No.1, as reflected
. by the decrease in the core melt frequency due to internally initiated events from 8.07 X 10-4/yr to 5.18 X 10-4/yr, a 35% reduction. Additionally, based on the hardware and procedural modifications made to date, NNECO has concluded that continued operation of Millstone Unit No. I does not pose any undue risk to the health and safety of the public.
_m,
" + -
TAILE l.
CORE MLT F1EQUENCIES BY INITIMDBS Symbol Description CM Frequencies /
Percent year Contribution T
Loss of Normal Power 1.59E-4 30.74 g
T Reactor Transients 2
-with Main Condenser Available 8.56E-5 16.53
-with Main Condenser Isolated 1.26E-5 2.42
-Reactor Trips 4.62E-5 8.93 T
L ss of Feedwater 9.41E-5 18.19 3
T Loss of Service Water 1.21E-5 2.33 4
T L ss of Reactor Building closed 5
Cooling System 1.77E-8
<0.01 T
L ss of Turbine Building Secondary 6
Closed Cooling System 1 39E-6 0.27 T
Loss of 120V Vital AC 1.65E-5 3 19 7
SSB Small Small Break LOCA 3 18E-5 6.15 2
(Area < 0.01 ft )
l SB Small Break LOCA 3.76E-5 7JT 2
2 (0.01 ft < Area < 0.2 ft )
LB Large Break LOCA 4.41E-6 0.85 2
(Area > 0.2 ft )
IORV Inadvertent cpening of a Safety / Relief Valve 1.61E-5 3 11 Ursitigated IC Tube Rupture 1.50E-7 0.03 i
Unisolated LOCA in the RWCU 1 39E-8
<0.01 Interfacing System LOCA in the LPCI System 1.61E-8
<0.01 I
Unisolated LOCA in the Core Spray System
' 1.10E-7 0.02 l
TOTAL 5.18E-4 100 l
l
Docket No. 50-245 B12393 ATTACHMENT 2 Northeast Nuclear Energy Company Millstone Unit No.1 Long-Te,rm Cooling Analysis l
l L
i r
i February 1987 l
===.
Background===
By letter dated July 10, 1985, Northeast Utilities Service Company (NUSCO), on behalf of Northeast Nuclear Energy Company (NNECO), submitted to the Staff a plant-specific Probabilistic Safety Study (PSS) for Millstone Unit No.1. A brief summary of the results of this study and the NUSCO evaluation of these results were presented as attachments. It was determined that approximately 64% of the total calculated core melt frequency at Millstone Unit No. I was due to a failure to maintain adequate long-term decay heat removal capability.
The PSS modeled four long-term cooling systems. The conditions under which each of the systems could be used are summarized below.
o Main Condenser - The main condenser can be used only if the feedwater system is operating and no significant fuel damage has occurred during the transient.
o Isolation Condenser - The isolation condenser is a passive decay heat removal system consisting of a shell and tube heat exchanger and is sufficient to remove decay heat for approximately 45 minutes provided there are no breaks or significant leaks in the reactor coolant system. Operation of a makeup valve to the shell side of the IC from the fire water system allows decay heat removal for an indefinite period of time.
o Shutdown Cooling System (SDCS) - The SDCS can be used if RPV level is restored to the normal operating level. This system may not be used in the event of a LOCA where the RPV level cannot be restored to the normal level or when fission products in the coolant preclude local manual valve operation.
Alternate Shutdown Cooling (ASDC) (or Containment Cooling)(1) - ASDC can o
be used during all accident sequences except an ATWS event in which the standby liquid control system has becn used to reduce reactor power. This is to prevent dilution of the boron injected into the RPV, which could result in an increase in reactor power.
ASDC provides a mechanism for cooling the torus water and injecting it into the RPV.
In ASDC, coolant is transferred back to the torus via the auto depressurization system (ADS), in which steam is bled off through opened main steam line safety / relief valves. Following a LOCA, in the containment cooling mode, coolant is transferred to the torus via the break opening.
(1)
Alternate Shutdown Cooling or Containment Cooling is a combination of systems / components and procedures which are utilized to shut down and cool the reactor during abnormal operating situations. Future references to the ASDC system refer to the systems / components and procedures which comprise the means for alternate shutdown and cooling of the core during transient situations.
l l
Af ter reviewing the dominant core melt sequences, it was evident that the failure of the ASDC was one of the major contributors to the core melt frequency.
Unavailability of the main condenser and isola +. ion condenser systems also contributed to core melt risk; however, the calcu ated unavailabilities were an order of magnitude lower than that of the ASDC system. Therefore, to reduce core melt risk, it is most important to understand 4he major causes of failures of the ASDC system.
Alternate Shutdown Coolinn Emergency procedures require the use of two LPCI pumps (one in each train) to take suction from the torus and inject coolant into the RPV. Coolant returns to the torus via the S/R valves (which are manually opened) or via the break opening in the case of a LOCA. To establish torus cooling, LPCI flow is established through the LPCI heat exchangers. These heat exchangers are cooled by emergency service water (one out of two ESW pumps per heat exchanger).
If adequate torus cooling is not provided, the torus slowly heats up, resulting in a decrease in net positive suction head (NPSH) available to the LPCI pumps.
Following a LOCA, the overall containment pressure rises which increases the NPSH available. Analysis shows that if there is no increase in drywell pressure, the NPSH available falls below the value required to maintain full pump flow when the torus heats up to 1760F. At that time, the pumps wilt begin to cavitate unless flow is throttled by the operator. The resultant reduced pump flow (either due to cavitation or throttling) will result in a decrease in heat transfer through the LPCI heat exchangers, yielding an increase in the rate of heat-up of the torus.
Each LPCI heat exchanger is rated at 40 X 106 BTU /hr. at the following conditions:
Torus Temperature - 1650F ESW Temperature - 750F ESW Flow - 5000 GPM LPCI Pump Flow - 5000 GPM As a result of the limited heat removal rate of each LPCI heat exchanger, operation of both LPCI heat exchangers is required to keep the torus temperature below 1760F, in order to satisfy the LPCI pump NPSH requirements. Since 2 (out of 2) trains of the ASDC systems are required to maintain adequate long-term cooling capability, the calculated ASDC unavailability is high.
In an effort to decrease this calculated ASDC unavailability, NUSCO has instituted j
the following modifications:
i o One of the dominant causes of ASDC failure in which 2 out of 2 LPCI i
trains are required was the failure of the solenoid valves controlling LPCI pump lube oil cooling. Previously, surveillance procedures did not require j
the verification of the proper operation of the solenoid valves that permit cooling water to flow to the LPCI pump lobe oil coolers. Surveillance procedures have been revised to test the operation of these valves on a j
monthly basis. This will improve the reliability of the LPCI system and reduces the core melt frequency attributable to long-term cooling failure by about 20%
. o Modifications to the valve operator and limit switch logic of the isolation condenser condensate return line valve (IC-3) have been made to address one of the dominant predicted causes of isolation condenser failure.
These and other modifications have resulted in a net reduction of the core melt frequency by about 35% However, based upon NNECO's firm commitment to lower the core melt frequency even further, NNECO is currently continuing in-depth analyses of the following issues:
o Replacing the LPCI heat exchangers with much larger capacity units. This would allow decay heat removal capability utilizing only one LPCI and ESW train.
o Proceduralizing the use of the reactor water clean-up system as a long-term decay heat removal source, either as a sole source of decay heat removal or in conjunction with one train of the LPCI system.
o Proceduralizing the use of the control rod drive or fire system pumps in conjunction with the condensate w /.er storage tank inventory as an additional source of reactor vessel makeup.
o Permitting containment venting as a decay heat removal scheme if torus cooling capability is lost.
o Splitting the power sources for the pumps supplying flow to each LPCI heat exchanger (i.e., one LPCI and one ESW pump being powered by the emergency diesel generator and the other set of pumps being powered by the gas turbine generator).
This would make each train electrically independent.
o Making piping modifications which would make the system mechanically single failure proof.
In a letter dated December 23, 1985,(2) NNECO discussed potential actions and procedural modifications that could possibly have favorable impacts on lowering the ;alculated core melt frequency. Additionally, several long-term activities were identified which were scheduled to support implementation of hardware and/or procedural modifications no 1 ter than the 1987 refueling outage. NNECO 9
further discussed in an April 30, 198613) letter, tentative plans to replace the LPCI heat exchangers with heat exchangers of about twice the existing heat removal capacity.
This would improve the capability to maintain adequate long-term cooling. The necessary design work required to enable procurement of there heat exchangers to support their installation during the 1987 refueling outage was on-going.
However, a final decision on whether or not to replace these heat exchangers had not yet been'made.
e (2)
- 3. F. Opeka letter to C. I. Grimes, " Millstone Unit No.1 - Probabilistic Safety Study," dated December 23,1985.
(3)
- 3. F. Opeka letter to C.1. Grimes, " Millstone Unit No.1 - Probabilistic Safety Study," dated April 30,1986.
4 Design, engineering, as well as cost benefit analyses, continued toward completion.
In May,1986, it was determined that replacement of the heat exchangers would cost $3.5 million, resulting in a decrease in the core melt frequency of approximately 16% (based on the calculated core melt frequency at that time). A second option, installing redundant valving which.would make the valving mechanically single failure proof, would cost $2 million with an 8% reduction in core melt frequency. In addition, the recent update to the PSS models, which corrected a few earlier inaccuracies in the SDC and ASDC modeling, has shown that the failure to maintain adequate long-term cooling at the plant now contributes approximately 42% of the total CMF. Based upon these preliminary conclusions and the uncertainty in completing the design, engineering and procurement of the required materials to support installation during the 1987 outage, it was decided that these two options and other possible modifications / procedural changes would be evaluated under the auspices of the ISAP methodology to determine the optimum solution, which would then be included in the subsequent integrated implementation schedule.
9
o..s Docket No. 50-245 B12393 O
ATTACHMENT 3 Northeast Nuclear Energy Company Millstone Unit No.1 PSS Update February 1987 m