05000305/LER-2004-001

From kanterella
Jump to navigation Jump to search
LER-2004-001,
Docket Number
Event date: 01-16-2004
Report date: 09-14-2004
Reporting criterion: 10 CFR 50.73(a)(2)(vii), Common Cause Inoperability
Initial Reporting
3052004001R01 - NRC Website
v • d • e

DESCRIPTION

On January 15, 2004, with the plant operating at 100% power, the discovery of significant biofouling (blockage by biological matter) of both trains of Safety Injection (SI)[BQ] Pump [P] lube oil (LO) coolers [CLR] resulted in initiating a Technical Specifications (TS) forced shutdown of the Kewaunee Nuclear Power Plant (KNPP). KNPP was taken offline at approximately 0600 on Friday, 1/16/04. This event was initially reported on January 16, 2004, at 0020 according to 10CFR50.72(b)(2)(i), TS Required Shutdown, 10CFR50.72 (b)(3)(ii)(B), Unanalyzed Condition, and 10CFR50.72(b)(3)(v)(D), Accident Mitigation.

During maintenance and inspection activities on SI pump A on January 15, 2004, the lube oil cooler was found to be biofouled. After the cooler was cleaned, very similar conditions were found in SI pump B. In fact, visible flow was only occurring in 3 of the 20 tubes in the B cooler inlet pass. Even though the coolers involved were cleaned and operational, Nuclear Management Company (NMC) decided to shut down the plant.

Service water (SW)[B11 flow through the coolers before and after cleaning was measured at approximately four and six gallons per minute, pre and post-cleaning, respectively. Previous surveillances were considered acceptable based on visible flow through a sight glass [FG]. A review of operating experience from Lake Michigan plants indicated that coolers with 3/8-inch tubes could become substantially blocked within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of cleaning. Although partial blockage tended to increase velocity through the remaining tubes, the surface area available for heat exchange was reduced, potentially affecting the cooler's ability to remove the design basis heat load. The event also showed that low service water velocity can contribute to the buildup of debris in the coolers. The safety injection pump lube oil coolers were operated with continuous, low-velocity flow; consequently, they were susceptible to debris accumulation.

The SI pump LO cooler was a two-pass heat exchanger. It was approximately 18 inches long and contained 20 3/8-inch diameter tubes for each pass. Tubes were contained in a tubesheet at each end; each end also had a hemispherical chamber. The inlet/outlet chamber contained a horizontal divider plate with an inlet nozzle on the top half and an outlet nozzle on the bottom. The opposite end, the return chamber, did not have a divider plate. Service water flowed continuously into the cooler through a 3/4­ inch pipe into the top half of the inlet/outlet chamber and continued through the 20 tubes on the top half of the cooler. Flow exited the top tubes, entered the return end chamber and continued into the bottom 20 tubes for the second pass. The service water then exited the second pass tubes into the outlet chamber and outlet nozzle [NZL].

The small size of the LO cooler tubes, coupled with the close tube-to-tube spacing of the tubesheet, and the relatively low flow velocity, resulted in lake weed becoming trapped against the tubesheet. This biofouling condition existed on the top half of the inlet/outlet tubesheet and the bottom half of the return end tubesheet.

The historical experience with the SI LO coolers demonstrated that this event was not a new phenomenon with the coolers. However, new performance criteria were being applied to the clean and inspect activity to ensure heat removal capability of the heat exchanger. Considering KNPP and industry operating experience (OE), it was concluded that the design of the SI lube oil cooler may not have been sufficient given the low SW flow velocity, tube size and system configuration. The remaining safety FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (6) PAGE (3) II 2004 — 001� 01 related components in the SW System have not demonstrated similar fouling and have velocity characteristics that support this conclusion.

CAUSE OF THE EVENT

The root cause of the LO cooler becoming plugged was an inadequate LO cooler design for the conditions and foreign materials that Lake Michigan could introduce into the SW system. The most significant contributing factors were the low flow velocities at the cooler and the tube sheet limiting dimensions.

In addition to the design application deficiency, a review of historical data and industry performance information related to heat exchanger fouling revealed a number of previous opportunities to identify and correct the conditions that led to the plant shutdown. The following is a summation of some of the issues identified by the root cause evaluation team:

The Nuclear Regulatory Commission (NRC) issued Generic Letter 89-13 requesting plants to conduct performance testing on heat exchangers cooled by service water to verify heat transfer capability.

Alternate approaches such as "inspect and clean" were considered acceptable for small, accessible components such as lube oil coolers. KNPP elected to implement this requirement for the SI pump LO coolers by establishing a recurring preventive maintenance action item that would be performed annually consistent with the NRC directive that inspections be conducted at every refueling outage. However, a comprehensive program document with assigned program owner that described the program was not established. There was also no specific acceptance criterion established for biofouling.

Initial inspection of the two SI pump LO coolers for Generic Letter 89-13 was performed during the spring 1992 refueling outage. The coolers were found to be approximately 50% fouled with "green plant matter" and Incident Report (IR)92-045 was established to address the issue. One of the corrective action recommendations from the IR was to determine maximum temperature differences between the service water and the lube oil to establish criteria to clean the coolers. On December 5, 1994, the Plant Operations Review Committee (PORC) approved canceling this IR corrective action. Consequently the PORC may have contributed to plant overconfidence that the issue was not of major concern.

This inspection also set the tone for future acceptances of the biofouling conditions, specifically:

  • The SI pumps were operable based on the lube oil temperature measurements, which were on the order of 120 degrees F during surveillance testing.
  • Visible flow though the sight glass was considered acceptable flow.
  • Annual inspection was believed to be excessive and future inspections would show that the frequency could be reduced.
  • No specific acceptance criteria for biofouling were identified.

FACILITY NAME (1)�

  • DOCKET NUMBER (2)� LER NUMBER (6� PAGE (3) The LO coolers were inspected and cleaned on roughly an annual basis between 1992 and 2000.

Fouling was present to varying degrees. Since no acceptance criterion was specified in the work orders, no corrective action reports were issued.

In 1997, as a result of radiography performed on the service water piping, flow to the LO coolers and SI pump stuffing boxes was measured to confirm that nodules seen on the radiographs were not significantly impeding flow. The flow measurements revealed that the 14 gallons per minute (gpm) minimum flow required by the vendor drawing (10 gpm to the cooler and 2 gpm to each stuffing box) was not being met. Kewaunee Assessment Process (KAP) 97-0783 was initiated to evaluate the condition.

A calculation included in the KAP 97-0783 evaluation concluded that the cooler required heat load removal capacity would be met with as low as 1.5 gpm flow through the cooler. The calculation assumed that all tubes would be open when calculating minimum required flow that the pipe must be able to pass.

C11423 was the calculation that had established the minimum flow requirements for the SI LO coolers to be 1.5 gpm. In late 2003 a review of calculation C11423 was performed. The calc was reviewed as a precautionary measure prior to Radiograph Testing (RT) of the SW piping to the LO coolers.

Unacceptable RT results would require installation of ultrasonic flow measuring devices (UFMs) to quantify minimum flow to the SI LO coolers. This calculation review questioned the minimum flow required and initiated the calculation improvement which established the new performance criteria for cooler plugging. The calculation concluded the following:

  • Three (3) gpm is sufficient, even for the maximum analyzed service water temperature of 80 degrees F, if the cooler has less than half of the tubes blocked per pass.
  • If more than half of the tubes per pass are blocked, the cooler may not be able to remove the required heat.

The calculation was approved on January 12, 2004 and once both LO coolers were found to be substantially blocked, the decision was made to declare the SI pumps inoperable, and subsequently required shutdown of the plant.

ANALYSIS OF THE EVENT

This event is reportable under 10CFR50.73(a)(2)(vii), "The completion of any nuclear plant shutdown required by the plant's Technical Specifications.

This event required a unit shutdown in accordance with the Kewaunee Technical Specifications due to both trains of Safety Injection being declared inoperable. At the time of the shutdown, the SI LO coolers were determined to have failed to meet the performance criteria, based on the number of blocked tubes versus service water temperature. A full loss of heat removal capability did not occur based on as-found evidence found to support that the coolers were ever completely blocked at any time.

FACILITY NAME (1)�t , DOCKET NUMBER (2) h «�LER NUMBER (6t PAGE (3) The past operability of the SI pumps was evaluated by performing a series of calculations and analyses to determine the impact of the degradation on the heat removal capabilities of the LO coolers, and the impact of increased lube oil temperature on pump operation. The calculations and analyses are documented in Report MPR-2658, "Safety Significance Evaluation for Kewaunee High Head Safety Injection Pump Lube Oil Cooler Degradation", Revision 0. A summary of the report is provided as follows:

  • The debris buildup on the tube sheets degrades service water flow through the coolers. The typical as-found condition during inspections has been the equivalent of about three tubes of flow through each tube pass. Based on past inspections of the lube oil coolers, and operator observations during regular rounds, it was concluded that there was always SW flow through the coolers. The actual flow conditions and blockage of the cooler are difficult to determine, so the evaluation in the report conservatively assumed the bounding case of full blockage and no SW flow.
  • The allowable lube oil temperature is limited by oil performance at elevated temperatures, the effect of increased temperatures on components in the lube oil system, and the required operating time for the SI pumps. Based on data from the pump thrust bearing vendor, as well as additional evaluations of the pump components, the SI pumps can operate reliably and are capable of performing all required safety functions for at least 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> for bulk lube oil temperatures up to 230 degrees F, and local lube oil temperatures at the bearings of 250 degrees F.
  • There are two events that were evaluated as being limiting: 1) A small break loss-of-coolant-accident (LOCA), and 2) An Appendix R event. A small break LOCA results in higher lube oil temperatures, but the event duration and required operating time are shorter than an Appendix R event. An Appendix R event results in lower, but still elevated lube oil temperatures that last for a longer duration. The required operating time for the SI pumps for a small break LOCA is about seven hours of continuous service. During an Appendix R event, the SI pump will be cycled on and off during the event to provide reactor coolant system inventory control. The longest single continuous run cycle is 90 minutes, and the total run time during the event is about 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br />. The total time at elevated temperatures could be longer, depending on the amount of cooling during the idle cycles when the pump is not running. For conservatism, the evaluation assumed the pump must operate for the entire 80-hour Appendix R event. In addition, because the allowable lube oil temperature is dependent on required operating time, and these two events have different operating times, the evaluation considered both of these events.
  • In the bounding case of total loss of service water flow through the lube oil cooler, the bulk lube oil temperature will remain below 220 degrees F for the small break LOCA, and below 211 degrees F for an Appendix R event. For the best estimate limiting case of 3 tubes and 3 gpm of SW flow, the maximum lube oil temperature would be less than 179 degrees F, even for long-term operation at the maximum recorded SW temperature and small break LOCA conditions. The corresponding maximum temperature for a single tube of SW flow is less than 202 degrees F.

These results show that the allowable lube oil operating time of 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> based on a conservative maximum lube oil temperature (250 degrees F is greater than the small break LOCA maximum lube oil temperature) is greater than the limiting required operating time (80 hours9.259259e-4 days <br />0.0222 hours <br />1.322751e-4 weeks <br />3.044e-5 months <br /> for the Appendix R event).

Therefore, the lake weed degradation of the lube oil coolers would not have prevented the ability of the SI pumps to perform all required safety functions.

FACILITY NAME (1) DOCKET NUMBER (2)� LER NUMBER (61 PAGE (3)

CORRECTIVE ACTIONS

An extensive extent of condition review was performed to determine potential effects of plugging on the remainder of the components cooled by the SW system. This included:

  • Comparing the internal dimensions of tubes in all the plant safety related heat exchangers and coolers and assessing the flow velocities based on available flow test data. Some non-safety related plant heat exchangers were also reviewed.
  • Visually inspecting a number of safety and non-safety related plant heat exchangers and area fan coil coolers.
  • Re-examining heat exchanger performance data for signs of degradation.
  • Measuring the internal clearances of the rotating SW strainers to ensure bypass potential for lake debris did not exist.

Based upon review of plant data, industry OE, and prior evaluations at KNPP and Point Beach Nuclear Plant (PBNP), NMC concluded that SW cooled safety related heat exchangers with tube sizes of 1/2-inch and greater are significantly less susceptible to similar lake weed fouling. As described above, the SI lube oil cooler was a two-pass exchanger with twenty (20) 3/8-inch tubes in each pass. At the design flow rate of 6 gallons per minute (GPM), the tube flow velocity was approximately 1.2 feet per second (FPS).

NUREG/CR-5210 guidance to minimize silting and biofouling, recommends maintaining flow velocity above 3 FPS. The design of the heat exchanger did not meet this criterion. In addition to the low velocity in the heat exchanger, industry and KNPP operating experience indicate that 3/8-inch tubes and smaller have a higher occurrence of fouling. This was due to a number of factors including flow velocity and distance between the tubes (spacing on the tube sheet). Of the safety related components normally supplied with SW, the SI LO cooler is the only safety-related cooler with 3/8-inch tubes, with the next smaller size being 1/2-inch tubes in the Turbine [TRB] Driven Auxiliary Feedwater (TDAFW)[BA] pump LO cooler.

The SI LO coolers were replaced with a different design under Design Change Request DCR 3518.

The new coolers are of a single tube coil design with a 3/4 - inch diameter tube (ID = 0.62 inch). The new cooler configuration with a larger, single tube, along with the elimination of tubesheets that can catch debris, serve to decrease the potential for accumulation of lake weed in the cooler. The single tube design is essentially an extension of the SW piping, and will pass debris that may travel through the system. Flow velocities through the new cooler prevent silt and debris accumulation from occurring, and provide the required heat removal capability.

SIMILAR EVENTS

None.

FACILITY NAME (1) - DOCKET NUMBER (2) LER NUMBER (6 . PAGE (3)

EQUIPMENT INFORMATION

SI Pumps:

  • Bingham-Willamette Co. (now Sulzer Bingham Co.), Model 4x6x9, Type CP, 11 stage, 3600 RPM pump.

Old Lube OH Coolers:

  • Thermxchanger, Inc, Type BF 2-Pass Liquid Cooler.

Retrieved from "https://kanterella.com/w/index.php?title=05000305/LER-2004-001&oldid=199050"