ML20100Q279
| ML20100Q279 | |
| Person / Time | |
|---|---|
| Site: | Limerick  |
| Issue date: | 03/01/1996 |
| From: | Walter MacFarland PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9603110559 | |
| Download: ML20100Q279 (7) | |
Text
_
Wittir G. MrcFarland, lV, P.E.
Vce President Limerick Generating Station
+
=-
.y,
PECO NUCLEAR esco eme<ov comneer Po Box 2300 A UNtr or PLCO ENacy Sanatoga, PA 19464-0920 610 718 3000 Fax 610 718 3008 Pager 1800 672 2285 #8320 March 1,1996 Docket Nos. 50-352 50-353 License Nos. NPF-39 NPF-85 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555
SUBJECT:
Limerick Generating Station, Units 1 and 2 Response to a Request for Additional Information Main Steam Safety Relief Valve and Emergency Core Cooling Systems Action Plans Attached is PECO Energy Company's response to the NRC's January 31,1996 Request for Additional information conceming the Limerick Generating Station (LGS), Units 1 and 2, action plans for monitoring Main Steam Safety Relief Valve (MSRV) tall pipe temperatures, Emergency Core Cooling System (ECCS) pump suction strainer differential pressures, and suppression pool cleanliness. The action plans were submitted to the NRC by letter dated October 6,1995. The NRC has concluded that additional information is needed to complete their review. The attachment to this letter provides a restatcment of the NRC's questions followed by our response to each question.
The attached response discusses changes to: 1) the MSRV tall pipe temperature monitoring action plan as originally described in our Octeoer 6,1995 letter, and 2) the exclusion of the High Pressure Coolant injection (HPCI) pump from the ECCS pump suction strainer differential pressure monitoring as originally described both in our October 6,1995 letter and in our November 16,1995 response to NRC Bulletin 95-02, " Unexpected Clogging of a Residual Heat Removal (RHR) Pump Suction Strainer While Operating in Suppression Pool Cooling Mode," for LGS, Units 1 and 2. The bases for these changes are provided in the attached response.
If you have any questions, please do not hesitate to contact us.
Very truly yours, 1
la n k$;
MW c
~
Attachment cc:
T. T. Martin, Administrator, Region I, USNRC w/ attachment N. S. Perry, USNRC Senior Resident inspector, LGS 9603110559 960301 O
r PDR ADOCK 05000352 jg P
ppy y
a 1
/
_.___.m Attachment Docket Nos. 50-352 and 50-353 March 1,1996 Page 1 of 6 Response to a Request for Additional Information Main Steam Safety Relief Valve and Emergency Core Cooling Systems Action Plans Limerick Generating Station, Units 1 and 2 1.
Alert and Action level tal pipe temperatures are provided by the licensee which would determine the proper action necessary regarding the occurrence of MSRV leakage. Please describe how the tai pipe temperature is correlated to MSRV leakage for Limerick 1 and 27 la this correlation based on actual test data or is it analytical? If it is analytical, what are the correlation parameter sensitivities and what is the overall uncertainty in the correlation? What are some of the key parameter values assumed, and what are their expected variabHities?
RESPONSE
The MSRV tailpipe temperature action levels are based on the combination of: 1) General Electric (GE) testing and experience with Limerick Generating Station's (LGS's) Main Steam i
Safety Relief Valve (MSRV) leakage,2) Target Rock industry experience with MSRV pilot valve leakage, and 3) LGS's recent experience with MSRV phot valve leakage.
Regarding the recent experience at LGS, subsequent testing of the five MSRVs replaced after the September 11,1995, spurious opening event revealed phot valve leakage on three of the five valves that exhibited high taHpipe temperatures. These were the only three MSRV phot valves in LGS history that leaked enough to cause elevated taHpipe temperatures. The normal amount of phot valve leakage identified during as-found MSRV testing, i.e., droplets on a mirror or minor whisping of steam, does not result in elevated taHpipe temperatures.
The 1M MSRV pilot valve had leaked for 18 months prior to spurk>usly opening. The MSRV 1
tailpipe temperature was 295'F when the MSRV opened. The 1M MSRV main valve did not leak.
The leakage of the pilot valve could not be measured but was estimated at 3000 lb/hr. The opening was caused by the erosion of the pilot rod which transmits setpoint spring force to the pilot disc.
The 1F MSRV main valve did not leak, but the pilot valve had significant leakage which could not be measured. The leakage prevented performance of as-found testing for setpoint verification.
The MSRV tailpipe temperature was 247'F at the time the 1M MSRV opened. Severe pilot erosion had occurred but the pilot disc was intact. The pilot rod had indications that erosion had started.
The 1D MSRV main valve did not leak, but the pilot valve leaked at 7 lb/hr. The MSRV tailpipe temperature was 215'F at the time the 1M MSRV opened. As-found setpoint testing showed an initial lift preesure at 2.5% above the setpoint pressure. The Technical Specification (TS) limit is
+/- 1% of the setpoint pressure. Subsequentlift pressureswere at 1.3% below the setpoint I
pressure. This behavior indicates that the initial lift pressure was affected by corrosion bonding i
of the pilot disc.
Attachment Docket Nos. 50-352 and 50-353 l
March 1,1996 Page 2 of 6 Future as-found testing of the MSRVs will quantify the leakage where possible on a routine basis.
There are many variables in plant configuration, such as thermocouple location and local drywell temperature, that prevent exact correlation of leakage with tailpipe temperature. This conclusion is consistent with the experience 04 the BWR Owner's Group.
2.
The licensee's Alert level for the tail pipe temperature begins at 225*F which would require that the temperature be trended in order to project when 275*F would be reached. The Action level begins at 250aF and would require that a planned outage be scheduled for when the i
temperatureis projected to reach 275*F. However, at 250*F, the licenses states that the leakage is in the range of 500 to 1000 pounds per hour (Ibm /hr). Leakage tests performed on Target Rock 2-Stage MSRVs in 1983, indicated that at 1000 lbm/hr leakage, the setpoint of the valves would be reduced by more than 10%, resulting in little or no simmer margin for normal operating system pressure. In addition, at some leakage less than 1000 lbm/hr, these tests indicate that the setpoint would drift downward to less than that required by the plant Technical j
Specifications (TS). Therefore, discuss why 250*F was chosen as the Action level when a 1
significantly lower temperature would appear to be necessary to prevent the spurious opening of a MSRV at power and assure that plant TS are met?
RESPONSE
There is no widespread agreement within the widastry that pilot valve leakage causes a reduction in MSRV setpoint. Experience at some plants indicates that MSRV setpoints may increase due to pilot valve leakage. The spurious opening of the 1M MSRV that occurred at LGS, Unit 1, was caused by a mechanistic failure of the MSRV pilot rod assembly, not by setpoint drift. The BWR Owner's Group has evaluated the issues of MSRV setpoint drift versus leakage and pilot valve leakage versus MSRV tailpipe temperature, and concluded that no exact correlatons exist at the present time. The MSRV tailpipe temperature versus pilot valve leakage, in particular, is dependent upon each MSRV's unique configuration.
However, the results of a visualinspection of the 1F MSRV and the condition of the associated pilot valve led PECO Energy to be concerned about the ability of the MSRV to self actuate within the required TS setpoint pressure limits since the loss of pilot disc geometry is assumed to have some affect on MSRV setpoint. As a result, the October 6,1995 MSRV tailpipe temperature action plan was modified.
The modified plan requires evaluating MSRV operability at a tailpipe temperature of 225*F rather than 250 F. This evaluation includes setpoint considerations,taking into account the MSRV tailpipe temperature trending data, the unique configuration of the MSRV and the MSRV leakage history. The appropriateTS action will be taken based on the results of this evaluation considering that the LGS TS require at least 11 of 14 MSRVs to be operable in Operational Conditions 1,2, and 3.
Until instrumentation is available to distinguish between pilot valve leakage and main valve leakage, all leakage is conservatively considered to be pilot valve leakage.
p.
Attachment Docket Nos. 50-352 and 50-353 March 1,1996 Page 3 of 6 in addition, because of the concern over the possible impact of pilot valve leakage, LGS will not continue to operate for an extended penod of time with any MSRV tailpipe temperature above 2501. Based on experience with the 1M and 1F MSRVs, PECO Energy expects that a spurious MSRV opening will not occur before any MSRV tailpipe temperature reaches 275T.
s 3.
The licensee's action plan indicates that a leaking MSRV would be replaced before the associated ta!! pipe temperature is expected to reach 275T. The licensee also states that on September 11,1995, the Unit 1 'M' MSRV lifted at 295T and that it took 6 months for the tail pipe temperatureto rise from 275*F to 295T. However, examination of photographs of the 'M' MSRV pilot disk reveals that the disk was completely eroded into two pieces and that steam had cut completely through the disk wall thickness and was actually eroding the pilot stem for some time prior to the sudden opening of the valve. It appears that signihcant damage to the pilot disk may have already occurred prior to reaching a tail pipe temperature of 2757. Therefore, what would be the maximum tail pipe temperature and maximum operating time criteria which would prevent significant erosion damage of the pilot disk?
RESPONSE
PECO Energy anticipates, based on previous experience, that tailpipe temperature increases for leaking MSRVs will slowly trend upward. Erosion of a pilot rod to the point of causing inadvertent opening of an MSRV takes a long time; approximately 18 months in the case of the 1M MSRV which had operated with an elevated tailpipe temperature since the beginning of the I
operating cycle. As indicated in the response to Question No.1, this elevated tailpipe temperature was caused solely by pilot valve leakage. Although significant erosion had occurred on the 1F MSRV pilot disc, the erosion was not as severe as the 1M MSRV even though the 1F MSRV pilot valve had also leaked since the beginning of the operating cycle, and failure of the 1F MSRV pilot was not imminent. Therefore, erosion damage of the pilot disc is not a concern for spurious MSRV opening over a short period of time.
As a result of the possible impact of severe erosion on the setpoint of an MSRV, PECO Energy has modified the action plan for high MSRV tailpipe temperatures as indicated in the response to Question No. 2 such that the evaluation of MSRV operability will occur at a lower tailpipe temperature than originally identified in the October 6,1995 letter. The modified action plan now requires the evaluation to be performed at 225T rather than at 2507. This evaluation will include MSRV setpoint considerations and the likelihood that an inadvertent MSRV spurious opening would occur.
As indicated in our October 6,1995 letter, plans for replacing a leaky MSRV before the tailpipe temperature reaches 2757 would be in place and executed appropriately, although, continued plant operation over long periods of time with any MSRV tailpipe temperature above 250T would not be permitted. The leaky MSRV would be replaced during a planned maintenance outage or during a forced outage (i.e., for other reasons), whichever occurs first.
A rapid increase in MSRV tailpipe temperature has not been experienced at LGS and is not expected to occur. Such an occurrence would require a more urgent implementation of a maintenance outage to replace the affected MSRV.
l w
Attachment Docket Nos. 50-352 and 50-353 March 1,1996 L
Page 4 of 6 i
l 4.
The licensee stated that the ECCS pump suction strainer differential proosure acceptance criteria l
will be available by November 1,1995. Please submit the acceptance criteria, including the maximum allowable differential pressure, for staff review.
i t
RESPONSE
Emergency Core Cooling System (ECCS)/ Reactor Core Isolation Cooling (RCIC) pump suction l
strainer differential pressure acceptance criteria has been developed and is provided in the table l
below. At this time, the acceptance criteria has been incorporated into the quarterly pump, l
valve and flow test for each ECCS/RCIC pump as indicated in our October 6,1995 letter and in l
our November 16,1995 response to NRC Bulletin 95-02, " Unexpected Clogging of a Residual l
Heat Removal (RHR) Pump Suction Strainer While Operating in Suppression Pool Cooling l
Mode,' for LGS, Units 1 and 2, with the exception of the High Pressure Coolant injection (HPCI) l pump.
Design Basis Conditions Test Conditions Acceptance Criteria O
T Z,
O T
Max.'
ALERT' (gpm/ pump)
( F)
(feet)
(gpm/ pump)
(*F)
RCIC 600 170 201.47 1600
$105 4.0 3.75 RHR 11,000 212 199.96 110,000 1 95 3.3 1.70 Core 3,950 212 199.96 23,175 5 95 2.9 2.30 Spray Notes: 1 Maximum allowable change in suction pressure between static conditions (with suction aligned to the suppression pool) and Test Conditions. This maximum allowable change includes a maximum pressure drop across the suction strainer at the test conditions such that the pressure drop across the strainer under indicated Design Basis Conditions is not greater than 2.0 psid, as stated in the LGS Updated Final Safety Analysis Report (UFSAR), Section 6.2.2.2.
l 2
Maximum value of the measured change in suction pressure between statia conditions and flow conditions, beyond which engineering evaluation is required.
HPCI pump suction strainer differential pressure testing has been determined to be of little benefit based on the inability to attain full flow conditions in the suppression pool to suppression pool mode. The reduced flowrates (nominally 1000 gpm versus rated flow of 5600 gpm) specified during previous tests are consistent with the system design and minimize the rate of flow accelerated corrosion (FAC) in the HPCI system flush line. The HPCI system flush line is too small (i.e., 4 inches) to pass full HPCI system flow. The low test flowrate results in low acceptance criteria, poor repeatability, and data that is easily distorted due to constant changes l
in suppression pool conditions (i.e., suppression pool level and temperature). Data from the L
tested ECCS/RCIC pump suction strainers will be used as an indicator of HPCI pump suction j
strainer cleanliness.
I I
.m m P
s-Attachment Docket Nos. 50452 and 50-353 March 1,1996 Page 5 of 6 Use of RHR, Core Spray (CS), and RCIC pump suction strainer differential pressure test data as an indicator of HPCI pump suction strainer performance is adequate sirms strainer clogging is accelerated by pump operating time and the HPCI system is normally adgned to the condensate storage tank. As a result, the HPCI system is nominally aligned to the suppressson pool for less than eight (8) hours per cycle. Also, data obtained from tested ECCS/RCIC pump suction strainers will be capable of identifying a change lo suppression pool conditions and possible degradation of the HPCI pump suction strainers. Therefore, as evidenced by recent suppression pool inspections / cleaning on both units, HPCI suction strainers are and should remain in a very clean conditkys throughout normal plant operation.
5.
RHR pumps are used for Suppression Pool cooling when the Suppression Pool temperature increases due to MSRV leakage. Therefore,it appearsthat the RHR pumps are required to run more frequently than originally designed. Describe how the capability and reliability of the RHR
)
pumps are affected due to the frequent Suppressson Pool cooling. Describe the licensee's action plan to assure the capabi&y and reliability of the RHR pumps.
RESPONSE
The RHR pumps which have suppression pool cooling capability (i.e., the 1 A,1B, 2A, 28), have accumulated a greater number of operating hours than the dedicated Low Pressure Cociant injection (LPCI) pumps (i.e., the 10,1D, 20,2D). Reviews of heavy operating periods show the 1 A RHR pump has significantly more operating hours and is used as the bounding example in this discussion.
The operating hours accumulated since initial plant startup on the 1 A RHR pump are estimated at 8000 hours0.0926 days <br />2.222 hours <br />0.0132 weeks <br />0.00304 months <br /> and are approximately 25% greater than the next closest RHR pump.
The LGS UFSAR indicates that the expected operating hours for a 40 year plant life are 31892 hours. This was based on BWR operating experience and engineeringjudgement at the time the plant was designed. Subsequentinformation developedfor the ECCS Motor Qualification Program (GE Document 22A4722) estimates the lifetime operating hours to be greater than
)
61,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />. This information was utilized in the current analysis detailing preventive maintenance under the Reliability Centered Maintenance (RCM) program.
The lifetime operating hours are estimates provided to ensure the life expectancy of the equipment exceeds the estimated required operating time. However, these estimates are not bounding and the component condition is monitored and trended by the following programs:
Quarterly Performance Testing (including inservice Testing)
Vibration Monitoring Ferrography Motor ElectricalTesting Environmental Qualification /RCM (Partial Disassembly)
e-Attachment Docket Nos. 50452 and 50-353 March 1,1996 Page G of 6 A review of these programs indicates no adverse trends. Comparisons were made between the 1 A RHR pump data and the 1D RHR pump data. There are no differences which would indicate that an adverse trend exists. The 1D RHR pump was chosen for comparison becauseits condition is v 0!' documented by a partial disassembly pedaivred in October,1995, and provides a good baseline.
In summary, the 1 A RHR pump / motor is in good condition. Programs are in place to continually monitor anct assess the condition of the RHR pumps and provide indication of adverse trends indicative of fatigue or imminent failure.
_