Part 21 Report - 1996-602

ACCESSION #: 9607240244 LICENSEE EVENT REPORT (LER) FACILITY NAME: Donald C. Cook Nuclear Plant - Unit 2 PAGE: 1 OF 6 DOCKET NUMBER: 05000316 TITLE: Two Cam Follower Springs Found Broken on Unit 2CD Emergency Diesel Generator Due to Manufacturing Flaw, Part 21 Follow-up Report EVENT DATE: 04/13/96 LER #: 96-008-00 REPORT DATE: 07/19/96 OTHER FACILITIES INVOLVED: None DOCKET NO: 05000 OPERATING MODE: 6 POWER LEVEL: 0 THIS REPORT IS SUBMITTED PURSUANT TO THE REQUIREMENTS OF 10 CFR SECTION: 50.73(a)(2)(v) & Other LICENSEE CONTACT FOR THIS LER: NAME: Mr. Ken Baker, Production TELEPHONE: (616) 465-5901, Engineering Manager x2402 COMPONENT FAILURE DESCRIPTION: CAUSE: SYSTEM: COMPONENT: MANUFACTURER: REPORTABLE NPRDS: SUPPLEMENTAL REPORT EXPECTED: NO ABSTRACT: In accordance with 10CFR21.21(c)(3)(ii) and 10CFR50.73(a)(2)(v), the following information is submitted to fulfill both Part 21 reporting requirements and Licensee Event Report requirements. On June 19, 1996, a 10CFR Part 21 report was made to the NRC Operations Center regarding broken cam follower springs discovered on the Unit 2 CD Emergency Diesel Generator (EDG) during the Unit 2 spring refueling outage. These spring failures were analyzed and determined to possibly constitute a significant safety hazard, as they could prevent the EDG from performing its design function. In accordance with the guidance presented in NUREG 0302 on Part 21 reporting, this event was determined to be reportable under 10CFR50.73(a)(2)(v), as an event or condition which alone could prevent the fulfillment of the safety function of a system. The defect found on both failed springs is thought to be the result of a manufacturing flaw, which appeared as a surface discontinuity or seam, running for at least 1 coil parallel to the axis of the wire down to the corner of the fracture surface. The broken cam follower springs found on the Number 4 Rear Bank and the Number 6 Front Bank cylinders were replaced and the 2CD EDG returned to service. The Unit 2AB EDG and the Unit 1AB and 1CD Emergency Diesel Generators were also monitored for signs of failed springs. No evidence was found to indicate a similar problem on the other EDGs. Based on the results of the investigation, it was determined that the event was of minimal safety significance. The event did not affect safe operation of the plant, and posed no threat to the health or safety of the public. END OF ABSTRACT TEXT PAGE 2 OF 6 Condition Prior to Event Unit Two was in Mode 6, Refuel, with all fuel removed from the core. Description of Event On April 13,1996, the Unit 2 CD Emergency Diesel Generator (EDG) (EIIS/EK) was started following replacement of the Number 4 Rear Bank (4RB) fuel pump and injector. A knocking sound was heard coming from the engine, loud enough to be heard by personnel on the other side of the engine. The noise appeared to be coming from the vicinity of the 4RB cylinder, and the cylinder exhaust temperatures indicated that the cylinder was not functioning. The engine was immediately shutdown. Upon disassembly the aluminum cam follower was found seized in its bore in the cast iron engine frame, and the cam follower spring broken into three pieces. The physical evidence indicated that one of the spring pieces wedged the aluminum cam follower in its bore, causing it to seize. The cam lobe continued to rotate and strike the cam follower, causing the noise, and damaging the follower and loosening the set screws protruding from the fuel pump. Based on the results of that inspection it was decided to disassemble and inspect all cylinders on the 2CD engine. A second broken spring was found on the Number 6 Front Bank (6FB) cylinder. This spring was broken into two pieces, with the follower and cover plate exhibiting evidence of damage similar to that of the 4RB cylinder. The follower, however, had not seized, nor were the set screws of the fuel pump loosened. The 6FB follower was able to continue to function properly based on the following action afforded by its own mass and the assistance of the fuel pump spring. Cause of the Event Examination of the failed springs revealed a surface discontinuity, or seam, running for at least one coil parallel to the axis of the wire down to the corner of the fracture surface. This flaw is thought to be a manufacturing flaw. The physical appearance of the two springs removed from the Unit 2CD EDG support that the springs had been broken for some period of time. The fracture surfaces of both springs were oxidized, had rounded edges, and the 4RB spring had a burnished appearance on the underside of the coils, indicating collision damage with itself and surrounding material after it fractured. Both of these mechanism would require a length of time to wear chrome-vanadium valve spring quality steel, which typically has a hardness in the range of 40 to 55 Rockwell C scale. TEXT PAGE 3 OF 6 Cause of the Event (cont'd) The nature of the fatigue failure and the endurance limit of ferritic materials supports that the springs had been fractured for some time, and that springs with similar discontinuities would be expected to have failed by now. At the time of discovery of this failure, each of the springs in the Unit 2CD engine had approximately 25 million stress cycles. While there is a lack of detailed material specifications for this spring material and there is a high degree of variability in the individual data points used to generate stress-cycle data presented in "S-N curves" , by comparison with published endurance limits and typical design practices it would be expected that springs with surface discontinuities such as the seams noted in the 4RB and 6FB springs to have a fatigue life less than 10 million cycles. Of the remaining ten springs removed from the 2CD EDG for visual inspection and magnetic particle inspection, none were damaged or showed signs of surface discontinuities, supporting the conclusion that any flawed springs would have failed by this point in their life. This conclusion would be equally applicable to the springs in the 2AB EDG, which have experienced approximately 23 million stress cycles, and is even stronger for the Unit 1 AB and CD EDGs, whose springs have experienced approximately 33 and 34 million stress cycles, respectively. Analysis of Event This event is being reported in accordance with 10CFR21.21 (c)(3)(ii) and 10CFR50.73(a)(2)(v). The event had an adverse effect on the ability of the 2CD EDG to carry full load for the period assumed in the safety analyses. Due to this adverse, but unquantifiable, effect on the engine, it was determined that a substantial safety hazard could exist, and a 10CFR Part 21 report was appropriate. 10CFR21.21 (c)(3)(ii) requires a written report within 30 days following receipt of the information by the responsible corporate officer on the identification of a defect or failure to comply. The written report is required to include the following information: (i) Name and address of the individual informing the Commission: Eugene E. Fitzpatrick Vice President, Indiana Michigan Power 1 Riverside Plaza Columbus, OH 43215-2373 (ii) Identification of the facility which fails to comply or contains a defect: Donald C. Cook Nuclear Plant, Unit 2 Indiana Michigan Power Emergency Diesel Generator (EDG) 2CD TEXT PAGE 4 OF 6 Analysis of Event (cont'd) (iii) Identification of the firm constructing the facility or supplying the basic components which contains a defect: Worthington Compressor now owned by Dresser Rand Co. Buffalo, New York 100 Chemung Street Painted Post, NY 14870 (iv) Nature of the defect and the safety hazard which is created or could be created by such defect: The Unit 2CD EDG number 4 rear bank (4RB) cylinder fuel injection pump cam follower spring failed. This spring has surface discontinuity, or seam, running for at least one coil parallel to the axis of the wire down to the corner of the fracture surface. The catastrophic failure of this spring caused the aluminum cam follower to seize. The resulting loss of power on the 4RB cylinder represents an approximately 8.33 percent loss. The engine governor would automatically feed the remaining 11 cylinders extra fuel to compensate for the 4RB cylinder loss. Tests have been conducted which indicate that the EDGs can sustain a 10 percent overload. Therefore, it is believed that the remaining 11 cylinders on the EDG could have performed their intended function for at least two hours. The spring failure and subsequent 4RB cylinder power loss would, however, have an adverse, but indeterminate, effect on the EDG to carry a full load for the period assumed in the safety analysis. (v) The date on which the information of such defect was obtained: June 19, 1996 (vi) In the case of a basic component which contains a defect, the number and location of all such components in use at the facility subject to the regulations in this part: The Donald C. Cook Nuclear plant has four 12 cylinder EDGs, for a total of 48 springs. (vii) The corrective action which has been, is being, or will be taken; the name of the individual or organization responsible for the action; and the length of time that has been or will be taken to complete the action: This information provided under the "Corrective Action" section of this LER. (viii) Any advice related to the defect that has been, is being or will be given to purchasers or licensees: None TEXT PAGE 5 OF 6 Analysis of Event (cont'd) In accordance with the guidance present in NUREG 0302, Revision 1, "Remarks Presented at Public Regional Meetings to Discuss Regulations (10CFR Part 21) for Reporting of Defects and Noncompliance", regarding reportability of Part 21 events as LERs, it was determined that this event was conservatively reportable under 10CFR50.73(a)(2)(v) as a condition that alone could have prevented the fulfillment of the safety function of structure or systems that are need to: (A) Shut down the reactor and maintain it in a safety shutdown condition; (B) Remove residual heat; (C) Control the release of radioactive material; or (D) Mitigate the consequences of an accident. This event occurred while the unit was shutdown and defueled. With the unit in this configuration, in Mode 6, there is no requirement for the EDGs to be operable. Subsequent evaluation of the failure of both the 4RB and 6FB springs led to the conclusion that the fracture of the springs had occurred early in their life cycle. During the test run on April 13, 1996 the 4RB cylinder seized, even though the spring is assumed to have been broken for some period of time. Previous surveillance tests on the Unit 2CD EDG were successful, and supports the operability of the EDG up until the time that the cylinder noise was heard. It has been concluded that the unit was in an inherently safe, defueled condition in which the safe operation of the unit was not affected. The event is concluded to be of minimal safety significance, and one which posed no threat to the health or safety of the public. Corrective Action The remaining 10 springs on the Unit 2CD EDG were examined by magnetic particle examination. No surface defects of cracks were discovered. Replacement springs, cam followers, and cover plates were obtained to replace those that were damaged. All parts were cleaned and reinstalled. The engine was operated a slow speed, then at half load as combustion pressure were checked. A full load run was made and pressure again checked. All parameters were satisfactory. Acoustic monitoring was performed on the 2AB EDG. The acoustic monitoring revealed an anomalous noise in one of the cam followers. The cam follower was disassembled and inspected, however, no problems were found. The Unit 1 EDGs were run at reduced load while acoustically monitoring the cam followers. This monitoring was performed using a mechanical stethoscope, an acoustic monitor and a vibration monitor. No abnormal noises were identified in any of the cam followers, and all exhibited similar acoustic signatures with a strong, clear signal. The vibration readings were consistent between cylinders, ranging between 0.5g and 1.7g on the 1CD engine and 0.4g and 1.8g on the 1AB engine. The set screws for the spring cover plate had been found to be loose on the 2CD EDG 4RB follower, therefore an inspection plug at each follower bore was removed prior to the engine runs so that the plate and setscrews could be observed during the runs. All were found to be tight, and had maintained the plate in position on top of the follower bore. As a further preventive action, the cam follower springs from the 1AB, 1CD and 2AB EDGs will be removed to inspect them or surface discontinuities at the next outage of sufficient duration. TEXT PAGE 6 OF 6 Failed Component Identification None Previous Similar Events None ATTACHMENT TO 9607240244 PAGE 1 OF 1 American Electric Power Cook Nuclear Plant One Cook Place Bridgman, MI 49106 616 465 5901 AEP AMERICAN ELECTRIC POWER July 19, 1996 United States Nuclear Regulatory Commission Document Control Desk Rockville, Maryland 20852 Operating Licenses DPR-74 Docket No. 50-316 Document Control Manager: In accordance with the criteria established by 10 CFR 50.73 entitled Licensee Event Report System the following report is being submitted: 96-008-00 Sincerely, A. A. Blind Site Vice President /mbd Attachment C: H. J. Miller, Region III E. E. Fitzpatrick P. A. Barrett R. F. Kroeger S. J. Brewer M. R. Padgett G. Charnoff, Esq. D. Hahn Records Center, INPO NRC Resident Inspector *** END OF DOCUMENT ***

Page Last Reviewed/Updated Friday, January 31, 2020