ML19325D654

AEOD/E908, Excessive Valve Body Erosion at Brunswick, Technical Evaluation Rept

ML19325D654
Person / Time
Site:	Brunswick
Issue date:	09/29/1989
From:	Ellen Brown NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
To:
Shared Package
ML19325D653	List: ML19325D652 ML19325D654
References
TASK-AE, TASK-E908 AEOD-E908, GL-89-13, IEB-88-004, IEB-88-4, IEIN-83-55, IEIN-89-001, IEIN-89-008, IEIN-89-1, IEIN-89-8, NUDOCS 8910250236
Download: ML19325D654 (7)

Text

( :. A f

bpe V  :.,

,s..

'Y Lef,.

j t

'AE03 TECHNICAL REVIEW REPORT j

v 1<

' UNIT:

Brunswick 1 TR REPORT NO : MOD /E908

. +

DOCKET NO.

50-325 DATE: September 29, 1989 LICENSEE:

Carolina Power and Light Company-EVALUATOR / CONTACT:

E. J. Brown 1

MSSS/AE:

. General Electric /'JE&C y,

SUBJECT:

EXCESSIVE VALVE BODY EROSION AT BP.UNSWICK I

EVENT DATE: December 14, 1988-

SUMMARY

n Excessive valve body orosion was discovered in the LPCI injection valve, r

E11-F0178, at Brunswick Unit 1.

The licensee investigation indicated that valve

-throttling was a' major contributor to the erosion. A review of previous AE00 reports, the. Sequence Coding and Search System, cnd NPRDS identified four xorevious AE00 reports that addressed erosion events and more than 200 other erosior.' events. The event data indicate a patterr rf erosion damage and degrMation of components in several safety systems (RHR, HPCI, RCIC, service L

,* y water, and RHR service water). The primary causes appear to be cavitation related to' tither throttling. devices such as valves, orifices, and reducers or

. low flow conditions. An NRC'information notice, IN 89-01, was issued on the

+

valve erosion event and Generic Letter 89-13 was isst.ed on service water system problems.

ItcappearsthatactionbytheBWRownersgroupwill concentrate on " target, valves in>various systems.

HoWyer..it appears that i

lowf flow damage vachanisms for pumps may not be adequa1.ely covered by either

- the BWR owners group action or Generic Letter 89-13 for service water systems.

Also <itt is not clear that industry efforts to monitor PWR plants har alved i

>as anticipated.

DISCUSSION On December.14,-1988, the licensee discovered excessive erosion of the valve c

Lody downstream of the seat during disassembly of the LPCI injection valve, l-

' L' n E11-F017B, for Brur,swick Unit 1.

The valve disassembly was being conducted as part of the plant long term maintenance program. The valve body had erosion l

t.

' pockets with a minimum wall thickness of 1.7 inches. The original wall L

thickness was between 3.5 and 4 inches.

Inspection of the identical valve, E11-017A.on the other'LPCI train revealed a wall thickness of 1.4 inches

. /:,

in the same area of the valve body. Preliminary investigation by the licensee indicated that a major contributor to the erosion of these valve bodies could be

?

the throttling operation of the valves. The licensee discovered erosion in other ' throttle valves in the LPCI system. NRC Information Notice JN 89-001

.(Ref. 1) was issued on th 9 event.

The effort to ascerhin the extent of this problem included a review of previous AE00 reports, a search of the SCSS LER database, and a search of the i

HPRDS equipment failure database. This review identified three engineering

, evaluation (EE) report? that addressed erosion and effects of valve throttling in nuclear plants and one casa study that included erosion as a degradation E

,w

+

m.

vg M102E0236 891018 PDR

<g i,',

ADOCK 05000324 c <1

e. u m;j W ',

J.

o

p',.

7 3;

[

y j Nt-mechanism.. The three EE reports are:

" Misuse of Valve Resulting in Yibration and Pipe Supports," E315 (Ref. 2); " Erosion in Damage to the. Valve Assembly (Ref. 3); arid " Pump Damage Due to Low Flow Cavita-Nuclear Power Plants," E416 V

tion," E807 (Ref. 4). The case study report was:

" Service Weter System Failures

.andDegradationsinLightWaterReactors,C801(Ref.5). The LER search using a

.theSequenceCodingSearchSystem(SCSR) identified 64valveerosionevents f rom 1980 through 1988.' The NPRDS search identified over 140 events covering the period from 1985 to 1983.

AE00 Reports AE00/E315 This report wac an evaluation of two events involving operation of the residual heat removal system (RHR) in the shutdown cooling mode at the Suquehanna Unit 1 plant. The F017B valve (the same valve with erosion at Brunswick) was i

used to throttle flow in the shutdown cooling mode of the RHR system. This valve was observed vibrating severely.

The packing was lost, the position indicator had fallen off, the adjacent pipe hanger had two broken welds, and another ptpe hanger had cracked welds.. Several pipe hanger weld cracks were observed on the other train also.

Four months later the valve disc separated from the stem. The investigation found that-RHR flow for the shutdown cooling 1

mode could be distributed as a minimum 1/4 heat exchanger flow and 3/4 bypass flow. This flou distribution in combination with a very low level of decay heat crebted.a situation in which it was desirable to reduce flow through the RHR t

heat exchanger, but this could only be accomplished by throttling valve F017B.

R The report suggested that an infctmation notice be issued.

IE IN 83-055-(Ref. 6) wts issued on the event.

It was concluded that the severe damage to the LPCI system injection valve, F017.B. was directly related to PHR system flow limitations that result from a combination of system design, configuration, flow control, and the low level of decay heat. The report suggested that NRR review system operation for compatib'lity of ulve assembly design and qualification including frequency of operation and vibration, together with the adequacy of g

^

the shutdown,to'ing mode system flow control.

AE00/E416' This study was an evoluation of more than 140 reports involving erosion in There were 31 reports about valves p(umps, valves, heat exchangers, and piping.a report may include several valve HTGR. The erosion events involved beth steam and water systems. The erosion

problems included wear through the wall of valve bodies, deterioration of rubber seats, and wear of retaining devices. A major finding was that " erosion events ippe'ir related to the specific water source with suspended solids (raw water, radwaste, etc.), the use of throttling devices (valves and orifir.es), or a combination of then effects." This study alco indicated that piping erosion events were related to use of throttling devices such as valves, orifices, and reducers. Approximate % "I percent of the 60 piping erosion reports were related to flow throt...r; Thus, if valve throttling is used, the program to monitor erosion ' +, indude both the valve and downstream piping.

r i

f m

~~~

9 -

[

o r -

4 The report concluded that constructive actions to address erosion problems could include:

(1) cognizar.ce of the phenomenon for certain systems and

' sites, (2) identification of specific plant equipment and physical configurations that may be susceptible to erosion, and (3) implementation of monitoring programs to detect degradation of equipment (pumps, valves, heat l

exchngers, and piping).

AEOD/E807 Erosion and vibration resulted in failure of an emergency service water pump-at Susquehanna Unit 1.

Subsequent disassembly.of the pump revealed that the bottom portion of the pump suction bell had separated from the body and had fallen into the pump pit.

Inspection of the residual heat removal service water pumps revealed cavitation damage similar to that found in the service i

water pumps. The cavitation was caused by flow recirculation due to operating the pumps at low flow rates.

The RHRSW pumps had been operated at less than 50 percent flow most of the time. A search of operatiorial experience database files identified similar pump degradation from low flow operation at Vermont Yankee, H. B. Robinson 2, Turkey Point 3, Haddam Neck, and a foreign plant.

The study concludes that operation of ce'trifugal pumps at low flow conditions for extended periods of time can cause cavitation damage from flow recirculation.

The cavitation erodes the impeller and the internal pump casing wall. The low flow conditiers could occur during aerformance of the inservice surveillance tests by restricting flow through tie mini flow bypass line, or running the pumps in a low flow mode for a system designed for a wide range of flows. NRC Bulletin 88-04 (Ref. 7) addressed the miniflow test configuration. NRC Informa-tion Notice 89-008 (Ref. 8) was issued to highlight pump damage caused by low-flow operation.

AE00/0801 This report was a comprehensive review and evalbation of service water sy* tem failures and degradations observed in LWRs from 1980 to 1987. The causes of failuresanddegraoationsincludevariousfoulingmechanisms(sediment deposition, biofouling, corrosion and erosion, pipe coating failures, calcium carbonate, foreign material and debris intrusion); single failure and other design deficiencies; flooding; multiple equipment failures; personnel and procedural errors; and seismic deficiencies. Of 980 operational events during this period, 276 were deemed to have potential generic safety significance.

A majority (58 percent) of the events involved system fouling. The fouling mechanism included corrosion and erosion (27 percent), biofculing (10 percent),

for+ign macerial and-debris intrusion (10 percent), sediment deposition (9 percent), and pipe coating failure and calcium carbonate deposition (1 percent).

The recommendations from the study were:

(1) Conduct, on a regular basis, performance testing of all heat exchangers l-which are cooled by the service water system and perform a safety function to verify heat exchanger 1" r transfer capability.

L

12) Require licensees to verify that their service water systems are not vulnerable to a single failure of an active component.

l l

l V

1 L s

.,p-t' (3)

Inspect, on'a regular basis, important portions of the oiping of the service water system for corrosior,, erosion, and biofouling.

(4) Reduce human errore in ?te operation, repair and maintenance of the service water' system.

Generic Letteer 89-13 (Ref. 9) was issued July 18,19P9 to address the issues identified in Case Study C801.

1 L.

LER SCSS Search All LERs were searched using the automated database (SCSS) to identify those

- LERs in which the words " erosion" and " valves" Doth appear.

Sixty-four LERs from 1980 through 1988 were identified. About half the events identified by this search occurred after our 1984 study of erosion, E416. These events did not appear to include excessive wear similar to that reported at Brunswick.

i In general,. the data was similar to that reviewed in the E416 study. There were many events with valves that failed leak rate tests due to er,' ion of the disc. This was corrected by lapping. Other reported causes were erosion due to steam cutting, and expansion of a small leak from a gasket "0"

ring, or steam packing.

h

~ NPRDS Search The search strategy for the NPRDS was to identify valves with normal / abnormal

wear, it was further refined with a narrative search for the word " erosion."

There were over 140 reports identified from 1985 through 1988 for all safety classes (1, 2, 3, and 4). The type of reported erosion was similar to that in the AE00 st W es and the recent SCSS review. There were no narrative descrip-tions that !rdicated erosion as extensive as that at Brunswick Unit 1.

However, there were reports that identified other valves in the RHR system with erosion i

damage. Some of the valves were F015B (this valve is in series with F017B),

F024B (RHR test line control valve), 1001-28A and B (outboard isolation valves),

t and 1001-36A (RHR pump test line). Thus, the events reported to NPRDS illustrate erosion patterra similar to those previously identified in the AE00 reports E315, E416, and E807.

Industry Program i

The BWR Own es Group (BWROG) has established an effort to address valve erosion l

due to cavitation induced by valve throttling.

The BWROG objective is to

" Provide a guidance document for individual utilities to establish and implement g'

a valve assessment / inspection program in order to eliminate valve erosion pro-1

. blems." The BWROG program is a multistep process including meeting with NRC staff to present the plan, schedule, and discuss results prior to final disposi-tion for use by licensee; surveying all BWROG participatii9 utilities to identify safety-related valves used for throttling; request / collect data or cavitation erosion; identify methods used by utilities to eliminate erosion i

problems; and prepue a BWROG Guidance Report.

Information about the program was presented to ACRS on April 27, 1989.

l 1

lr l

1_r-

___m

j r

u Based on data reviewed to date, the effort has identified aporoximately 15 or 16 " target" valves in BWR safety-related water systems. Ite appropriate 4

numbert of valves depends on whether the BWR has HPCI or HPCS. The approach i

addresses eight valves in the RHR system, two in the RHR service water system, two in the core spray system, one to three in the HPC1/HPCS systems, and one in the RCIC syttem.

For all syste'ns other than RHR, the " target" valves are test return valves excent for two heat exchanger flew control valves.

AEOD staff have participated with URR during the BWROG presentations. The previously cned AE00 reports E315, E416, and E807 were identified for industry information. The concerns expressed by AE0D were that cavitation induced damage has exhibited a broader scope than simple valve body erosion.

In addition to valve body erosion, the AEOD studies have identified cavitation induced damage as valve essembly degradation due to vibration, piping support damage due to vibration, pipini; erosion downstream of throttled valves, severe degracation of pumps in the service water system and the RHR service water system, and low flow related damage to puaos in other system:.

It is out understanding that the BWROG program will be limited to " target" valve body erosion. Further, it was indicated by the BWROG that pumps will not be addressed ond, if service water pumps are a concern, they should be covered as part of any NRC actior related to service water systems.. In addition, a representative from NUMARC indicated that efforts to address erosion in PWR plants would be revie,,ed with the various owners groups. We are not aware of subsequent action (after April 1989) by any PWR owners group.

t OVERVIEW The AE00 reports have identified a relatively broad scope of erosion related problems involving the RHR system, service water system, and RHR service water system.- The problems include cavitation induced valve body erosion, valve assemoly degradation due to vibra'. ion, piping support damage due to vibration, piping erosion downstream of throttled valves, severe degradation of pumps in the service water system and RHR service water system, low flow related damage to pumps in other systems, and corrosion / erosion fouling that may adversely impact heat transfer capability of sers,e water system heat exchangers or piping rather than pump on valve operability.

NRC generic communications that pertain to these issues include Information Notices83-055, l

89-001 and 89-008;Bulletin 88-04, and Generic Letter 89-13.

It appears that there are two efforts established to address these problem areas.

First, the BWROG has established an effort to cover " target" valves l

in systems such as RHR, HPCI, RCIC, and HPCS as well as.two RHR service water I

system valve. This BWROG effort appears to be a thcrough effort that will L

concentrate on valve erosion and tavitation but will not address other L

components. Presumably, if valve cavitation is reduced, there is tne potential l

for beneficial reduction in vii, ration problems with piping supports and valve e.ssembly operation.

l The second effort involves licensee response to Generic Letter 89-13 pertaining 1:

to service water systems.

As previously indicated, the corrosion / erosion aspect of.this effort appears to emphasize fouling that may impact heat transfer capability rather than degradation of components such as pumps and valvts. Thus, it would seem that low flow pump erosion may not receive appropriate monitoring to address the concerns identified in E807 (Ref. 4).

1 L

l r.

l.

o.

I>

f c

L.

p.

i L-FINDINGS 1.

AE0D studies have identified a broad scope of cavitation r lated component i

damage. These include valve body erosion, valve assembly cegradation due to vibration. piping support damage due to vibration, piping erosion c

downstream of throttled valves, severe degradation of pumps in the service t

water system and RHR service water system, low flow damage to pumps in other systems, and service water system failures and degradations involving fouling mechanisms that include corrosion and erosion.

2.

The conclusions, suggested actions, and recommendations in AE00 reports E315, E416, E807, and C801 appear to be reasonable approaches to minimize effects of this damage in the future.

3.

NRC has issued several generic comunicetions addressing these erosion issues. These documents are Information Notices83-055, 89-001 and 89-008;Bulletin 88-04; and Generic Letter 89-13.

4.

The EWROG effort appears to be a thorough effort pertaining to erosion of

" target"' valves.

NRR is following this effort. We understand the effort will not include monitoring of pumps.

5.

It does not appear that any formal effort has been establisheo by PWR owners groups to review these erosion issues.

6.

The damage mechanism to centrifugal pumps under icw flow conditions 1dentified in E807 (Ref. 4) does not appear to be appropriately addressed by either the BWROG action or Generic Letter 89-13.

CONCLUSIONS The event data from 1980 through 1988 exhibit a consistent pattern of erosion damage and degradation of components in several safety related systems. The primary causes appear to be cavitation related to either throttling devices such as valves, orifices, and reducers or low flow conditions. This results in either long term erosion wear or a more immediate destructive vibr6 tion problem.

However, the root causes lecding to erosion or cavitation damage appear related to system flow limitations with special concern at low flows relative to full system capacity. A combination of system design, configuration, flow control and low flow requirements conspire to cause equipment damage. The conclusions, suggested actions and recomraendetions in AE00 reports E315, E416, E807, C801, and the NRR Generic Letter 89-?') appear useful to monitor damage t.nd degradation due to erosion.

Howevet,.t appears that flow damage mechanisms for pumps may not be adequately addressed by either the BWROG or GL 89-13.

Also, it is not clear that industry efforts to monitor PWR plants has evolved as anticipated.

w REFERENCES 1.

U.S. Nuclear Regulatory Commission, NRC Information Notice No. 89-001, Valve Body Erosion," January 4,1989.

2.

U.S. Nuclear Regulatory Commission, E. J. Brown, " Misuse of Valve Resulting in Vibration and Damage to the Valve Assembly and Pipe Supports, "AE00/E315, July 7, 1983.

l,.

ci ;

-7 l

s f

3.

U.S. Nuclear Regulatory Conu.ission, E. J. Brown, " Erosion in Nuclear Power Plants," AE0D/E416, June 11, 1984 L

4.

U.S. Nuclear Regulatory Commission, C. Hsu, " Pump Dmage Due to low Flow L

Cavitation," AE00/E807, October 18, 1988.

6.

U.S. Nuclar Regulatory Coinmission, P. Lam and E. I.eeds, "Servire Water System Failures and Degradations n Light '..ater Reactors," AE0D/Ca01, August 1988.

6.

U.S. Nur lear Regulatory Connission, IE Information Notice 83-055, "Misappiication of Valves by. Throttling Beyond Design Range," August 22, 1983.

7.

U.S. Nuclear Regulatory Cornission, NRC Bulletin No. 88-04, " Potential Safety-Related Pump Loss," Hay 5,1988.

8.

U.S. Nuclear Regulatory Consnission, NRC Information Nc.89-008, " Pump Damage Caused by Low-Flow Operations," January 26, 1989.

9.

U.S. Nuclear Regulatory Connission, J. G. Partlow to All I.icensees, " Service

-Water System Problems Affecting Safety-Related Equipment (Generic Letter 89-13)." July _18,1989.

1 1

1.

L I

i l

I j

L