ML20044F855
| ML20044F855 | |
| Person / Time | |
|---|---|
| Site: | Vermont Yankee File:NorthStar Vermont Yankee icon.png |
| Issue date: | 05/18/1993 |
| From: | Drysdale P, Eapen P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20044F850 | List: |
| References | |
| 50-271-93-09, 50-271-93-9, NUDOCS 9306010126 | |
| Download: ML20044F855 (9) | |
See also: IR 05000271/1993009
Text
{{#Wiki_filter:: ' . i . U.S. NUCLEAR REGULATORY COMMISSION , REGION I Docket No. 50-271 , Report No. 50-271/93-09 , License No. DPR-28 Licensee: Vermont Yankee Nuclear Power Corporation
t RD 5, Box 169 Brattleboro, VT 05301 { , Facility Name: Vermont Yankee Nuclear Power Station i In'spection at: Vernon, Vermont and King of Prussia, Pennsylvania l Inspection Conducted: April 14 - May 11,1993
Inspectors: P. D. Drysdale, Sr. Reactor Engineer, RI
T. J. Kenny, Sr. Reactor Engineer, RI B. R. Whitacre, Reactor Engineer, RI
R. E. DePriest, Reactor Engineer Intern l
O m [/
- 7b3
Izad Inspector: P. D. Drysdale, Sr/ Reactor Engineer, Date Systems Section, EB, DRS l /[I f3 Approved By: . . / ' Dr. P. K. Eapen, Chief, Systems Section - Date .; Engineering Branch, DRS > -{ ! ! ' 9306010126 930524 ! PDR ADOCK 05000271 G PDR i
. . . . i h . 2 i Insnection Summary: . Areas Inspected: The licensce's evaluations and corrective actions for control rod average , scram insertion times that exceeded the Technical Specification limits.
e ! Results: The licensee's analysis and actions to correct plant equipment deficiencies were j adequate for returning the plant to power operations. An apparent violation was identified ! for a concern that no corrective actions were taken for six months after it was determined that the average control rod scram time of 0.391 seconds for one two-by-two array exceeded l the Technical Specification limit of 0.379 seconds. The Vermont Yankee plant was operated outside the Technical Specification limit for this entire period. One unresolved item was identified in the area of procurement, control of service life limited items and receipt inspection practices related to the use of ASCO solenoid operated pilot valves. A second unresolved item related to design control practices for modifications made to ASCO solenoid ! scram pilot valves installed in the plant was also identified.
.! I i , 9 -
1 . . DETAIIS 1.0 CONTROL ROD INSERTION TIME IN EXCESS OF TECHNICAL i SPECIFICATIONS 1 ' The control rod drive (CRD) system assures that the reactor will remain suberitical by driving in the control rods within the specified insertion time limit. The actual scram times for all control rods are determined during each operating cycle. The Technical Specificadons require that within a 16 to 32 week interval,50% of the control rods in all four reactor quadrants be surveillance tested to determine the core-wide average scram time and the average scram times for control rod two-by-two arrays. The maximum average times permitted for those tests are specified in Section 3.3.C. of the Vermont Yankee Technical Specifications. A post-test evaluation is required to provide reasonable assurance that proper control rod drive performance is being maintained. The results of measurements performed on the control rod drives are submitted in each startup test report. The purpose of this inspection was to assess the licensee's evaluations and corrective actions ta resolve control rod scram insertions times that were greater than Technical Specification (TS) limits during surveillance tests conducted in October 1992 and April 1993. The average j scram insertion times for one two-by-two array was greater than the applicable Technical Specification limit since October 1992. The facility's Technical Specifications require an immediate reactor shutdown when the average scram time for a two-by-two array exceed the specified limit. However, the licensee did not shut the reactor down when the average scram time for one two-by-two array exceeded the Technical Specification limit. In April 1993, the average scram times for seven two-by-two arrays and the core-wide c average were in excess of the Technical Specification limit. The condition was identified to . NRC and a request for enforcement discretion was made. The plant was shut down for an ' unrelated problem. Inspectors reviewed the licensee's corrective actions for adequacy prior to returning the plant to power operations. 1.1 Description of Events , On October 15,1992, reactor power was reduced to approximately 70% for a scheduled ' bi-monthly control rod pattern exchange and individual rod scram time surveillance testing in accordance with TS Section 4.3.C.2. The core-wide average insertion time for the first - ' 4.51% of rod motion (i.e., "from notch 48 to notch 46") was 0.340 seconds and was acceptable to meet the TS required time of 50.358 seconds. However, one two-by-two array resulted in an average insertion time to notch 46 of 0.391 seconds which did not meet the TS , required time of s0.379 seconds. The licensee initiated a Potential Reportable Occurrence (PRO #92-083) to review the need to report the unsatisfactory test results to the NRC. As a ! ' result of their review, Reactor Engineering and Computer Department personnel concluded that the two-by-two array average scram time was outside the scope of TS Section 3.3.C.3 , and, consequently, did not require an immediate shut down of the reactor upon determination -
. 1 h 4 that the average scram time was deficient. In addition, the licensee did not initiate a root cause determination of the deficient equipment performance or pursue corrective actions required to assure continued adequate performance. The licensee's management concurred with the conclusion in the PRO and determined that a shutdown was not necessary and that the deficient scram times did not have to be reported to the NRC. On April 6,1993, the licensee performed another scheduled control rod pattern exchange and scram time surveillance test. The core-wide average time of 0.369 seconds for the rods tested did not meet the TS maximum limit of 0.358 seconds. Also, seven two-by-two arrays average times (ranging from 0.380 to 0.405 seconds) did not meet the TS maximum limit of 0.379 seconds. The licensee initiated a new PRO to determine the reportability of these , ' results. The licensee recognized on April 12, 1993, that the test results from October 1992 should have been reported and that it may have been necessary to shut the plant down. At 7:11 p.m. on April 12, 1993, the licensee made a 10 CFR 50.72 call and reported the failure to shut the reactor down when the scram time for one two-by-two array exceeded the Technical Specification 3.3.C limit on October 15, 1992. The PRO was converted to a Licensee Event Report (LER #93-005) after concluding that the NRC must be notified of the high core-wide average time. TS Sections 3.3.C.I.1 and 3.3.C.I.2 contain the specifications for both the core-wide average and the two-by-two array average scram insertion times. The Section 3.3.C.3. Limiting Condition for Operation (LCO) states "if Specification 3.3.C.1.2 cannot be met, if operating, the reactor shall be shut down immediately upon determination that [the] average scram time is deficient." In addition, the Acceptance Criteria and Final Conditions stated in the licensee's test procedure OP-4424, " Control Rod Scram Testing and Data Reduction," requires a plant shutdown if the scram time results are not within the limits specified by TS Section 3.3.C.1.2. Since surveillance test results obtained on October 15, 1992, did not ! satisfy the maximum insertion time to the notch 46 position for one two-by-two array as specified in both TS Sections 3.3.C.1.1 and 3.3.C.1.2, the licensee was required to take the actions specified in Section 3.3.C.3. In addition,10 CFR 50, Appendix B, Criterion XI, " Test Control," and Criterion XVI, " Corrective Action," require that deficient test results must be evaluated and measures must be taken to assure deficient conditions are corrected. However, the reactor was not shut down and was operated continuously at power until April 7,1993, when the reactor was shut down due to an unrelated feedwater system leak. Failure to shutdown the reactor as required by the TS and to correct deficient control rod conditions is an apparent violation of NRC requirements (EEI 50-271/93-09-01). After the test results of Apri! 6,1993, were obtained, the licensee requested that the NRC
provide enforcement discretion for the TS required actions after deficient scram times were , ' discovered at notch 46. The request proposed a temporary time limit of 0.425 seconds for the core-wide average and 0.450 seconds for the two-by-two arrays everage to be used for _' the remainder of the current operating cycle. The licensee did not consider the high average , I w
- , . 5 scram time test results to be safety significant based upon the safety study performed by ' Yankee Nuclear Services Division which concluded that notch 46 insertion times could increase to 0.500 seconds without impact on the ability of control rod system to protect against fuel damage. Interim disemtion was granted for 48 hours by NRC on the condition ' that the licensee must submit a written request to document their justification for the , enforcement discretion and to demonstrate that the slower scram times did not constitute a
significant degradation of the CRD system. However, the enforcement discretion was not , imp!cmented as the plant was shut down due to an unrelated concern. On April 13, 1993, the NRC initiated an independent inspection to review the causes of the
failed scram times and related factors, and to review the licensee's evaluations and corrective actions to recover proper scram insertion times prior to returning the plant to power ' operations. The licensee indicated that investigations would be ongoing to analyze the slow scram times and that results would be providu! to the NRC prior to plant startup.
1.2 Short-term Actions The licensee chartered three task teams to troubleshoot the slow scram times, to isolate the specific equipment problems responsible for the slow scram times, and to examine the basis for the TS scram time requirements. These teams were composed of General Electric (GE),
Automatic Switch Company (ASCO), Yankee Atomic Electric Company (YAEC) and
Vermont Yankee personnel who were required to provide recommendations to plant , management to resolve all equipment issues and to investigate programmatic issues [ contributing to the events described above. , ! One team investigated various CRD conditions that could potentially cause slow scram times ! such as drive line friction, system hydraulics, air entrapment, valve blockage, valve leakage,
, internal CRD flow blockage, instrument air parameters, control rod weight and accumulator ( pmssure. l Based on subsequent analyses, the slow scram times appeared only from notch 48 to 46 and t implied that the problem was only in the CRD's Start of Motion (SOM) region, i.e., the time l from deenergization of the Scram Solenoid Pilot Valves (SSPVs) to the beginning of actual ! rod drop-out movement from position 48. This distance represents only a fraction of an inch l of rod travel. The apparent cause was then narrowed down to the CRD Hydraulic Control , ! Units (HCUs). The task team investigated physical characteristics that may affect scram times to determine if a correlation existed between the SSPV physical characteristics and the scram response time. New replacement parts kits for SSPVs were procured for pilot head and air valve and installed on about half of all HCUs. Tests performed at cold hydrostatic l pressure (=950 psi) resulted in a significant improvement in scram times. The SSPVs were ! I replaced on all remaining SSPVs. This resulted in a significant improvement in the notch 46 scram times that were well within the TS limits. The task team's final root cause and test results were being documented in accordance with the licensee's corrective action process; ' however, as of this inspection, the results were not yet finalized for NRC review. , , i , = . - , 5
-_ . 4 6 The licensee performed surface hardness tests on the SSPV air port diaphragms, but the results did not correlate diaphragm stiffness directly with scram response time for particular SSPVs. Four SSPV assemblies were sent to a GE laboratory for detailed testing and material analyses. The exact cause was not yet known at the conclusion of this inspection; however, preliminary laboratory results indicated accelerated deterioration of the BUNA-N diaphragms that was induced by thermal degradation. There also appeared to be a close relationship between the degree of thermal degradation and the slow scram response times of the associated SSPVs. Installation testing was completed satisfactorily on all 89 CRDs with the refurbished pilot valves. The licensee then applied cold hydrostatic pressure to the reactor and retested all control rods, with CRD pumps isolated, and performed a detailed review of the test data to validate their analysis and conclusions. Additional testing, equipment replacements, and administrative controls were established prior to plant startup. The licensee also performed additional scram tests of all 89 CRDs during power ascension. A retest of 50% of the CRDs is scheduled during the next rod pattern exchange in June 1993. Prior to returning the reactor to power, the licensee fully demonstrated and assured the capability of all CRDs to meet the TS scram insertion times. The results of the licensee's efforts to investigate and correct the CRD scram time performance discrepancies were presented in a letter (BVY 93-41) to the NRC on April 15, 1993. Based upon the licensee's analysis and corrective actions to restore the CRDs to full performance capability, the NRC considered that the plant equipment was capable of returning to power operations. On April 16,1993, the plant manager issued a memorandum to all departments heads . highlighting the recent course of events leading to an apparent breakdown in the management system that allowed the CRD performance to approach and finally exceed Technical Specification limits. The memo requested that all departments with Technical Specification and surveillance testing responsibilities perform a critical review and re-examination of those responsibilities in advance of an evaluation in these areas requested by the Plant Operations Review Committee (PORC). A PORC follow-up item was established during a prestart-up meeting to require all departments to evaluate their Technical Specification and surveillance test practices and to ensure that margin to Technical Specification limits was monitored and maintained, and that corrective actions and management notifications were made in a timely
manner. Based on the above actions, the NRC considered that the licensee had taken the appropriate actions to prepare the plant and the staff for a return to power. The reactor was made critical on April 16,1993, and later the same day, all 89 control rods were successfully scram tested under hot plant conditions. The core-wide average time was 0.312 seconds, within the TS required time limit. Additional two-by-two array testing will take place in June 1993 during the scheduled rod pattern exchange. . i' l
. . 7 1.3 Procurtment, Receipt Inspection, and Storage of Scram Pilot Solenoid Valves and Replacement Parts % The NRC inspectors examir-d the procurement and receipt inspection activities at the site with respect to the ASCO solenoid va'ves. The records for these valves that documented material requirements and dates related to shelf-life, service-life, valve assembly dates, and BUNA-N cure dates were not consistent. The information available did not show consistent program requirements between records at the site and other records provided by GE and ASCO. The inspectors reviewed VY purchase orders from 1988,1989, and April 1993. , GE apparatus requisition documents stated that the shelf-life of diaphragm kits was 10 years
and the shelf-life of the pilot head valve kits was 6 years. GE Service Information Letter (SIL) No.128 stated that the shelf-life of the diaphragm was 9 years from the cure date of BUNA-N material. A subsequent revision to SIL No.128 indicated that the BUNA-N parts ' should not be used in excess of 7 years beyond the assembly date. I The licensee did not appear to have a dermitive process to assure that the necessary beginning and end dates for the service and shelf-life of all ASCO components at the site
were within the limits established by GE or ASCO. For example, contrary to GE SIL No.128, some component labels in the site warehouse indicated that the material had a 10 year shelf-life beyond the assembly date rather than 7 years recommended by the GE SIL. l Some of the licensee's receipt inspection personnel were not aware that some ASCO part number changes had occurred until after material was received on site. They were also not aware of existing revisions to GE SILs on ASCO valves. It is apparent that ASCO ,' components assembled by GE in 1988 contained BUNA-N material that was cured in 1984, and that these components were installed in 1989, five years into the total nine year life established by GE in SIL No.128. Considering the 7 year limit beyond the assembly date,
this material was approaching its end oflife in 1992 and 1993. The above concerns related to procurement and control of limited service life SSPV spare , parts procurement, receipt inspection and storage remain unresolved (URI 50-271/93-09-02). T 1.4 Modifications to ASCO Pilot Valve Internals Seals r After the 1989 refueling outage, the licensee changed the SSPV body / pilot head interface passage seals material from BUNA-N to Viton to correct an observed premature degradation of the BUNA-N that was causing the scram air header to leak and slowly draw down. The licensee procured commercial grade Viton 0-rings and used these seats to replace the , i BUNA-N seals originally provided by ASCO. This change was made with the expectation that a full six year service life could be realized in these seals. The licensee installed the Viton seals without informing GE or ASCO or obtaining their concurrence prior to the , change. ASCO did not concur with the use of Viton in this location and later recommended ' that the licensee reinstall BUNA-N' seals as originally provided with the valves. ASCO
regarded seal leakage as a good indication that degradation was occurring in all BUNA-N parts and considered that this leakage indicated the SSPVs were nearing the end of their j i ,
- _
. . 8 service life. ASCO indicated that the licensee should have regarded other internal BUNA-N materials in the SSPVs (i.e., the air port diaphragms) as having degraded to the end of their service life and concluded that the SSPVs should have been replaced at that time. The inspectors reviewed the suitability of Viton for these seals with regard to the design basis function of the SSPVs. The licensee indicated that an equivalency evaluation was performed to determine their acceptability for this application. The material change was processed through the licensee's EQ program and a "One-For-One" evaluation. It was not apparent to the inspectors that this change was made with full recognition of all SSPV design requirements that would be reviewed in a modification program applicable to safety-related equipment. This item remains unresolved (URI 50-271/93-094)3). 2.0 UNRESOLVED ITEMS Unresolved Items are matters about which more information is required to ascertain whether they are acceptable items or violations. Two unresolved items identified during this inspection are discussed in Sections 1.3 and 1.4 of this report. 3.0 EXIT MEETING The inspectors discussed the results of the onsite inspection on April 16,1993, with site personnel identified in Attachment 1. The remainder of this inspection was conducted from the Region 1 office and the final results were communicated to licensee management on May 11,1993. The licensee acknowledged the inspection findings as detailed in this report. Some additional information was provided by the licensee during the exit meeting regarding the unresolved items.
. . . -
i 9
! ATTACHMENT 1 Persons Contacted ! Vermont Yankee Nuclear Power Corocration >
- M. Benoit, Reactor and Computer Engineering Manager
, B. Buteau, Engineering Superintendent
- C. Cameron, Senior Reactor Engineer
j R. Current, I&C Engineer j
- J. Herron, Technical Services Superintendent
~ , W. Limburger, Procurement Engineer - 3. Meyer, Project Engineer
R. Pagodin, Operations Superintendent D. Reid, Vice President, Operations
R. Sojka, Operations Suppon Manager
i R. Wanczyk, Plant Manager
Yankee Atomic Electric Company J. Hoffman, Engineering Manager l Others E. Gibo, Lead System Engineer, Genen! Electric Company K. Thomas, Automatic Switch Company United States Nuclear Regulatory Commission . D. Dorman, Vermont Yankee Project Manager, NRR f J. Durr, Chief, Engineering Branch, DRS - H. Eichenholz, Senior Resident Inspector, Vermont Yankee P. Harris, Resident Inspector, Vermont Yankee !
M. Hodges, Director, DRS j
E. Imbro, Acting Deputy Director, DRS E. Kelly, Chief, Section 3A, DRP .; D lew, Project Engineer, Branch 3, DRP l H. Ornstein, AEOD t ' J. Petrosino, Vendor Inspection Branch, NRR ' ' ' Designates those in attendance at the exit interview on May 11,1993. i f i . - . - - - }}