ML060960541
| ML060960541 | |
| Person / Time | |
|---|---|
| Site: | Braidwood  |
| Issue date: | 11/06/2000 |
| From: | Exelon Generation Co |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| FOIA/PA-2006-0115 | |
| Download: ML060960541 (19) | |
Text
oI TITLE:
Circ Water Blowdown Line Vacuum Breaker failure due to low stress, high cycle fatigue, resulting in flooding of Owner Controlled property and discharge outside of NPDES approved path.
UNIT:
Braidwood Station Unit Common EVENT DATE:
11/06/2000 EVENT TIME:
14:30 REPORT NUMBER:
AT # 38237 / CR # A2000-04281 REPORT DATE:
12/05/00 REVISION:
2 INVESTIGATORS:
Mike Riegel (Tcam Lead), Paul Uremiovic (Qualified Root Cause Investigator), Luis Rhoden, Kimberly Aleshire, Joe Tidmore, and Harry King ABSTRACT:
e Gril On 11/06/00, Circ Water Blowdown Vacuum Breaker, OCWI 36, was discovered leaking following a report from a local resident of water in a ditch adjacent to his property. The vacuum breaker had been leaking for several days. It is estimated that as much as 3 million gallons of water may have leaked from the failed valve. Since multiple radwaste releases may have been made while the valve was leaking some slightly radioactively contaminated water was discharged to site property and to a ditch immediately adjacent to site property. No Offsite Dose Calculation Manual (ODCM) or National Pollutant Discharge Elimination System (NPDES) limits were exceeded; therefore the event is not reportable.
The leaking valve was replaced and as much of the spilled water as possible was pumped back into the CW Blowdown piping. A root cause investigation was conducted to determine factors that contributed to the failure of the valve and subsequent release of water. Analytical techniques employed in performing the root cause investigation included Failure Modes and Effects Analysis, Barrier Analysis, Event and Causal Factor Charting and interviewing.
The vacuum breaker valve failed due to low stress high cycle fatigue. This valve had been in service since initial construction. Root causes for the valve failure were lack of an adequate preventive maintenance program for the vacuum breaker valves and an inappropriate configuration (lack of internal surge protection) of the currently installed vacuum breaker valves.
Additionally, system operating methodology is a significant contributing factor in that the operating procedure allows reinitiating blowdown flow following short duration shutdowns by using the motor operator on the isolation valves. Opening the valves using the motor operator rapidly establishes full flow and causes significant pressure surges since the piping will be depressurized and partially drained.
The failure mechanism and root causes would be applicable to all vacuum breaker valves in the Circ Water Blowdown and Makeup (M/U) systems. Similar systems are installed at Byron and LaSalle stations. However, Byron uses a different type valve.
Corrective Actions include implementing adequate preventive maintenance programs, replacing the valves and revising the operating procedure to prevent power opening of the isolation valves to establish blowdown flow.
CONDITION STATEMENT:
The float assembly for the Circ Water Blowdown Vacuum Breaker, 0CW136, experienced fatigue failure due to impact loading from excessive operating cycles al~d inadequate preventive maintenance. The failure of the float assembly resulted in the unapproved discharge of some slightly radioactively contaminated water to Site property and to the south side ditch of Smiley Road immediately adjacent to Site property. Levels were tip to 8.9E-7 Co-58, 2.3E-7 Co-60, 1.4E-7 Mn-54, and 8.1E-7 Te-123m (all yICi/g). Particulate activity was all contained within 50 meters of the failed vacuum breaker vault on Site property and was well within designated pathway release limits. Detailed radioactivity sample results and Lower Limits of Detection (LLD) are provided in Attachment 6. Activity found off site property was limited to Tritium in the water contained in the ditch along Smiley road. Tritium activity was all within designated pathway release limits, but slightly in excess of unrestricted drinking water limits. To the maximum extent feasible, all water was pumped back into the CW Blowdown line.
EVENT DESCRIPTION:
(Refer to detailed timeline (Attachment 1) for additional information.)
At approximately 14:30 on 11/6/00, the National Pollutant Discharge Elimination System (NPDES) Coordinator received a call from Illinois Environmental Protection Agency (1EPA) regarding standing water in a ditch adjacent to Site property along the south side of Smiley Road.
An area resident had reported the water and noted that the water was present in the ditch for approximately 7-10 days prior to ]EPA notification. Suspecting that a faulty vacuum breaker on either the Circ Water M/U or Blowdown System was the source of the water, the NPDES Coordinator notified the Shift Manager and OCC Director of the IEPA notification.
The NPDES Coordinator walked down the Circ Water Blowdown system at approximately 15:00 and identified that the water was coming from a valve vault that houses the OCW135/6 vacuum breaker/isolation valve assembly. The NPDES Coordinator assessed the site and concluded that the water was confined to Site property and the ditch along the south side of Smiley Road immediately adjacent to Site property.
The water in the ditch was confined by the resident's driveway to the west and by higher elevation to the east. The total length of ditch containing water was approximately 600 - 800 feet, and the total volume in the ditch was later estimated to be approximately 80,000 gallons.
The NPDES Coordinator notified the IEPA of his findings regarding the water source and the boundaries of the discharge. Station NPDES monitoring requirements were discussed and the IEPA concluded that no additional sampling was required and that there were no NPDES concerns since the water was contained and not discharging to 'Waters of the State'.
Between 16:00 - 17:00, a meeting was held with senior station management, the Shift manager and the OCC staff. The NPDES coordinator briefed the attendees on the results of his field observations of the area surrounding the vacuum breaker valve. Station management was also briefed on the discussions between the NPDES coordinator and the IEPA. Senior management directed the following actions be taken:
- 1.
Operating personnel were to evaluate water inventories and to explore potential alternate release options.
- 2.
Isolate the CW Blowdown System
- 3.
Make preparations to take the CW blowdown system out of service, drain the piping section and replace the failed vacuum breaker valve.
The Circ Water Blowdown system was then isolated in preparation for draining and repairs.
There was no discussion at this time of any need to sample for radioactivity in the water that had been discharged.
At approximately 0615 on 11/7/00, the RP Manager was contacted by the Operations Manager that there was a blowdown line leak and that RP was requested to meet with the Chemistry Manager to look at potential alternate release paths for radwaste. The reason for this request was that radwaste releases would not be possible via the blowdown system while blowdown, was isolated for repairs to the vacuum breaker valve.
Following this phone conversation, the RP Manager spoke with the RP Technical Superintendent and discussed the need to pull water samples. Included in this discussion was a conclusion that the samples should contain no radioactivity because of'the belief that the spilled water was "only lake water".
A decision to conduct water sampling of the water leaking from the manway cover of the vacuum breaker structure was made at 0800. The sample was taken at approximately 0845 and the results of the gamma isotopic analysis indicated no quantifiable peaks found (NQPF), and the tritium result was less than the lower limit of detection (LLD) of 1.87E-06 uCi/ml.
At approximately 1130, RP received infornation that the leak may have occurred for a period of 7-10 days and that the water that leaked was from the circulating water blowdown line which carries the liquid radwaste discharges from the station to the river. At 1230 on 11//7100, a decision was made to initiate soil sampling in the vicinity of the vacuum breaker structure, and to obtain a water sample from the standing water that was onsite, but near the Smiley Road ditch.
At approximately 0830, Mechanical Maintenance (MMD) personnel with assistance from System Engineering pumped out the OCW135/136 vault and began draining the blowdown piping to facilitate work on OCW135/136. At approximately 1200, after the CW Blowdown line had drained sufficiently, the entire OCW 135/136 isolation valve and vacuum breaker assembly was replaced.
A total of 5 soil samples were obtained within approximately 30 feet of the vacuum breaker structure, and 2 of the 5 samples had detectable levels of radioactivity. The onsite soil sample obtained near the Smiley Road ditch was analyzed and the gamma isotopic indicated NQPF and the water analysis indicated tritium at a level of 3.5E-5 uCi/ml.
The results of these samples were discussed with corporate Generation Support Department (GSD) RP Manager at 1900 on 11-7-00. Corporate GSD agreed to discuss the issue with the corporate Generation Support General Manager. At 1945, the Station Manager and Site Vice President were notified of the sample results.
At 0830 on 11-8-00, the RPM discussed the sample results on the morning call. At 1400, the RP, Chemistry, Regulatory Assurance, Station Manager, and Site VP met to discuss the current status, next steps, and sampling for the event.
On I 1-8-00, additional onsite sampling of the standing water in the area leading to the Smiley Road ditch was performed. Four water samples were taken at approximately 1600 and results indicated tritium levels 3.7E-05 to 5.3E-05 uCi/ml.
On 1 1-9-00 at 1000, a conference call was held with the site and corporate and an Offsite Sampling Plan, Remediation Plan, and Communications Plan was agreed to. At 1200, discussions were held with site NRC and regional NRC. At 1210, notification of the offsite release was made to Will County authorities and to the Reed Township Highway Commissioner.
At 1245, RP was dispatched to obtain water samples from the Smiley Road ditch.
At approximately 1400, four water samples were obtained from the Smiley Road ditch. Gamma isotopic analysis indicated NQPF and the tritium analyses ranged from 1.9E-05 to 2.51E-05 uCi/ml. These samples were also analyzed by Teledyne Isotopes Midwest Laboratory with similar results.
The evening of 11-9-00, the NRC Regional Office and IDNS were notified of the Smiley Road ditch sample analyses results.
At approximately 1 100 on 1 1-10-00, pumping of the water back to the blowdown line commenced. Pumping continued using a 600 gpm pump, approximately 18 hours2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br /> per day, until 2000 on 11-15-00 when all possible water had been pumped back into the blowdown line.
BACKGROUND INFORMATION:
The primary function of the Circ Water Blowdown System is to provide for Lake turnover to prevent undesirable chemical buildup in Lake. The secondary function of the Circ Water Blowdown System is to provide dilution for liquid releases.
The Circ Water Blowdown System (Attachment 7) is designed to return Cooling Lake water back to the Kankakee River. Processed fluids from the Sewage Treatment System and the Radwaste Treatment Systems discharge directly to the Circ Water Blowdown system where dilution occurs prior to release to the Kankakee River. The Wastewater Treatment Plant and the Demineralizer Regenerant Waste systems along with various strainer/filter backwashes are returned to the Cooling Lake and thus are indirectly returned to the Kankakee River.through the Blowdown line after dilution by the Cooling Lake.
The Circ Water Blowdown system begins at the Circ Water System supply to the condenser.
Two 24" carbon steel pipes tap off the Circ Water supply piping (one from each unit) and combine into a 36" common header. Motor operated isolation valves, 1/2CW018, are provided on each 24" line. The 6" Radwaste Treatment System discharge pipe connects to the 36" Blowdown header. Downstream of the Radwaste connection, the Bloivdown pipe is expanded to 48" prior to connection of the 3" Sewage Treatment Plant discharge pipe. The 48" diameter Blowdown pipe is reinforced concrete pipe (RCP) and runs along owner controlled property to the Blowdown River Screen House. Eleven vacuum breaker assemblies are incorporated at the high points along the 48" diameter RCP to prevent pipe implosion. The 48" RCP is eventually split and reduced to two 24" discharge pipes at the Kankakee River. Each 24" discharge pipe was originally equipped with a motor operated spray valve, OCWOI 8A/B. Tihe entire piping network is approximately 29,000 ft long and is operated at about 12,000 gpm (-2.5 ft/s).
A typical vacuum breaker is shown on attachment 4. On system startup, the vacuum breaker exhausts air from the piping system until the float assembly rises with water level to close and seal for system operation. Upon system shutdown, the vacuum breaker is designed to open as water level decreases. The air release or 'pilot'valve provides two functions. The primary pilot valve function is to release entrained air that accumulates at the high points during normal system operation, air that would increase head loss and reduce process flow if not removed. The pilot valve also facilitates earlier opening of the vacuum breaker on system shutdown. On shutdowns, air pockets that develop at high points may be at positive pressure, tending to hold the vacuum breaker on its seat even though water level is below the float assembly. However, the pilot valve will release the air and allow the vacuum breaker to open as soon as level drops.
Each vacuum breaker is provided with a butterfly isolation valve to facilitate vacuum breaker maintenance.
The Circ Water Blowdown system was originally designed to be maintained full and pressurized.
This was accomplished through manipulation of tle Blowdown Spray Valves, OCWO1 8A/B.
These valves were susceptible to freezing due to design and system operation requirements.
Based on this, other maintenance issues, and parts obsolescence, these valves were eventually abandoned in the full open position in the mid 1 980s. System control was transferred to the upstream motor operator isolation valves, 1/2CWO1 8. The system would no longer be maintained full and pressurized on shutdown as a result of this change. Minimal technical review was performed on the hydraulic effects on the vacuum breakers from this method of operation (ie: surge check valves were not evaluated for system incorporation).
In 1997, Chemical Feed System was relocated from the Turbine Building to the Lake Screen House under Modification M20-0-95-003. One of the primary reasons for centralization of the Chemical Feed system to the Lake Screen House was to reduce maintenance cost via system size. This design change necessitated isolating the Circ Water Blowdown System o01 a daily basis to accommodate biocide injections into the Circ Water System, because our permits do not authorize discharge of biocide to the Kankakee River.
The daily requirement to isolate Circ Water Blowdown for biocide injection prompted the Operations Department to challenge the BwOP CW-1 2 procedural requirement to slowly open the 1/2CW018 valves for'system start-up. BwOP CW-1 2 was revised to allow fast motorized operation of 1/2CWO18, in lieu of slower manual throttling, following short periods of system shutdown (i.e.: biocide injections). Minimal technical review wa,8 perforned on the hydraulic effects on the vacuum breakers from this method of operation.
Work history on the Circ Water Blowdown System vacuum breakers was reviewed. There were no recorded vacuum breaker float assembly failures prior to this event. Several instances of leaking pilot valves were noted from the review. The OCW060 pilot valve was discovered leaking in 12/98. PIF # A 1998-04324 was generated to address the flooding of site property and the Smilcy Rd ditch immediately adjacent to site property. The piping to the air release valve on the OCW058 failed in 12/96. The complete vacuum breaker assembly including pilot valve was replaced with a new assembly in 6/97. It should be noted that the OCW058 vacuum breaker failed again on about 11/20/2000 while this root cause investigation was in progress. The float assembly broke at the bowl to guide bar weld. No other significant work history was identified.
The failure ofthe OCW136 float assembly was discussed with the vendor. Based on the failure description, the vendor indicated that it appeared to be consistent with the effects of a pressure surge (i.e.: water hammer). The vendor indicated that surge protection check valves should be considered for a vacuum breaker when pipe flows exceed 6 ft/s and are required when flow velocities exceed 10 fl/s. The vendor also recommended a 7-10 year PM frequency to address valve elastomer degradation.
ROOT CAUSE ANALYSIS AND CORRECTIVE ACTIONS:
A.
Investigation and Root Cause Analysis Techniques Methods utilized for perfonning this investigation included interviewing, Barrier Analysis, Event and Causal Factor charting (Attachment 2), and Failure Modes and Effects Analysis (Attachment 3). Task analysis had also been specified in the root cause report charter. However, since task analysis is a tool that is used on investigations where problems during performance of tasks contribute to the event and no such perfonnance issues were identified during this investigation, the task analysis technique was not utilized. The majority of the conclusions in this report are based upon outcome of the Failure Modes and Effects Analysis and Barrier Analysis. Event and Causal Factor charting did lead to the development of some questions, and the identification of potential barriers, but otherwise did not provide much useful assistance in this investigation.
Root Cause Analysis The investigation determined that inadequate material condition of the 0CW136 Vacuum Breaker Valve resulting from inadequate preventive maintenance, inappropriate configuration of the vacuum breaker valve assembly and system operation methodology led to valve failure and release of water from the blowdown system.
Event and Causal Factor Chart An Event and Causal Factor Chart (E&CF) is included as Attachment 2. As stated above, the E&CF chart was used principally to identify potential barriers and to develop questions for interviews.
Barrier Analysis Barrier analysis shows 3 barriers which could have prevented or mitigated the vacuum breaker valve failure. A discussion of each barrier and its failure method/mode is provided below.
Interview results contributed significantly to the barrier analysis A. Preventive Maintenance - This program cnsures that equipment and systems are checked periodically and maintained within acceptable parameters so the) ivill function as designed. The preventive maintenance programz has no requirement to perform anya kind of internal valve inspection or operational check and no requirement to periodically replace the valves.
The vacuum breaker valves were essentially installed as run to failure components.
There are no Technical Specification requirements or NRC commitments to conduct periodic maintenance. Prior to 1999, informal walkdowns of the blowdown system were perfonned on an annual basis. For the most part, results of the walkdowns were not logged and no records maintained of the observed material condition of the valves, other than action requests for repair of observed deficiencies (e.g. leaking pilot valve). A handwritten record of a 1995 walkdown of both the CW Blowdown and Makeup lines was found. This record noted whether valves were open or shut, any leaks found, and any water in the valve vaults.
In July 1999, a preventive maintenance template from STANDARD NES-G-08, CoinEd Perfonnance Centered Maintenance (PCM) Templates, was adopted for application to the vacuum breaker valves.. The particular template chosen is specifically applicable to spring actuated safety relief valves,.and contains no discussion of applicability to float type valves. The predefine task description is "perform setpoint verification and seat leak check, or replace valve". The periodicity was set at 10 years. Although most of the valves had been in service since initial construction and had no previous maintenance history, due dates for maintenance were set well into the future, and no consideration given to replacing any of the valves based upon age or time in service. This issue is also discussed under Root Cause #1 The template chosen was the closest match from all those available in the standard PCM template index. Time pressure to complete the project was given as the reason for choosing this default as opposed to developing an appropriate template applicable to the vacuum breakers. CA 1, CA 2 and CAPR I have been generated to correct specific maintenance issues associated with the vacuum breakers. CA 4 has been generated to review a sample of preventive maintenance templates to look for additional instances of incorrectly assigned templates.
B. Desi.gnApplication - Components utilized in plant construction should (be of appropriate design for a given application to ensure acceptable service. Tile biarrier was challenged wvhen system operation wvas changedl wt'ithout changing the design or configuration of the vacuum breaker assemblies.
Original CW blowdown system operation provided for controlling blowdown flow using valves at the river screen house, thus the system would always remain full of water. This method of operation was abandoned within the first two years of operation due to repetitive failures of the control valves. Current operation of the system provides for controlling blowdown flow using valves located in the plant near the main condensers and when flow is secured, the blowdown line will depressurize and partially drain resulting in a potential pressure surge xwhen flow is reinitiated.
Discussion with the valve manufacturer revealed that if the valves are subjected to significant pressure surges, they should be equipped with surge protection. The current configuration has no surge protection. This issue is discussed more fuilly in the section titled Root Cause #2.
The reason why the system operation was changed rather than correcting the material condition of the valves at the river screen house will not be pursued since that decision was made so long ago. Similarly, the reason the change was made without considering impact on the vacuum breaker design/configuration cannot be determined. CAPR 2 has been generated to replace the current design vacuum breaker assembly with a surge-protected configuration.
In developing corrective actions, consideration was given to replacing the vacuum breakers with a totally passive standpipe system. However, some of the vacuum breakers are under pressure and the resulting stagnant column of water in the associated standpipe would necessitate installation and maintenance of heat tracing to prevent freezing in winter weather. Therefore, this potential solution will not be pursued any further.
Additionally, consideration was given to restoring control to the OCWOI 8A/B valves at the river screen house. These valves are abandoned in the full open position, which reduces head loss and provides for maximum flow. During operation currently, most of the blowdown piping is not water solid and the valves woyld have to be throttled down to maintain the system full, which would also reduce blowcdown flow. Reduced blowdown flow is not desirable from a cooling lake chemistry control standpoint.
The valves could be used to bottle up the blowdown system when shutting it down.
However, this would require closing the valves prior to securing blowdown flow in order to fill the piping, potentially resulting in damaging wate'r hammer to the system.
Therefore, this potential solution will also not be pursued any further.
C.
o
- Procedures provide the appropriate instructions to ensure actions are carried out correctly with appropriate limits and acceptance criteria. This barrierfailed ien thie procedure was pnodifiecl to allov operation of the systemn that could result in significant pressure surges (fatigiue cycles,) on the vacuum breaker valves.
The CW Blowdown System Startup, Operation and Shutdown procedure, BwOP CW-12 allows opening the blowdown isolation valves with power following "short duration shutdowns" such as routine daily chemical biocide additions. However, during a typical 2-hour shutdown of the system for biocide injection, the blowdown system can partially drain and thus a significant surge (water hammer) can result when flow is rapidly restored using the motor operators. This issue is discussed more fully under the section titled Significant Contributing Cause.
CA 3 has been developed to change the blowdown procedure to always require opening the isolation valves in stages over a few minutes to prevent water hammer.
Failure Modes and] Effects Analysis This technique %'as used to develop and analyze potential failure modes for the failed vacuum breaker assembly. A fishbone diagram used in this analysis is provided as Attachment 3.
Brittle Failurc. This mode was eliminated due to the valve float being constructed of Stainless Steel and not susceptible to brittle failure in the system operating environment.
Alanufacture. This mode was eliminated due to lack of any noted manufacturing defects in the failed valve.
Preventive Maintenance. Failure to perform adequate preventive maintenance certainly contributed to failure of the valve, and in fact lack of an adequate preventive maintenance program is one of the root causes. However, there is no evidence that any preventive maintenance ever performed on the valve contributed to the failure, therefore improper performance of preventive maintenance was eliminated as a failure mode.
Corrective Maintenance. Similarly, lack of any significant corrective maintenance history rules this out as a failure mode.
Aging. The failed valve was examined closely and did not show signs of stress corrosion cracking or other corrosion mechanisms. There was evidence of age related wear in the valve bushings, however this is not considered a contributor to failure. The length of service does come into consideration when coupled with the fatigue mechanism.
Design. The lack of surge protection, given the way the system is operated, is a strong contributor to valve failure. This information was provided via interview with the valve manufacturer. Beyond that, the valve design, specifically that of a float operated vacuum breaker, is an appropriate application for protection of the CW blowdown system.
Fatigue. Post failure analysis shows that the most probable failure mechanism was low stress high cycle fatigue. Visual inspection reveals a large fatigue and small fracture cross section, consistent with this failure mechanism.
System Operation. This mode is a strong contributor to failure, due to the fatigue cycles caused by operation. System operation is discussed in great detail in other sections of the report and therefore will not be elaborated on in this section.
B.
Summary of Causes and Corrective Actions Root Cause #1 The preventive maintenance program for the CW Blowdown vacuum breaker valves is inadequate. The valves, with one exception, had all been in service since initial construction of the plant prior to the failure and subsequent replacement of the OCW136 Valve. The other valve replacement was due to problems with the pilot valve sensing line failing (due to general corrosion), and not with the main vacuum breaker valve itself. Preventive Maintenance essentially consisted of an annual walkdown of the valves by the system engineer, and if no leakage was observed, the valves were deemed to be OK; results of the walkdowns were not logged.
A preventive maintenance template for the valves was adopted from Performance Centered Maintenance (PCM) standard templates in July 1999. The template chosen was for "Spring Actuated Pressure Relief Valves" which was the closest match from the choices available. The "why" given for this was time pressure to complete the preventive maintenance templates. This cause will not be pursued further because the project is completed and therefore no relevant CAPR would be developed. CA 4 has been generated to sample for additional inappropriate preventive maintenance templates.
CA I has been generated to replace all vacuum breaker valves and thus restore current material condition CA 2 has been generated to develop an adequate preventive maintenance program for the valves which includes periodic inspection of the valves including internals or provides for valve replacement at appropriate intervals. This item will also include the development of system walkdown inspection requirements including specified frequency of walkdowns and documentation/reporting of walkdown results.
CAPR I has been generated to implement the revised preventive maintenance program and system walkdown inspection requirements.
Root Cause #2 The configuration of the current vacuum breaker valve assembly is inappropriate. The current valve assembly (Attachment 4) consists of an integral butterfly isolation valve and vacuum breaker float valve with an attached pilot or "air release" valve. Discussion with the valve manufacturer revealed that the current configuration would be susceptible to premature failure if the valves were subjected to significant repetitive pressure surges during operation, such as would be experienced during startup with rapid filling/pressurization of the system. The manufacturer stated that with pressure surge conditions, a valve with built in surge protection (Attachment 5) would be required.
The original system operation provided for controlling blowdown flow using valves at the river screen house, thus the system would always remain fill of water. This method of operating the system was abandoned within the first two years of operation due to repetitive failures (caused principally by freezing) of the control valves at the river screen house, and difficulty in remotely controlling the valves at the river screen house. Current operation of the system provides for controlling blowdown flow using valves located in the plant near the main condensers, and thus when flow is secured, the blowdown line will depressurize and at least partially drain resulting in a potential pressure surge when flow is re initiated. The reason for why system operation was changed as opposed to correcting the material condition of the valves at the river screen house is beyond the scope of this investigation due to the large time interval from initial construction and will not be evaluated further.
CAPR 2 has been generated to replace the current design vacuum breaker assembly with a surge protected (Attachment 5) configuration. This action can be performed coincident with CA 1, but is listed separately to provide a definite link of CAPRs to root causes.
Sigznificant Contributing Cause A significant contributing cause, if in fact not a third root cause, is system operating methodology. As stated earlier, blowdown flow is controlled using valves in the plant located near the condensers, specifically the l/2CW018 valves, which are 24-inch motor, operated butterfly valves. If blowdown flow is initiated by opening these valves using the motor operators, a significant pressure surge can result.
The Chemical Feed System configuration was modified in 1997, changing the injection point where chemicals are added to the CW system, and now CW blowdown flow must be secured whenever chlorination of the CW system is accomplished. This is done to comply with NPDES permits, which forbid discharging of biocide to the river. Chlorination is performed for each unit on a daily basis and therefore it is necessary to alter blowdown flow on a daily basis to ensure there is no blo~vdown from the unit being chlorinated.
The CW Blowdown System Startup, Operation and Shutdown procedure, BwOP CW-12, contains a note that states in part "Slowly opening MOV CWOI 8, Units i & 2 CW Blowdown Isol Vlv, in stages over a few minutes will prevent damage to the CW Blowdown piping/components due to water hammering of tile drained piping." To provide some relief to the operators from having to manually operate the isolation valves on a daily basis in support of chorination operations, the procedure was changed in 1998 to also state "For short duration shutdowns of CW blowdown (i.e.: routine daily chemical biocide' additions, etc.), reopening with power from MCR is acceptable." This change was authorized without sufficient technical analysis, and in fact following a typical two hour shutdown for chlorination there is a significant surge when flow is restored, particularly if only single unit blowdown is in operation and blowdown is totally secured (as opposed to shifting to the opposite unit) for chlorination.
CA 3 has been developed to change the blowdown procedure, BwOP CW 12, to always require slowly opening the isolation valves in stages over a few minutes when initiating flow from a no blowdown flow condition. This CA does not apply when shifting blowdown from one unit to another as long as blowdown is not secured.
C.
Equipnicnt Failurcs (EF)
EF # Summary of EF Associated Causal Factor
- 1.
OCW136 vacuum breaker float assembly failure (date of 1, 2, and 3 failure estimated to be between 10/27/2000 and 10/30/2000)
- 2.
OCW058 vacuum breaker float assembly failure (date of 1, 2, and 3 failure estimated to be 11/20/2000)
Both the failed CW blowdown valves, OCWXI36 and OCW058 (failed 11/20) were analyzed to detennine failure mode, and both were determined to have failed from low stress high cycle fatigue. This type of failure mode is consistent with the causes listed above.
D.
Causal Factors (CFs)
(Root Causes are identified by Asterisks)
CF # Summarv of Causal Factors Associated Corrective Action 1
- Root Cause - The preventive maintenance programs for the CW Make-Up and Blowdown System vacuum breakers are inadequate. An effective preventive maintenance program needs to be developed and implemented for both systems.
2*
Root Cause - The design configuration of the current vacuum breaker valve assembly is inappropriate. The current valve assembly (Attachment 4) consists of an integral butterfly isolation valve and vacuum breaker float valve with an attached pilot or "air release" valve.
Discussion with the valve manufacturer revealed that the current configuration would be susceptible to premature failure if the valves were subjected to significant repetitive pressure surges during-operation, such as would be experienced during startup with rapid filling/pressurization of the system. The manufacturer stated that with pressure surge conditions, a valve with built in surge protection (Attachment 5) would be required.
3 Significant Contributing Cause - Current system operating procedure (BwOP CW-12) allows initiating CW blowdown flow by opening the isolation valves using the motor operators if blowdown flow has only been shutdown for a short duration (i.e. routine daily chemical biocide additions, etc.). Operating the blowdown system in this manner can result in significant pressure surges.
'4 Contributing Cause - In July 1999, a preventive maintenance template from STANDARD NES-G-08, ComEd Perfornance Centered Maintenance (PCM)
Templates, was adopted for application to the vacuum breaker valves. The particular template chosen is specifically applicable to spring actuated safety relief valves, and contains no discussion of applicability to float type valves. Time pressure to complete the project was given as the reason for choosing this default as opposed to developing an appropriate template applicable to the vacuum breakers.
CAPRI, CAI, and CA2 CAPR2 and CAI CA3 CA2 and CA4
E.
Corrective Actions (Corrective Actions to Prevent Recurrence are labeled CAPRs) 1nimediate Corrective Actions:
- 0 B
S 0
0 0
- These Act ions The CW blowdown system was isolated to stop the leak A team was assembled to recover from the leak and restore the system Appropriate notifications were made The leaking vacuum breaker valve was replaced Spilled water was pumped back into the CW Blowdown System Radioactivity samples were taken to comply with I OCFR50.75(G).
A root cause investigation was commenced are all listed as immediate actions even though some occurred over several days.
Corrective Action Assignee A8930TT Due Date 03/01/01 CAPR I Description of corrective action to be taken:
Implement a revised preventive maintenance program for the float operated vacuum breaker valve assemblies for the CNN' Blowdown and Makeup Systems. This PIM will be developed by CA 2 and will include specific intervals for inspection of valve internals or provide for periodic replacement of the valves. This item wvill also include the implementation of system walkdown inspection requirements including specified frequency of walkdowns and docutnientation/reporting of walkdown results.
CAPR 2 Description of corrective action to be taken:
Replace the current design vacuum breaker assembly with a surge-protected configuration, Valve an(l Primer Company Cat. ID 1036974 (6 inch),
1036975 (8 inch), 1036976 (10 inch).
This CAPR should be accomplished coincident with CA 1 to minimize the number of voalve replacements.
A8930TT (design approval/doc) 03/01/01 A8922MM (physical work)
Actions Corrective Action Assignee Duc Datc CAI Description of corrective action to be A8922MM 03/01/01 taken:
Replace all vacuum breaker valve assemblies in the CNN' Blowdowion and Makeup Systems to restore system material condition. This CA should be accomplished coincident with CAPR 2 to minimize the number of valve replacements.
CA2 Description of corrective action to be A8930TT 02/01/01 taken:
Develop an adequate preventive maintenance program for the CW Blowdown and Makeup System vacuum breaker valve assemblics which includes periodic inspection of the alves including valve internals or provides for valve replacement at appropriate intervals. This item will also include the development of system walkdowvn inspection req(uirelmlents including specified frequency of wialkdoowns and documentation/reporting of valkdown results. This program will be implemented by CAPR 1.
CA3 Description of corrective action to be A89100P 01/05/01 taken:
Revise BwOP CWV'-12 to always require slowly opening the blowvdowin isolation valves (1/2CW'018) in stages ovcr a few minutes when initiating blowvdown flowv from a no blowdown nowv condition.
This CA does not apply when shifting blowdown from one unit to another as long as blowvdown is not secured. This procedure revision shall be identified as a corrective action per this AT' item.
Actions Corrective Action Assignee Due Date CA4 Description of corrective action to be A8930TZ 05/01/01 taken:
Perform a review of a representative sample of preventive maintenance templates to determine if there are additional inappropriately assigned preventiv*e maintenance templates.
EFRI Description of corrective action to be A8930TT 08/01/01 taken:
Perform an effectiveness review of the CAPRs.
F.
Extent of Condition Other Exelon/Amergen Nuclear sites were contacted to determine how those plants are configured for circ water blowdown and makeup and if they have experienced any similar problems with vacuum breaker float assembly failures. Byron and LaSalle stations were the only stations confirmed to have circ water blowdown and makeup systems that utilize vacuum breakers in their design. For CW blowdown and makeup systems, the extent of condition is limited to Byron and LaSalle.
Byron Station replaced their fiberglass blowdown and makeup piping in 1987 with carbon steel due to a line failure associated with ground shifting. 12" Golden Anderson's model GH-7K vacuum breakers with line surge protection were installed at that time. System Engineering walkdowns are conducted annually and Operations regularly drives down the lines when they make River Screen House rounds. No vacuum breaker failures have been identified and only minor amounts of water have been discovered contained within the valve vaults.
LaSalle's Operating Department performs inspections on their circ water blowdown and makeup systems on a semi annual basis. The inspections consist of leak checks and flushing the air release valve of any debris. The vacuum relief float is also checked and cleaned. The majority of the problems they experience are related to plugging and freezing. The smaller air release valve float is the component that has frozen and it only affected the air release and not the vacuum relief. Corrosion has also been identified as an issue on the piping from the vacuum break to the air release valve. There is no history of vacuum breaker float assembly failures at LaSalle.
There are differences in system operation among the sites. In general, LaSalle operates their system continuously, whereas Byron and Braidwood must cycle their blowdown systems to accommodate chemical feed additions. Additionally, Byron still controls blowdown flow using the OCWI 8A/13 spray valves at their river screen house, thus their blowdown system is maintained full at all times. Byron's spray valves are typically throttled to 20-22% open to maintain 12-13k gpm blowdown flow. The Byron spray valves are reportedly difficult to control and require frequent maintenance to remain operable.
Other potential off site release paths were evaluated for extent of condition. The summary of the evaluation is as follows:
NPDES Outfalls 001 (a) - Wastewater Treatment Plant This system discharges to the U-2 Circ Water return line (to the lake)
-Discharge of 'waste'water from the non-radiological portion of the plant. This discharge consists mainly of:
Turbine Building Drains (TE,TF), Fire and Oil sump, Tendon Tunnel sumps, Aux Bir blowdown, Secondary side drains, Pretreatment drains and can be alternate discharge path for regenerant waste from Muds and CPs.
001(d) - Demineralizer Regenerant Waste
-Discharge of regenerant waste. This discharge consists mainly of:
Muds and CP regencrant waste, Regenerant chemical area drains (acid & caustic) and portable demineralizer regenerant waste (RO waste)
This discharges to the cooling lake via the circ water return to the lake 00 I(e) - River Intake screen backwash
-river screen house backwash is directed to the make-up pump suction (forebay) only during pump operation and is ultimately pumped to the lake with the make-tip. This is includes as a point source only with no monitoring required.
002 - North Site Stornnwater Runoff This discharge consists mainly of stormwater from the Parking lots, Transformer areas, Station and North Site area as well as Station roof drains. The discharge path is via the runoff ditch along the west side of the property and is ultimately discharged to the Mazon River. No monitoring is required for stormwater 003 - South Site Stormwater Runoff This discharge consists of stormwater runoff from the area south and east of the main site (near the LSH). The discharge path is via the runoff ditch along the west side of the property and is ultimately discharged to the Mazon River. No monitoring is required for stormwvater 004 - Switchyard Area Runoff This discharge consists of stormwater runoff from the switchyard area. An oil separator exists in this area to remove oil in the event of leakage from the switchyard equipment.
The discharge path is via the runoff ditch along the west side of the property and is ultimately discharged to the Mazon River. No monitoring is required for stormwater
There are two special conditions-that identify flowpaths not included as 'normal' discharges.
Special Condition 9: Discharge of station cooling pond water to adjacent impoundments owned by the permittee, to replace water which is withdrawn from these impoundments for station operation during periods of low flow in the Kankakee River when the station must decouple its operation from the river, is hereby permitted for these emergency periods.
No monitoring is required however the Agency (IEPA) must be notified if this occurs.
This simply says we can refill Monster Lake if we have to pump it down during drought conditions.
Special Condition 12: An emergency cooling pond overflow exists tributary to an unnamed drainage ditch that is tributary to the Mazon River. Discharges to this overflow shall be subject to the bypass provisions of 40 CFR 122.41 (m)
-This states that monitoring of 001 (Cooling Pond Blowdown) parameters must be performed daily during discharge.
Additionally the Make-up to the lake (river water) would be a potential source of 'discharge' should we develop a leak in that line.
The NPDES related treatment systems discharge either directly into the blowdown line or to the Cire Water return to the lake and then ultimately discharge through the blowdown.
The Wastewater Treatment and Sewage Treatment processes are separated from the Main Plant or Turb/Aux buildings. Any leakage or failures within these systems would be contained within their respective buildings and/or general area. The Demineralizer Waste process is contained within the Turbine Bldg and the Radwaste systems are located primarily within the Aux Bldg.,
however some components (release tanks) are located in the Turbine building. Any leakage or failures within these systems would be contained within their respective areas.
The connections to the Circ Water return to the lake (Wastewater and Dcmin Waste) are located underground in the area west of the Turb Bldg. The Sewage Treatment Plant connection is underground in the area north of the Sewage Plant and the radwaste tank discharge ties in to the blowdown within the Turb/Aux Bldg. Any catastrophic failure of these connections should be visible by localized bubbling or saturation of the ground in which case the discharge could be isolated. This would prompt investigation by excavation to determine cause of failure and initiation of repairs. Preventive measures could be taken at that time (berms or other barriers) to prevent this leakage from entering the runoff ditches.
Periodic rounds of these systems should provide the early detection of any unusual conditions.
Preventive measures could then be taken to eliminate potential for release of water from these systems..
G.
Risk Assessment There were no plant specific risks associated with this issue. There were no risks to the CW Blowdown system as a result of this issue, since in the failed condition the vacuum breaker assembly would still function to prevent a vacuum from forming and causing damage to the blowdown piping.
PREVIOUS EVENTS:
Search of CAP, EWCS, OPEX, etc. was conducted. The OPEX search wvas limited to the previous 2 years and confined to the variables of "unplanned releases of liquids", "vacuum breaker failure and releases", and "blowdown vacuum breaker". The EWCS search was limited to both Circ Water Makeup and Blowdown vacuum breaker failures. The CAP system search was limited to the Circ Water System. No similar vacuum breaker float failure events were identified via these searches.
EVALUATOR COMMENTS:
The station was slow to implement Event Response Guidelines, CWPI-NSP-AP-1-l, orNGG Issues Management, OP-AA-I 91-503. The initial facts of water iln the ditch oil Smiley Road and the source of the wvater being a leaking vacuum breaker on CW bloxvdown piping were known at approximately 1600 on 11/06/00. NGG Issues Management was not entered until sometime on 11/09/00 after results of radioactive sampling showed both isotopic. and Tritium samples above LLD, a delay of about 3 days. The decision to forml a root cause team to investigate the issue was not made until 3 days after the event. The station did establish separate teams for event recovery and root cause.
Interviews with station management personnel responding to the initial event reports revealed there was no consideration for use of the Event Response Guidelines or NGG Issues Management procedures for general guidance. OP-AA-101-503, NGG Issues Management was developed as a corrective action to the Braidwood Oil Spill Event in order to respond appropriately to significant issues, including those that have potential media or public interest.
Since it was known that the blowdown line had most likely been leaking for more than a week (based on the information from the resident who reported the water in the Smiley Road ditch) and that radioactive releases from the Site are conducted via the blowdown line, the potential for radioactivity in the Smiley Road ditch should have prompted entry into the NGG Issues Management Procedure due to the possibility of public and media interest.
Entry into the NGG Issues and Management Procedure did eventually occur on 11-9-00, three days following the initial event. The station did detennine that a prompt investigation was not required.
The establishment of separate teams for event recovery and root cause is an apparent lesson learned from the Oil Separator event (CR # A2000-2683) which occurred in June 2000.
Condition Report #A2000-04465 has been written to document slow station response in implementing NGG Issues Management or Event Response Guidelines.