ML20086P031
| ML20086P031 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs  |
| Issue date: | 12/17/1991 |
| From: | Creel G BALTIMORE GAS & ELECTRIC CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9112260144 | |
| Download: ML20086P031 (5) | |
Text
7 f
BALTIMORC
_3(
. GAS"AND ELECTRl
^ 1650 CALVERT CLIFFS PARKWAY
- LUSBY, MARYLAND 20657 4702 Gconot C.' CRt'EL~~
vics PatstotNT NUCLE Ah ENEROV (4io) eso-4 ass December 17,1991 U. S. Nuclear Regulatory Commission Washington,DC 20555
. ATTENTION: =
Document Control Desk
SUBJECT:
Calvert Cliffs Nuclear Power Plant Unit Nos.1 & 2; Docket Nos. 50 317 & 50-318 Comments Concerning AEOD Draft Report, Experience with Byron Jackson
.' Pump Seals
REFERENCE:
(a)
Letter from Mr. T. M. Novak (NRC) to Mr. R. E. Denton (BG&E),
dated September 27,1991, same subject Gentlemen:
As requested in the referenced letter, we are providing comments concerning the data contained in the draft, AEOD report. The NRC draft AEOD Special Report, " Experience with Pump Seals Installed in Reactor Coolant Pumps Manufactured by Byron Jackson," examines Reactor Coolant
- Pump (RCP) seal operating expenence in Byron Jackson pumps through August 1990, It is intended 1to update industry operating experience in RCP seals for the period since the proposed resolution to Generic Issue (GI) 23 was published, and to serve as an aid in coming to resolution of GI 23.
-Reactor Coolant Pump seal failures, invoMng one or more seal stages, have been a significant contributor to plant down-time. More importantly, a gross seal failure can lead to a small break loss of coolant accident (LOCA). The AEOD report concludes that, although industry experience with RCP seal failures has dramatically improved since 1980, the frequency of RCP seal failures since--
1985 is considerably higher in Byron Jackson pumps than in pumps manufactured by other vendors.
This is based upon a review of NPRDS data, as well as site visits to Calvert Cliffs, Waterford 3, and Arkansas Nuclear One plants.
Attachment (1) represents Baltimore Gas and Electric Company's comments on the draft AEOD report. These comments include editorial remarks which are solely intended to ensure the report is -
. properly balanced, and provide additional technical information which was missing from the report.
hS1Y
- I 00 gg..Afod I
L nnn.
- 9112260144 911217 PDR ADOCK 05000317 S
)
r Document Control Desk December 17,1991 P. age 2 g
Should you have any questions regarding this matter, we will be pleased to discuss them with you.
Very trulyyours, GCC/ PSF / dim
[
}
Attachment cc:
D. A. Brune, Esquire J. E. Silberg. Esquire i
R. A. Capra, NRC D. G. Mcdonald, Jr., NRC T. T. Martin, NRC L E. Nicholson, NRC R. I. mci.can, DNR J. H. Walter, PSC
a A'ITACIIMENT (1) llattimore Gas and Electric Company's Comments Concerning AEOD Draft Report, Experience with Ilyrtm Jackson Pump Seals The following specific comments are provided:
Executive Summary: It states: "... loss of CCW, even for very short periods, is suf6cient to cause seal damage." _ While the loss of component cooling water (CCW) is extremely critical for SU seals, extensive testing in the new Sulzer Bingham seal has proven that it is less sensitive to this event. The tests showed that the SBPI seal can operate for over 30 minutes without CCW and have no appreciable increase in seal leakage.
ExecuF ' Summary; The summary infers that the SBPI seal is not trouble-free, it should be noted u... the two SBPI seal failures could not be attributed to the seal design. In one case the problem was traced to a maintenance error. The other case, a seal failure on San Onofre 2, was a result of the pump's rotating bafDe becoming diskxiged and blocking the controlled bleed-off tubing. We believe that the Sulzer Bingham seal has many excellent features which will imprcwe our seal reliability record. It is a much simpler design and easier to maintain, thus reducing the probability of maintenance errors. It has also been designed to withs:and plant transients which, in the past, have caused seal degradation or failures.
Report. Section 1.0: It should be noted that the majority of Byron Jackson RCP sei urcs have occurred at Babcock & Wilcox (B&W) plants and not Combustion Engineerii.g (CE) plants. In fact, CE and B&W plants have experienced even fewer gross seal failures than Westinghouse plants.
Report. Section 1.1: It states that licensees have " frequently" operated plants with one or more failed seal stages. We consider operating in this manner a highly unusual condition, and do not make it a practice to operate with degraded seals on a " frequent" basis. While our procedures w:ll allow us to do so, every situation is carefully analyzed by Operations and Plant Engineering.
Report. Section 2.0_; Table 1 shows a summary of gross RCP seal leakage events since January 1974. Although not shown, four of the plants are Westinghouse plants and two are B&W plants. Arkansas 2 is the only CE plant listed with a " gross" seal failure (8/88 40 gpm).
Report. Section 2.1: The last paragraph is very important to note. It states that the term
" failure" is defined as the degradation or loss of at least one stage of a multi-stage seal. The same paragraph continues and states that seal failures resulting in external leakage have occurred only twice in Byron Jackson RCPs. The ANO Unit 1 (B&W) failure resulted from operating a three-stage SU seal in a degraded state. The seal had been subjected to at least two significant plant transients, and after it was identified that the seal had degraded, the plant continued to operate. The other external leakage event, ANO Unit 2 (CE), was a result of a sensing line failure leading to progressive middle / upper / vapor seal stage failures.
Although it could be argued that the Byron Jackson SU seal has not been a very reliable design, its reliability record should not be confused with its safety record on " gross"/ external seal leakage. The four-stage scal's overall factor of safety / redundancy, accomplished by having three pressure breakdown seals rated for full Reactor Coolant System pressure with an additional vapor seal, is greater than a Westinghouse seal with one pressure breakdown seal, a redundant prmure breakdown seal, and a vapor seal.
I
r A'ITACIIMENT (1)
Baltimore Gas and Electric Company's Comments Concerning AEOD Draft Report, Experience with Byron Jackson Pump Seals Renort. Section 2.1 Section 2.2 discusses SU seal experience and concludes that the majority of seal failures can be attributed to maintenance errors. We believe that the seal's maintenance difficulties can be traced to inherent SU design characteristics, such as the use of "U-cups,' which in the past have made the seal difficult to maintain. Since seal manufacturers havs improved their designs, we expect better performance industry-wide.
Renort. Section 21 Section 2.5.1 discusses plant-specific experiences with the Sulzer Bingham seal. The report states that at San Onofre 2, the SBPI seal experienced staging difficulties at startap. Based upon conver.;ations with Sulzer Bingham, they indicate that the staging difficulties have occurred at startup and at other operating conditions as well. The only correlation San Onofre's engineers can find is in the variation of the ultimate heat sink temperature, ivhich affect: CCW temperatures. Perturbation in staging pressures have occurred with simultareous fluctuations in CCW temperatures. This anomaly is still being -
investigated. The other seal-related problem, as discussed earlier, involved the loss of con'.ralled bleed off (CBO) flow through the seal as a result of all six bolts, which hold the pump's rotating baffic, fracturing and then blocking the CPO pressure breakdown coils. It should be noted that rone of these failures were a direct result of the seal's design.
Renort. Section 3.3: Section 3.3.1 comments on RCP seal instrumentation and brings up some good points. The report states that operators frequently rely an CBO temperatures as the most dependable indication of seal performance. Although CBC temperature is an important indicator, it is suggested that operators rely more on CBO flow rates as a better indicator. The report also recognizes the difficulty in instrumenting the relatively low flow rates inherent with controlled bleed off seals. At Calvert Cliffs, we have installed new CBO flow meters. These flowmeters, manufactured by FCI, use thermal dispersion technology, which is presently one of the only viable means to accurately measure CBO flow rates. This information is provided to control room operators, and together with seal staging pressures and CBO temperatures, allows operators to accurately assess seal performance.
Table 5 Data Table 5 indicates that information on operating days was not available. Based upon the best available information the following input is provided (Plant 7 is CCNPP Unit 1):
Days Seal Discovery Operating Plant Type Date (Approx.)
Comments Plant 7 SU 07/30/85
< 30 Leaking seal forced shutdown.
Plant 7 SU 10/09/85
< 100 Leaking seal forced shutdown.
Plant 7 SU 10/31/88 123 Continued operating with one failed seal stage.
Plant 7 SU 03/02/89 120 Continued operating until second stage failed.
l l
1 l.
2
3 A'ITACitMENT 11)
Baltimore Gas and Electric Cmnpany's Comments Concerning AEOD Draft Report, Experience with Byron Jackson Pump Seals
SUMMARY
We generally agree with the report's conclusion that Byron Jackson RCP "SU* seals have not been very reliable. Improving RCP seal reliability was the main justification for replacing our SU seals with Sulzer Bingham seals. However, since SU seals have been involved in only two " gross" seal failures to date, we feel that there is an important distinction between reliability and their safety record on gross seal failures.
The report also raises some concern about utilities " frequently" operating four-stage seals with one failed seal stage. While this situation represents an unusual operating condition and deserves careful analysis to justify continued operat!on, the report is unclear about whether future restrictions will be considered. Byron Jackson and Sulzer Bingham seals are designed with greater redundancy than Westinghouse seals; therefore, if future restrictions are pursued, these differences should be accounted for.
l r
i, 3
. ~..
.