ML20154F305
| ML20154F305 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 09/08/1988 |
| From: | Shelton D TOLEDO EDISON CO. |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| 1-823, IEB-88-004, IEB-88-4, NUDOCS 8809190344 | |
| Download: ML20154F305 (10) | |
Text
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .
TDLEDO
% sum EDISON A Ca.enar Erew Creary DONALD C. SHELTON w.r,w.4,=
M'9] 249 2300 Docket No. 50-346 License Fo. NPF-3 Serial No. 1-823 September 8, 1988 United States Nuclear Regulatory Commission Document Coutrol Desk Vashington, D. C. 20555
Subject:
Response to NRC Bulletin No. 88-04: Potential Safety-Related Pump Loss Gentlemen On May 12, 1988, Toledo Edison (TE) received NRC Bulletin No. 88-04 entitled "Potential Safety-Related Pump Loss", (Log No. 1-1PJ3). This Bulletin requests that licensees provide a vritten response, within 60 days of receipt, that summarizes the investigation of the concerns discussed in the bulletin.
However, upon learning that vendor information could not be received in time to permit TE to meet the original due date, Mr. A. V. DeAgazio, the NRC/NRP.
Davis-Besse Project Manager, agreed to extend the due date.
NRC Bulletin 88-04 identifies two miniflow design concerns that each licensee is required to investigate and take corrective actions for, if applicable.
The first concern involves the potential for one or more safety-related pumps i (in systems that have a common miniflow configuration) to be(cme "dead-headed" as a result of puso-to-pump interaction during miniflow operation. The second I concern is whether current miniflov line capacities for safety-related puups are adequate for single pump operation.
TE identified seven systems that contain pumps that could be affected by the concerns addressed in the bulletin. These systems are: Auxiliary Teodwater (AFV), Motor Driven Feedvater (MDTV), Service Vater (SV), Containment Spray (CS), Component Cooling Vater (CCV) Decay Heat /Lov Pressure Injection (DH/LPI), and High Pressure Injection (HPI). Each of these systems was evaluated for the two concerns presented in the bulletin. The attachment contains a summary of the evaluation process, as well as details concerning the calculations, tests, and reviews conducted.
8809190344 080900 PDR ADOCK 05000346 G PNV lif THE TOLEDO EDSON COMPANY EDSON PLAZA 300 MADSON AVENUE TOLEDO. OH'O 43652
-- - - - i
Dockot No. 50-346
- l. Lic;nso No. NPF-3 l Serial No. 1-823
( Page 2 Toledo Edison has concluded that Devis-Besse does not have any cafety-related systems in which a pump or pumps vill become dead-headed as.a result of pump-to-pump interactions during miniflow operation. Verification of the adequacy of minimum flow bypass lines has been confirmed for all but the DH/LPI and HPI systems. Tg is currently developing a test plan to verify the adequacy of miniflow for these two systems. Testing vill be cumpleted as soon as practicable following restart from the current refueling outage. A final report vill be submitted by December 15, 1988.
Very trul ours, W
CFH/ tit Attachment cci A. B. Davis, Regional Administrator A. V. DeAgazio, Project Manager DB-1 NRC Resident Inspector i
RESPONSE TO NRC BULLETIN 88-04 POR DAVIS-BESSE NUCLEAR POWER STATJON UNIT !Ks.1 This letter is submitted in conformance with the Atcaic Energy Act of 1954 Section 182a, in response to NRC Bulletin 88-04, "Potential Safety-Related Pump Loss."
,,, /l aLA D.'C. Shelton, Vice President, Nuclear Sworn and subscribed before me this 8th day of September,1988.
Alf Y$ / 4 otary Public, State of Ohio Wom4 HLASCH w w state 04 ONo g g_ y ,ge.eJ.no W.IMS I ~
Dockot No. 50-346 Licenso No. NPF-3 Serici Ns. 1-823 .
Attachment l Page 1 RESP 0 HSE TO NRC BULLETIN NO. 88-04 NRC Item 1 Promptly determine whether or not your facility has any safety-related system '
with a pump and piping system configuration that does not p eclude pump-to-pump interaction during miniflov operation and could therefore result in dead-heading of one or more of the pumps.
TE Response Seven systems were identified that contain pumps that could be affected by the '
concerns identified in the ballatin. These systems are: Auxiliary Feedvater (
(AFV), Motor Driven Feedvater (MDFV), Service Vater (SV), Containment Spray (CS), Component Cooling Vater (CCV), Decay Heat /Lov Pressure Injection (DH/LPI), and High Pressure Injection (HPI). The potential for pump-to-pump j interaction during miniflow operation exists for the AFV, MDFV, and SV systems.
NRC Item 2
- If the situation described in Item 1 exists, evaluate the system for flov division taking into consideration (a) the actual line and component resistance for the as-built configuration of the identified systems (b) the head versus flow characteristics of the installed pumps, including cetual test data for "strong" and "veak" pump flows
- (c) the effect of test instrument i error and reading errors and (d) the vorst case allevances for deviation of pump test parameters as allowed by the American Society of Mechanical 4 Engineers Boiler and Pressure Vessei Code (ASME Code)Section XI, Paragraph IVP-3100.
]
TE Response TE evaluated the three systems identified above by performing calculations, 4
based on certified pump performance curves, to determine the approximate
- flovrates through the systems for various pump combinations and conditions.
1 The results of these calculations were then compared against the pump vendor's confirmed minimum flovrates. A summary of the results of this ef fort is provided in Tables 1, 2, and 3.
i A review of these tables shows that there is no case in which tne calculated flovrate is less than the flovrate considered adequate by the applicable pump vendor. Therefore, TE concludes that "dead-heading" or one or more pumps as a result of pump-to-pump interaction is not a concern at Davis-Besse (DB-1).
NRC Iter 3 i Evaluate the adequacy of the minimum flov bypass lines for safety-related centrifugal pumps with respect to damage resulting from operation and testing in the minimum flow mode. This evaluation should include consideration of the effects of cumulative operating hours in the minitum flov mode over the lifetime of the plant and during the postulated a cident scenatio involving j the largest time spent in this mode. The evaluation should be based on best
- = - . , . . _ _ - _ _ _
Docket No. 50-346 Liecnso No. NPF-3 1
- Sarial Ns. 1-823 Attachment Page 2 current estimates of potential pump damage from operatior of the specific pump models involved, derived from pertinent test data and field experienco on pump damage. The evaluation should also include verification from the pump suppliers that current miniflow rates (or any proposed modifications to j miniflow systems) are sufficient to ensure that there vill be no pump damage from lov flow operation. If the test data do not justify the existing capacity of the bypass lines (e.g., if the data does not come from flows
' comparable to the current capacity) or if the pump supplier does not verify the adequacy of the current miniflow capacity, the licensee should provide a pisn to obtain additienal test data and/or modify the miniflow capacity as needed.
TE Response The adequacy of miniflow rates was evaluated by requesting current miniflov rate information f' ton each pump vendor and comparing that information against the lovest normal, emergency, and test flovrate for each of the seven systems.
l Table 4 presents the vendor confirmed miniflow rate and the lovest normal flovrate for each of the seven systems. The CCV Pump vendor (Goulds Pumps) recommended a miniflow rate that was greater than the lowest flovrate for a normal mode of operation. The system lineup that results in the lovest i flovrate for a CCV pump (i.e., 1350 gpm) consists of the Emergency Diesel Generator Cooling Vater Heat Exchanger, the Containment Gas Analyzer, the
- DH/LPI Pump Bearing Housing Cooler, and the HPI Pump Bearing Oil Cooler. This lineup only occurs on the train dedicated to the essential header during
- 1) a Loss of Offsite Pover, 2) an SFAS Level 2 actuation without an SFAS Level 3
)
' actuation, or 3) testing of the Emergency Diesel Generator.
]
l Vhile Goulds Pumps recommended 3000 gpm as the minimum recirculation flov for l the CCV pumps, they qualified this recommendation by stating that lover
{
flovrates vould be acceptable for "intermittent operation", provided there was
' no excessive pump vibration at that flovrate. As part of DB-l's Inservice Test (IST) Program, vibration measurements are conducted quarterly on all i
pumps, including the three CCV pumps, but not at lov flow conditions.
I Therefore, a test of CCV Pump No. 3 was conducted in which the flow was i reduced in steps and vibration measurements for the pump vere taken at the i intermediate and the lovest flovrate. (Note: CCV Pump No. I and 2 vere not available for testing due to the outages they vill be tested later.) A
- l. comparison of CCV Pump No. 3 vibration measurements with those taken quarterly l
indicates that there is no appreciable increase in pump vibration at the lov i 21ovrate. A 'er the test, the vendor was again contacted, given the test results, and a ked to clarify the term "intermittent operation". The >endor 4
stated that operation for up to four hours per month at 1350 spa vould be l acceptable. To ensure that the four-heur limit is not exceeded. TE plans to i revise the CCV System Operating and 'lara Procedure prior to restart from the current refueling outage.
In addition to the test conducted on CCV Pump No. 3, a review of maintenance i records for all thtee CCV pumps was conducted. This review indicates that l during the more than 100,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> that these pumps have operated at various l
l l
i
Dockst N3. 50-346 Lic9nso No. NPF-3 S3 rial No. 1-823 Attachment Page 3 flovrates (the number of hours at miniflow condition is not knovn), there has been no history of bearing o; seal failures that could be attributed to vibration problems caused by impeller recirculation.
The vendor for the DH/LPI and HPI pumps (Hayward Tyler) stated that, while they had no definitive data that vould raise doubts over Davis-Besse's current miniflows, they were unable to confirm that these flows were adequate to ensure that the DH/LPI and HPI pumps vould not experience degradation as a result of impeller recirculation. The vendor recommended that a pump test to verify enduratice be conducted under actual of niflow conditions or, alternatively, (tor those pumps known to have operated at miniflow conditions), that service history data be examined.
TE examined performance test data for a seven year period .or the pumps in both systems and did not identify any abnorual pump performance or degradation that could be attributed to operation at minimum flovrates. This finding is not conclusive, however, as the pumps in DH/*PI and HPI systems have experienced very little run time at miniflow conditions. Therefore, TE is currently evaluating the vendor's recommendation and developing a test plan designed to obtain vibration measurements after stabilising DH/LPI and HPI pump operation at minimum flovrates.
Upon development of the tert plan, details concerning the plan vill be discussed with the NRC. Testing vill be conducted as soon as practicable following restart from the current refueling outage. TE vill submit a final report on this subject by December 15, 1988.
s Dock:t No. 50-346 Lic9nsa No. NFF-3 S:rici No. 1-823 Attachment Page 4 TABLE 1
SUMMARY
OF THE EVALUATION OF PUMP-TO-PUMP INTERACTIONS FOR THE AFV PUMPS CAI,CULATIONS CALCULATED spa (Pump /No.)
Case 1: Veak Pump 6 rated speed (3600 rpm) 242 (AF/1-1)
Case 2: Strong and weak pumps in parallel 239 (AP/1-2) d rated speed 236 (AF/1-1)
Case 3: Veak pump 6 rated speed (degraded 12%) 225 (AF/1-1) in parallel with strong pump e 4500 rpm Case 4: Veak pump 9 rated speed (degraded 12%) 225 (AF/1-1) and in parallel with MDFV pump (strongest)
Case 5: Venk pump 9 1100 rpm 72 (AF/1-1)"'
Case 6: Veak pump 9 1100 rps (degraded 12%) 64 (AF/1-1)"'
and in sarallel with MDFV pump Vendor Confirmed Miniflow Rates (AF) 225 spe 6 3600 rpm 64 spa 9 1100 rpm
- a. Operation of an AFV pump at 1100 rps is not required by normal, transient, or emergency procedures.
Dock 3t No. 50-346 Lic;nsa No. NPF-3 ,
Serial No. 1-823 :
Attachment Page 5 ,
f TABLE 2
SUMMARY
OF THE EVALUATION OF PUMP-TO-PUMP INTERACTIONS FOR THE MDFV PUMP CALCtlLATIONS CALCULATED spa Case la MDFV pump (degraded 12%) recirculating 216 i to the Condensate Storage Tank (CST) :
Case 2 MDFV Pump (degraded 12%) recirculating >216 Note: The flow i to the Dearator Storage Tank (DST) path to the DST is less restrictive.
i !
Case 3: MDFV Pump (degraded 12%) in parallel vith 213 (MDFV) {
. the strong AF pump operating 04500 rpm a
c Vendor Confirmed Miniflow Rate 180 gpa 4
}
4 l
}
l 1
D akot No. 50-346 ,
.Lic;nsJ No. NPF-3 Serici No. 1-823 Attachment i Page 6 !
TABLE 3 1
SUMMARY
OF THE EVALUATION OF PUMP-TO-PUMP INTERACTIONS FOR THE SV PUMPS CALCULATIONS CALCULATED spa (Pump /No.)
Case 1: Veak Pump 1202 (SV/1-2)
Case 2: Venk Pump (degraded 12%) 1129 (SV/1-2)
Case 3: Veak Pump in parallel with 1145 (SV/1-2) ,
strong pump 1168 (SV/1-1)
Case 4: Veak Pump (degraded 12%) in 1070 (SV/1-2) parallel with strong pump 1173 (SV/1-1) {
f Vendor Confirmed l Miniflow Rate (SV) t t
1000 spa l
[
i i
i l
I L
i k
l
,..- Dockot No. 50-346
. Licens3 No. NPF-3 l
. Striol No. 1-823 l Attechnent j Page 7 l
l TABLE 4 1
VF.NDOR CONFIRMED MINIFLOV RATE VERSUS THE LOVEST FLOVRATE BY SYSTEM l l
l VENDOR CONFIRMED LOVEST DB-1 !
SYSTEMS HINIFLOV RATE (spa) FLOVRATE (spa)"'
AFV 225 (93600 rpm) 225 64b*
64 (91100 rpm) l MDFV 180 213 ;
SV 1000 1070 CS 750 1175 CCW 3000 (1350 for four hours / month) 1350 DH/LPI 85** 96 HPI 35C ' 53
- a. The lovest flovrate is the lovest rate for any mode of operation (i.e.,
normal, emergency, or testing) including parallel operation of pumps.
- b. Operation of an AFV pump at 1100 rpm is not required by normal, transient, or emergency procedures.
- c. This is the miniflow rate recommended by the original pump vendort this number was not confirmed by the new pump vendor.
,