IR 05000272/1993029
| ML18100A910 | |
| Person / Time | |
|---|---|
| Site: | Salem, Hope Creek  |
| Issue date: | 01/31/1994 |
| From: | Lohmeier A, Modes M NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML18100A909 | List: |
| References | |
| 50-272-93-29, 50-311-93-29, 50-354-93-30, NUDOCS 9403080129 | |
| Download: ML18100A910 (11) | |
Text
U. S. NUCLEAR REGULATORY COMMISSION
REGION I
DOCKET/REPORT NOS:
50-272/93-29 50-311/93-29 so.:-354/93-30 LICENSE NOS:
LICENSEE:
DPR-70 DPR-75 NPF-57 Public Service Electric & Gas Company 80 Park.Plaza - 17C
. Newark, New Jersey FACILITIES:
PSE&G Company Research and Testing Laboratory Salem 1 & 2 and Hope Creek Stations INSPECTIONS AT:.
Mapleview, New Jersey Hancocks Bridge, New Jersey INSPECTION DATES:
December 3, 1993 December 13-17, 1993 INSPECTOR:
Alfred.~hmei~ Sr. Reactor Engineer Matena'l.s Section, EB, DRS
APPROVED BY: $Me&~
Michael C. Modes, Chief Materials Section, EB, DRS 9403080129 940224 ;_
PDR ADDCK 05000272 O
PDR 1-:J-'1 -. f rfue Date
. 2 Areas Inspected: The inspection scope included the root cause investigation of Salem EDG cylinder liner failure and operational cycle data retention _and evaluation at Salem and Hope Creek Nuclear Power Generation Station *
Results: The EDG failure investigation was comprehensive. The Salem Units 1 and 2 transients oceur at the same rate as that predicted. Two regions of Salem reactor pressure vessels have significant lifetime cumulative usage factors: vessel head closure studs and bottom head instrument penetrations. Salem engineering does not retain steam generator and pressurizer stress reports. Salem has no written procedure providing for engineering review of the monitoring data at regular intervals. Hope Creek operational transient records are being retained and monitored. Some Hope Creek transients are being applied at a greater rate than anticipated in the design of the plant. Hope Creek has a procedure for implementation of a thermal monitoring program for cyclic operational transients.. Hope Creek is developing an on-line transient and fatigue monitoring system.
DETAILS SCOPE OF INSPECTION (37700)
The soope of this inspection included. evaluation of the emergency diesel ge~erator cylinder liner failure investigations at the PSE&G Research and Testing Laboratory and preparation by Salem engineering of a root cause evaluation report. The inspection also included evaluation of data *retention of operational cycles in compliance with technical specification requirements and evaluation of the procedures and data retained for Salem and Hope Creek Nuclear Power Generation Station.0 El\\.IBRGENCY DIESEL GENERATOR ENGINE CYLINDER LINER FAILURE On December 2, 1993, with Unit 2 at 100 percent power, the 2C emergency diesel generator (EDG) was removed for maintenance not related to the power train. Several hours after completion of a post maintenance operability surveillance test, water was discovered in the vicinity of the 3R cylinder. On December 3, 1993, after discussions with the NRC, PSE&G began removal of the 3R cylinder and root cause analysis and repair project teams were formed. Unit 2 began a shut down and NRC was notified under 10 CPR 50.72; an incident report was initiated. On December 4, 1993, a failed liner sample was sent to the Public Service Electric and Gas (PSE&G) Research and Testing Laboratory (RTL). Inspection of Failed EDG Engine Cylinder Liners On December 8, 1993, the inspector visited the PSE&G Company RTL, to assess the metallographic inspection program and examine the failed EDG cylinder liner prior to extensive dissectio The following cylinder liners were available for inspection:
- N Cylinder Liner Manufacturer Serial Number (1)
3R (failed)
CAD*
5T-24-1-81-6 (2)
3R (original)
ALCO*
7271-9-71 (3).
SR (failed)
CAD ES-29-31-1 (4)
7R ALCO/CAD 7271-5-20 (5)
Never Installed CAD 5-21-1-6 (6)
Never Installed CAD E4-29-2-5
- CAD - Canadian Allied Diesel Company, ALCO - American Locomotive Company Parts 3R(l) and 5R(3) were undergoing metallographic examination at the time of the *
inspector's visit. Part 7R(4) was taken out of its shipping box and examined by the inspecto Parts (4) and (5) were liners that had never been installed in a diesel engine.
The Diesel engine had 18 pistons in 2 lines of 9 pistons each. The designation R refers to the right side of the engine, looking from the generator end. The liner number indicates the piston number beginning at the side opposite that of the generator. *
Examination of 3R(l) indicated that the cylinder liner flange had fractured completely around the circumference. It was removed in 3 segmental pieces around the circumference. The fracture surface began at the notch on the outside surface of the cylinder liner and proceeded at a 45 degree angle toward the inner surface of the cylinder liner. CoinCidentally, 3 radial through wall cracks were found which divided the failed flange into the three segments. A fourth radial crack proceeding from the outside of the cylinder toward the inside, but not through, was visibly noted. PSE&G personnel believed the circumferential fracture preceded the radial fractur Examination of the fracture surfaces of 3R(l) by the inspector indicated a brittle type of fracture surface (typical of cast iron) with only slight evidence of a starting point at one location around the cylinder. The inspector could not clearly discern any fatigue striations or brittle fracture chevron indications. The inner surface showed a small shining edge around the inner circumference that was indicative of the final fracture surfac Initial materials analysis indicated the material to be gray cast iron. The inspector did not review any further test data. Further analysis will be performed of 3R(l) to characterize the materials propertie Examination of the radial fracture surface of the cylinder liner flange indicated a surface similar to that of the circumferential crack. The crack began at the lower end of the cylinder liner flange in the upper cylinder groove. The fracture surface color ranged from black (with rust colored areas) at the start of the crack, changing to dark gray-yellow, and finally to a shining gray at the end of the crac Examination of the inner surfaces of the cylinder liners by the inspector revealed only highly polished surfaces along the length of piston travel. The outer surfaces of the cylinder liners showed no signs of distres Examination of the fracture surfaces of 5R(3) by the inspector indicated a 300 degree fracture surface in the same plane as that for 3R(l).
PSE&G laboratory personnel reviewed the RTL metallography program with the inspecto The program included photographing the piece parts, obtaining photomacrographs and photomicrographs of fracture surfaces, visual examination, black light inspection, and material analysis. Radiography (RT) could not be done at the laboratory because RT facilities were not available at RTL.
The inspector was informed by the *laboratory manager that the results of the materials investigation would be provided to the root cause investigation team at the Salem Nuclear Power Generation Station. The inspector found the PSE&G RTL investigation was proceeding carefully and the laboratory personnel to be knowledgeable in materials failure investigative technique.2 Review of the EDG Cylinder Liner Root Cause Program On December 13, 1993, the inspector reviewed the root cause determination for the EDG engine cylinder liner failure as reflected in a preliminary outline of the root cause report to be presented to NRC at the conclusion of the investigation. The report included background (problem description, manufacturing history, cause of replacement), root cause investigation methodology, description of failure event, areas investigated (operations, installation, materials investigation and evaluation), conclusions, and operability evaluation of failed unit and other similar unit PSE&G personnel discussed details of the preliminary outline report with the inspector. The inspector found that PSE&G believed a possible root cause factor to be unevenness or misalignment of the cylinder and flange surface causing high stresses at the stress concentration point under the flange at the juncture of the flange to cylindrical portion of the cylinder liners. The liner material is gray cast iron and is much more sensitive to fracture at stress concentration points or defects. The stress concentration in the CAD design appears to be higher than that for the ALCO design because the undercut radius in the CAD liner is smalle PSE&G has retained a specialist consultant to perform finite element analyses on the flange -
cylinder juncture (where fracture occurred) to determine the effects of flange mating surface unevenness, stress concentration for different fillet sizes, and bolt tightnes The inspector found that PSE&G has initiated a comprehensive investigative program to determine the root cause of EDG cylinder liner failur.0 CYCLIC OPERATIONAL DATA RECORDING AND MONITORING Background The primary system components are designed to meet the requirements of Section III of the ASME Boiler and Pressure Vessel Code for Nuclear Vessels. The Code requires a design by analysis approach to evaluate whether the components can sustain the prescribed steady state -
pressure and thermal loadings and the cyclic application of these loadings.
The utility (owner of the components) specifies the types and numbers of application of loadings which are anticipated during the plant lifetim ~oiilponents *are designed in*
accordance with these specifications. Therefore, in the case of cyclic loading, the specification will state the numbers and types of transient operi;ttion that can be anticipated throughout the plant lifetime. These transients are described. in the *Technical Specification (l'S) for the plant. In the case of Units 1 and 2 of Salem, the component cyclic or transient limits are given in Table 5.7-1 of the TS. In the case of Hope Creek, the component cyclic or transient limits are given in Table 5. 7.1-1 of the TS. Operation of the components within these transient limitations is required for TS complianc Since primary system components are designed to sustain limited numbers of transients, the plant technical specifications (TS) reflect the requirement that records and documents relating to the cyclic operation of the plant must be maintained throughout the plant licensed lifetim The transient record retention requirements for Salem Units 1 and 2 are given in 6.10.2 (e)
of the TS. The transient record retention requirement for Hope Creek is given in 6.10.3 (e)
of the T The criteria for exhaustion of fatigue life of a component is reflected in a cumulative usage factor (CUF), which is the summation of the ratio of expected numbers of cycles at the applied strain range to the cycles at that strain range necessary to cause fatigue failu.re. An appropriate factor of safety in terms of strain level or cycles is utilized in the same sense as a factor of safety for stress level* in relation to fracture stress.* The ASME Boiler and Pressure Vessel Code,Section III for Nuclear Vessels'limits the lifetim~ Cumulative Usage Factor to There is a tacit assumption that if the plant operation is within the bounds of the prescribed operating transient limits described within the TS, the lifetime cumulative usage factor will be less than 1.0. This is borne out by the stress analysis and fatigue evaluation performed on the component design details. If, however, the transient operation bounds are exceeded, either by greater transient strain excursions or application of a greater number of transients, the evaluation must be reconsidered. Most regions of the primary components have been designed with cumulative usage factors less than 1.0 when subjected to the strain levels and numbers of cycles prescribed in the technical specification. These regions are not critical when the prescribed transients are exceeded. It is the regions having design cumulative usage factors approaching 1.0 that must be monitored carefully for signs of fatigue deterioration when the cumulative usage factor exceeds.2 Salem Nuclear Power Generation Station - Units 1and2 the inspector reviewed the monitoring of operating cycles for units 1 and 2 to assess the status of cyclic life expenditure for the transients listed in Table 5. 7.1 of the T 'Shown in Attachment B is a comparison of design cycles, expected portion of design cycles expended to date, and actual cycles expended to date.
Unit 1 has operated 17 years of its 40 year design lifetime. The irispector found, with the exception of the reactor trip transient, the rate of transient expenditure of Unit 1 is equal to,.
or below' the expected rate.. 174 reactor trip. transients have occurred,. slightly above. the expected 170 transients expecte Unit 2 has operated 12 years of its 40 year design lifetime.* The inspector found, with the exception of the reactor trip transient, the rate of transient expenditure is equal to, or below, the expected rate. 122 reactor trip transients have occurred, slightly above the 120 transients
- expecte *
-
The inspector reviewed the Combustion Engineering reactor vessel stress reports CENC 1148 (September 1970) and _CENC 1155 (July 1971) for units 1 and 2, respectively. The reports were signed a'nd reviewed by a professional engineer, as required by the ASME Boiler and Pressure Vessel Code,Section III, for Nuclear Vessels. These reports provide for the calculation of cumulative usage factor for the evaluated regions within the reactor pressure vessel. The 40 year lifetime cumulative usage factors for the reactor pressure vessels of Units 1 and 2 were reviewed by the inspector.. The inspector found that only two regions of the reactor pressure vessel have significant lifetime cumulative usage factors (greater than 0.1). The vessel head closure studs have a lifetime cumulative usage factor of 0.40 for Unit 1 and 0.32 for Unit 2. The bottom head instrument penetrations have a lifetime cumulative usage factor of 0.10 for Unit 1 and 0.074 for Unit 2.
The Salem engineering department did not have copies of the steam generator and pressurizer stress reports available for review of cumulative usage factors in critical regions of these components, such as the tubesheet - shell - channel head complex, divider plate, primary nozzles, steam outlet nozzle,. tube to tubesheet welds, and pressurizer surge and spray nozzles. As a result of this, no justification of the increased number of reactor trip transients could be reviewed by the inspector. It is requested, as an unresolved item URI 272/93029-01, that justification by engineering be provided, allowing the increased number of cycles over the design number of cycle The inspector reviewed the collection process for data used in monitoring the transient operation of the plant. A plant engineer collects the operating data on a regular basis and compares the data with prescribed monitoring data obtained from the TS.. At a prescribed fraction of the total number of design cycles, the plant engineer will report the event to the engineering function responsible for review of transient operatio.3 Hope Creek Nuclear Power Generation Station The inspector.reviewed the monitoring of operating cycles at Hope Creek to assess the status of cyclic life expenditure for the transients listed in Table 5.7.1-1 of the T The inspector found operational transient records are being retained and continually monitore The inspector reviewed a graphic portrayal of the transient cycle histogram of
.
.
normal, upset, emergency, faulted, and additional normal and upset conditions utilized in. *
forming the basis for cyclic operating condition monitoring. These transients are more comprehensive descriptions of the plant cyclic operation than that portrayed in Table 5. 7.1-1. *
In reviewing the transient operating history to date of those transients specified in Table 5. 7.1-1, the inspector found that some transients are being applied at a greater rate than anticipated in the design of the plan The heatup and cooldown transients experienced to date, after 7 years operation are 44 cycles. Compared with the 120 cycle design lifetime, plant operation has used 37 percent of the heatup and cooldown transients. This would indicate that, at the end of 40 year life, plant operation would use 210 percent of the 120 design life heatup and cooldown transients, or 252 cycles. A similar case can be seen from the reactor trip transients, where 7 years operation has resulted in 46 reactor trip transients. At the same transient frequency,*
146 percent of the 180 design life transients, or 262 reactor trip transient In the case of the heatup and cooldown transients, or the reactor trip transients, the fatigue evaluation of component parts of the reactor is important, in that the design cumulative usage factor must be low in order to sustain the increased number of operating transients. Hope Creek did not have available the stress reports and fatigue evaluation of the reactor pressure vessel components for review by the inspector. As a result, no justification of the increased numbers of cycles over the design cycles for the heatup/cooldown and reactor trip transients could be reviewed by the inspector. It is requested, as an unresolved item URI 345/93030-01, that justification by engineering be provided, allowing the increased numbers of cycles over the design numbers of cycle The inspector found that Hope Creek has prepared a Procedure HC.TE-PR.ZZ-013 (Q) for implementation of a thermal monitoring program for cyclic operational transients. The procedure provides for monitoring the operating transients limited by the technical specification Table 5. 7.1-1 together with an additional ti:ansient related to loss of feedwater heating. In addition to recording and evaluating these transients, the procedure provides for specific notification of engineering management of the monitoring results such that corrective action may be taken to -alleviate the effects of excessive transient severit The inspector found: that Hope Creek is implementing a proactive program to develop an on-line transient and fatigue monitoring system. The inspector attended a vendor presentation of one such system under consideration by Hope Creek which has been installed in many PWR and BWR plants throughout the United States. The concept utilizes sensors applied to a structural component which provide specific data for a pre-analyzed solution to a stress and fatigue evaluation problem for that structur *
9 CONCLUSIONS
The PSE&G RTL EDG cylinder liner failure investigation is.proceeding carefully and the laboratory personnel are knowledgeable in materials failure investigative technique *
PSE&G has initiated a comprehensive investigative program to detemtlne the root cause of EDG cylinder liner failur *
The rates of Salem Units 1 and 2 transient expenditure are equal to, or below, the expected rat *
Two regions of the Salem reactor pressure vessels have significant -lifetime cumulative usage factors (greater than 0.1): vessel head closure studs and the bottom head instrument penetration *
Salem and Hope Creek operational transient records* are being retained and
continually monitore *
Some Hope Creek transients are being applied at a greater rate than anticipated in the design of the plan *
Hope Creek has prepared a procedure for implementation of a thermal monitoring program for cyclic operational transient *
Hope Creek is implementing a proactive program to develop an on-line transient and fatigue monitoring syste.0 MANAGEMENT MEETINGS The inspector met with Salem and Hope Creek representatives at the entrance meeting on December 13, 1993, and at the exit meeting on December 17, 1993, at the Salem Nuclear Power Generation Station in Hancocks Bridge, New Jersey. The names of PBAPS personnel
- contacted is shown on Attachment The findings of the inspection were discussed with licensee management at the December 1993 exit meeting. PSE&G personnel did not disagree with the findings of the
. inspector.
ATTACHMENT A*
PSE&G Research and -Testing Laboratories (RTL) Meeting
. *-~-
In attendance during the visit were the following PSE&G personnel:
C. Dzuba D. P. Hauth G. C. Moran H. L. Onorato N. C. Roy Senior Test Engineer, RTL _Metallurgy *
Senior Staff Engineer, QA and Procurement Manager, RTL Materials Division Senior Staff Engineer, Licensing and Regulation Senior Engineer, RTL Metallurgy Entrance and Exit Meeting Attendance List Salem Nuclear Power Generation Station
- M. Morroni K.Pike E. Villar C. A. Vondra
- E. Zegarowicz Teehnical Manager Technical Engineer, Reactor/Plant Performance Station Licensing Engineer General Manager, Salem Operations Engineer, Technical Department Hope Creek Nuclear Power Generation Station
- J. Clancy C. Fuhrmeister
- M. Gray P. Opsal Site Representatives
- M. Sesok Technical Manager System Engineer Engineer, Licensing Dept. Head, System Engineering, NSSS Site Representative, Atlantic Electric United States Nuclear Regulator_y Commission R. Summers Project Engineer Those names with an asterisk (*)attended the exit meeting on December 3, 1993.
ATTACHMENT B COMPARISON OF DESIGN - EXPECTED - ACTUAL OPERATIONAL CYCLES Transient Heatup Cooldown Loss of Load Loss of AC Power Loss of Flow One Loop
- 40 Year Unit 1 Design Expected Cycles Cycles 200
200
80
40
80
Unit 1 Unit2 Unit 2 Actual Expected Actual Cycles Cycles Cycles
60
47
48
24
17
6
24 O_
Large Load 200
17
3 Step Increase RC Pipe Break OB Earthquake DB Earthquake 2485 psig primary leak test 3107 psig primary leak test 1356 psig secondary leak Turbine Roll
0
21
4
21
2
-2
4
0
0
0
3
1
1
1
1
0
3 0