ML20151H266
| ML20151H266 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 07/18/1988 |
| From: | Cockfield D PORTLAND GENERAL ELECTRIC CO. |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| IEB-88-004, IEB-88-4, NUDOCS 8808010207 | |
| Download: ML20151H266 (4) | |
Text
.
M M
Poftland Gonoral Electric Conixwly l
E
~
David W. Cockheid Vice President, Nuclear July 18, 1988 Trojan Nuclear plant Docket 50-344 l
License NPF-1 l
U.S. Nuclear Regulatory Commission ATTN:
Document Control Dosk Washington DC 20555
Dear Sirs:
NRC Bulletin No. 88-04 Lotential Safety-Rela _ted pump 1.oy This response to Bulletin 88-04 sumarizes the initial actions taken by Portland Ceneral Electric to ensure adequate minimum flow for safety-related pumps. The Bulletin identified im.ediato and longer-term actions required to address minimum flow due to:
(1) the possibility of pump-to-pump interaction and (2) the adequacy of the current miniflow capacity.
These items are discussed in the attachment to this letter.
This response does not completely address all of the Bulletin items at this time because pump vendor information on minimum flow has not been received.
An updated response will be provided following receipt of the required information in order to completely respond to bolletin 88-04.
Sincerely.
Attachment I
et Mr. John B. Martin Regional Administrator, Region V U.S. Nuclear Regulatory Comission Mr. Bill Dixon State of Oregon Department of Energy T(/
Mr. R. C. Barr n\\{
NRC Resident inspector Trojan Nuclear Plant 1
@s Subscribed and sworn to before me this 18th day of July 1988.
0800010207 000710 h[, 4
/%
DR ADOCK 050 34 l
Notary public of Ot egon My Comission Expires:
c 4tw I / 'Yb Qs a
-, c, s *Wm 1r s
~;a
Trojan Nuclear Plant Document Control Desk Docket 50-344 Attachment Licensed NPF-1 July 18, 1988 Page 1 of 3
SUMMARY
OF EVALUATION OF NRC BULLET! _88-04 The following response presents the initial assessment of the actione requested by NRC Bulletin 88-04.
It is not possible to completely address each of the "Actions Requested" items at this time because requested information from pump vendors has not been received. Following receipt and evaluation of purp minimum flow recommendations, an upd6ted response will be provided.
A.
AFFECTED SYSTEMS AND ISSUES Safety-related centrifugal pumps with minimum flow recirculation paths were evaluated for minimum flow operation and pump-to-pump inter-action.
Because vpdated pump vendor recommendations were not available to incorpornte into this responso, it cannot be concluded that the existing vendor-specified minimum flow rates are acceptable. This situation will be reevaluated when further information from the pump vendors is available. Conclusions regarding the potential for pump-to-pump interaction under mininum flow conditions are discussed below.
1.
Auxiliary Feedwater Purps.
The auxiliary feedwater pumps have separate discharge headers and miniflow paths that are protected against backflow by check vulves, thus the minirum flow paths remain independent to the common return line to the Condensate Storage Tank.
2.
Esfety injection purps. Centrifugal chartba purfe.s The centrifugal charging pumps and safety injection pumps miniflow design has flow paths that are train-independent with a check valve separating the pump discharges, and separate high-pressure drop orifices in each line, prior to a common miniflow discharge line.
Given the high-discharge pressure of these pumps, the high-pressure drop orifices mitigate backpressure effects which could cause pump-to-purp interactions which would affect the minimum flow.
3.
Containment Spray purps.
The Containment spray pumps have separate headers and no miniflow under accident conditions, and are tested only under minimum flow conditions. Thus, the minimum flow considerations do not apply for accident conditions for these pumps.
I l
1 l
9 Trojan Nuclear plant Document Control Desk Docket 50-344 Attachment Licensed NpF-1 July 18, 3988 l
page 2 of 3 4.
Residual Heat Removal UHR) Fungg.
f i
The RHR pumps have a minifluw design with individual flow paths for i
each pump which cpen automatically on a low-flow signal (less than I
S00 spm). The mintflow valves automatically close ou a high-flow signal with flow greatse than 1.000 gpm. However. the RHR pump discharge and miniflow paths share a comon header suck that pump-to-pump interaction during miniflow operation is not prectuded.
l 5.
Boric Acid Transfer pumps.
2 i
{
The boric acid transfer pumps are operated and tested with a I
recirculation flow path that is manually aligned. A potential i
4 I
comon header flow path is not aligned.
The pumps do not receive i
)
an automatic start signal and are under manual control with one pump on recirculation and the other pump in standby. This pre-cludes pump interaction under miniflow conditions.
t l
B.
SHORT-AND LONG-TERM CORRECTIVE ACTIONS h
1 Minimum Rocirculation Flow.
I To address short-terio considerations about the adequacy of pump minimum i
flow rates, tne above-described pumps were reviewed and found to be in 1
conformance with pump vendor recommendations. As part of the long-tem l
)
corrective actions, the manufacturers of the affected pumps have been I
requested to update their recommendations for the mininum flow require-l l
ments. This informatinn will be evaluated when it is received to
[
determine if modifications are needed, and an updated response to this t
Bulletin will be provided.
[
pump interaction.
The RHR pumps were examined further to detemine whether the strong 1
pump / weak pump interaction is occurring at Trojan.
The most recent s
test data at miniflow conditions show that the
'A' RHR pump has a
[
pressure differential of 172.5 psi at 630 gpm and the
'B' pump has a I
pressure differential of 173.5 psi at 600 gpm. Had the flows been j
closer to the same value, it is expected that the two pressure dif-i forential values would have been even closer to one another. These
[
data appear typical since a review of past data show only small t
variations; this confirms the recent measured values. Thus, based on strong puep/ weak pump interaction is not occur-f current indications. '
j ring at this point in 'im a.
snort-tem action will be taken by
[
September 30, 1988 to rwise the pariodic operating test that monitors I
pump discharge pressure on a quarterly bar,is to require evaluation of I
pump differential pressure data to ensure that unacceptable pump-to-q putp interaction is not occurring.
I i
t t
\\
Trojan Nuclear Plant Document Control Desk Docket 50-344 Attachment Licensed NPF-1 July 18, 1988 Page 3 of 3 Long-term action is under consideration and will be further discussed in the updated response to this Bulletin.
Since pump wear occurs slowly under normal operating conditions, degradation in pump operating characteristics will be detected by the quarterly t6ats before it l
becomes significant.
L C.
SCHED'1LE FOR LONC-TERM RESOLUT10W The long-term resolution of this issue will be discussed in an update to this response, j
D.
JUSTIFICATION FOR CONTINUED OPERATION i
Continued operation with the potential for strong pump / weak pump inter-action in the RHR System has been evaluated.
The probability of a l
strong pump / weak pump interaction is extremely low, primarily because l
(1) the pumps are tested quarterly and past data have not indicated any i
significant changes in pump characteristics (2) the pumps are being operated under miniflow conditions in conformance with the manufac-turer's recommended miniflow criteria, and (3) the degradation of one
)
of the pumps leading to strong pump / weak pump interaction is expected to occur over a sufficiently long period of time to allow its identi-fication during periodic testing.
In addition, the type of event where the minimum flow interaction could take place is limited to small-break Loss-of-Coolant Accidents (LOCAs) where the RHR pumps are needed for long-term recirculation cooling of the ec.re.
Quality materials, l
construction, inspection, and leak-testing of the Reactor Coolant l
System provide reasonable assurance that a small-break LOCA is a low l
proba-bility event.
The nature of the pump-to-pump interaction is such that the degrae*d pump would not suddenly lose all minimum flow capability.
Rather, the minimum flow would decrease below the 500 Epm recommended value.
Eval-untion of the Tuckey Point RHR pump with no flow showed 10 minutes of operation was permissible. The Trojan RHR pumps are very similar (Ingersoll-Rand 8X20WD) to the Turkey Point RHR pumps (Ingersoll-Rand 8120W) and thus, pump deadheading would not cause prompt pump damage.
Furthermore, a small amount of minimum flow would extend the period l
before pump damage occurred.
Overall, it is judged that continued operation is acceptable based on the use of existing vendor ninimum flow criteria and because the current test data show that pump interactions that would lead to pump degradation are not occurring at Trojan.
TDW/DRS/CRA/mr 2111P
]