4.0 Summary of Major DCISC Review Topics, 16th Annual Report - July 1, 2005 thru June 30, 2006
4.2 Conduct of Maintenance
4.2.1 Overview and Previous Activities
The DCPP maintenance program has improved since the startup of the DCPP. The initially high and increasing Operating Capacity Factor, from 86 in 1985 to 89.4 in 2002, 97.4 in 2003, 97.5 in 2004, and 97.8 for the DCISC reporting period July 1, 2005 – June 30, 2006 demonstrates that the DCPP maintenance program has been effective. The DCISC has reviewed the DCPP maintenance program, or key elements of it, at public meetings and Fact-finding meetings.
The NRC Maintenance Rule (10 CFR 50.65) issued in 1991, required that commercial nuclear plant licensees monitor the performance or condition, or provide effective preventative maintenance of all risk significant structures, systems and components (SSCs) against licensee established goals. PG&E implemented the Maintenance Rule requirements on all risk significant SSCs, as a basis for its Maintenance Program.
DCPP has not changed on-line maintenance practices significantly in the last few years. Scheduling the maintenance of Systems, Structures, or Components in the 12-week rolling matrix according to their Train/Bus/Set relationship minimizes a large part of the Technical Specification (TS) conflict and risk factor. The on-line maintenance work is scheduled such that risk significant work is minimized and to help the asset teams better schedule their work.
PG&E reorganized the Maintenance Department into Asset Teams in the mid-1990s. The Asset Teams were established to improve the process of maintaining DCPP to reduce costs while maintaining or improving quality. The five teams have been reorganized into four teams, which are now: 1) Turbine Building Team, 2) Nuclear Steam Supply System (NSSS) Team, 3) Control Room/Electrical Team, and 4) Maintenance Support Team. During the previous reporting period, the DCISC reviewed the following maintenance-related areas:
During the previous reporting period, the DCISC reviewed the following maintenance-related areas:
- Work control Program
- Asset Team Update
- Troubleshooting Process Update
The DCISC concluded in previous periods that the Maintenance Program appeared satisfactory.
4.2.2 Current Period Activities
During the current period, the DCISC reviewed conduct of maintenance activities as described below.
- Implementation of On-Line Maintenance
- Meeting with New Maintenance Services Director
- Overview of the Maintenance Department
- Ten-Year In-Service Inspection for Outage 1R13
- Measuring and Test Equipment (MTE) Program Review
- Rigging Problems During 1R13
- Review Progress Made in Troubleshooting Program
Implementation of On-Line Maintenance
The DCISC met with Tim King, Assistant Maintenance Director, Ken Pazdan, Construction Manager, and Mark Frauenheim, Electrical Discipline Lead/Acting Maintenance Director to discuss on-line maintenance at the September 7-9, 2006 Fact-finding Meeting (Volume II, Exhibit D.2, Section 3.10).
DCPP has a Cycle Manager who plans the next two years for major projects and on-line maintenance. DCPP uses Probabilistic Risk Assessment (PRA) and Out-of-Service Risk Assessment Modeling (ORAM) to determine when to do on-line maintenance.
DCPP is giving more attention to the selection of equipment for on-line maintenance than in the past and are now using more detailed planning and scheduling for the equipment. They perform all Technical Specification (TS) required preventive maintenance (PM) and testing on-line. In general, they plan to perform work on a piece of equipment that takes only 50% of time allowed by T.S. On the secondary side of the plant, they perform on-line maintenance where it does not impact reliability.
DCPP uses Probabilistic Risk Analysis (PRA) to determine when to do on-line maintenance. They perform all Technical Specification (TS) required Preventive Maintenance (PM) and testing on-line and, in general, they plan to perform work on a piece of equipment that takes only 50% of the out-of-service time allowed by TS. It appears that PG&E is using the proper selection process to perform on-line maintenance of equipment.
Overview of the Maintenance Department
Mr. Purkis, Maintenance Services Director, reviewed the current organization of the Maintenance Services Department at DCPP at the October 12-13, 2005, DCISC Public Meeting (Volume II, Exhibit B.3) and at the September 21-22, 2005 Fact-finding Meeting (Volume II, Exhibit D.3, Section 3.3).
Following 1R13, DCPP would be returning from a process-based organization (Asset Teams) to a traditional organizational alignment for Maintenance which will divide the organization by disciplines such as electrical, instrumentation and control, mechanical, etc. He commented with the asset team organizational arrangement, DCPP has been challenged to keep enough personnel qualified on differing pieces of equipment. The change will be made in a trial period between Outages 1R13 and 2R14 and then permanently following 2R14. He believes Maintenance can perform more efficiently and more training can be accomplished with a functional organization. He plans to move more work to the back shift from the current four ten-hour day schedule to better support Operations and to achieve higher productivity. He expects to drive down the Maintenance backlog from about 700 to about 300 items, keep overtime down and stay under budget.
Mr. Purkis stated that, in common with other utilities, some DCPP maintenance activities, mainly involving emergent work, will be still undertaken as multiple-disciplined team activities, staffed by senior personnel. Work at the Intake Structure, due to the challenges posed by saltwater corrosion and its physical location, will also be accomplished by a dedicated team. A team will also be formed to address work needed on motor operated valves, which require expertise in electrical and mechanical disciplines, and air operated valves which will have mechanical and instrumentation and control disciplines sharing responsibility.
Changes made to the work control procedures and standards within Maintenance included the establishment of a Daily & Outage Planning Group to focus supervision, provide consistent performance expectations to improve work package quality, and balance work loads to meet work control needs. Performance issues within Maintenance as included pre-outage work completion, benchmarking and quality of the work packages.
Jack Purkis, new Maintenance Services Director, has a substantial amount of experience in the nuclear industry having worked at four other nuclear plants and having achieved the position of Plant Manager. He appeared knowledgeable about the maintenance area and appeared definitive and progressive in his plans to improve it.
Ten-Year In-Service Inspection for Outage 1R13
Mr. Dave Gonzalez, In-service Inspection (ISI) Manager, reviewed the Unit-1 Outage 1R13 reactor vessel in service inspection procedure and results at the November 9-10, 2005 Fact-finding Meeting (Volume II, Exhibit D.4, Section 3.5) and at the February 16, 2006 DCISC Public Meeting (Volume II, Exhibit B.6). This is an American Society of Mechanical Engineer’s (AMSE) Code required inspection of reactor vessel welds and interior surfaces.
The reactor vessel is fabricated from carbon-manganese, plated and welded together. Vessel segments range in thickness from approximately 11” in the nozzle areas to 5” on the bottom head. Ultrasonic examination of the full volume of all the welds and the nozzle attachments to the main coolant system is required by the ASME Code once every ten years and this inspection took place at DCPP during 1R13. The examination requires the removal of the reactor vessel lower internals structure (core barrel), a 325,000 pound highly radioactive structure, to allow access to the vessel interior. Robots are programmed to position ultrasonic transducer arrays on the vessel to interrogate the weld and base materials, searching for indications of defects. Examination procedures and personnel undergo rigorous testing in order to qualify to make these examinations. Data acquisition is by fiber optic link to data analysis, where it is digitally recorded, analyzed and archived for future comparison. This was the third time the U-1 reactor vessel had been ASME inspected and the weld characteristics are trended over the 20-year time period.
Examination of the reactor internals was conducted concurrently with the vessel volumetric examinations, with the internals on the stand to allow access to the outside and underside surfaces of the core barrel, areas not normally accessible for examination. A thorough visual examination was made, and digitally recorded for future reference, of all accessible inside diameter and outside diameter surfaces using a remote controlled submarine with a high resolution camera. No indications of flaws or structural distress were noted.
The results of the in-service inspections performed during 1R13 indicate all welds were examined per ASME Code and procedural requirements. No indications exceeding Code acceptance criteria were detected. Alloy 600 material locations in the nozzle-to-pipe welds and bottom mounted instrumentation tubes are an industry concern. Proactive examination of these bottom mounted instrumentation tubes was performed and no defects were detected. Supplementary examination of all nozzle-to-pipe dissimilar metal welds was also performed and no indications exceeding Code acceptance criteria were detected.
The DCPP Unit 1 Reactor Vessel In-Service Inspection (ISI), performed in Outage 1R13, appeared to have been performed effectively. No indications exceeding ASME Code acceptance criteria were found.
10-Year In-Service Reactor Head Inspection During 1R13: Reactor Head, Bottom and Internals
At the February 15-16, 2006 DCISC Public Meeting (Volume II, Exhibit B.6) Mr. Mike Leger reported on the 10-Year In-Service Reactor Head, Bottom and internals Inspection during 1R13. NRC requires specified inspection requirements for reactor heads at pressurized water reactors. Each plant is required to establish a formula based on reactor head temperature at 100 power and operating years, to establish risk factor for susceptibility to primary water stress corrosion cracking (PWSCC).
Volumetric exams and bare metal visual exams are required each Refueling Outage beginning in the 13th refueling cycle for DCPP. The primary inspection technique used involves time-of-flight diffraction ultrasonics, a volumetric technique, supplemented by eddy current surface examination. The vendor engaged for the inspection previously demonstrated its detection capabilities on the EPRI reactor head mockup and before 1R13, the vendor underwent additional qualifications training. All data acquisition is performed remotely.
The review included a photograph of the vendor’s mockup to demonstrate the difficulties of access to the control rod drive mechanism (CRDM) thermal sleeve-head penetration tube areas and a photo of the modified head shield ring showing the inspection tool in place, and the inspection tooling control system station which serves as the interface between the robot and the trailers outside of Containment from which the inspection commands are sent and the results recorded for later analysis.
During 1R13, all 80 penetrations for Unit-1 were examined volumetrically and there were no indications of PWSCC detected. The reactor top head bare metal visual examine was conducted with a robotic crawler and included a 360-degree exam around all tubes and complete base metal inspection. No indications of pressure boundary leakage were found. There was a slight staining from old boric acid or water spills visible on some tubes but these were found to have no impact on head integrity. The equipment used by DCPP would have discovered the degradation experienced at Ohio’s Davis-Besse nuclear plant, and the degradation at Davis-Besse was discovered through a volumetric examination which discovered cracking in the head penetration tubes and subsequently led to the discovery of a gaping hole on the outside of the reactor head.
DCPP will perform volumetric and bare metal examinations on each of its reactor heads during each Refueling Outage until the heads are replaced with heads of a different material. PG&E’s Board of Directors had approved DCPP’s request to replace both reactor heads during 2009 and 2010.
The DCPP Unit 1 Reactor Head In-Service Inspection (ISI), performed in Outage 1R13, appeared to have been performed effectively. No indications exceeding acceptance criteria were found.
Measuring and Test Equipment (MTE) Program Review
The DCISC met with Derek Bell, MTE Program Manager, to review the Measuring and Test Equipment (MTE) Program at the November 9-10, 2005 Fact-finding Meeting (Volume II, Exhibit D.4, Section 3.7).
As part of Maintenance Services, the MTE Group consists of a supervisor (Bell) ten technicians, and three tool clerks. The MTE Program controls the use and calibration of non-plant measuring devices used to assure that plant parameters are within established values. It also controls the calibration of non-plant radiation monitoring devices that are used to assure that radiation dose and count rates are accurately measured for plant areas and for personnel. Individual maintenance procedures are used to conduct calibrations for each type or model of instrument. The instruments and calibration schedules are contained in databases. Calibration activities are conducted in eight lab facilities in various locations in the plant and in maintenance shop buildings.
DCPP considered MTE Program health to be good based on adequacy of calibration facilities, number of qualified personnel, budget, instrument population and performance, management support, interface between MTE and users (mostly Maintenance), and industry participation. Two issues were (1) attrition of experienced personnel via retirement and (2) age of instruments and availability of equivalent replacements.
The Fact-finding Team reviewed a March 2, 2005 biennial audit of the MTE Program required by the Final Safety Analysis Report (FSAR). The audit was conducted by Quality Verification (QV) to verify that MTE was calibrated to verifiable standards, MTE use was controlled, out of tolerance MTE and standards were evaluated for impact on plant equipment, and MTE Program activities were properly documented. QV concluded that the effectiveness of the DCPP MTE Program was satisfactory and did not identify any significant problems. There were several minor problems identified in Action Requests.
The DCPP Measuring and Test Equipment (MTE) Program appeared satisfactory and in good health. Specific improvements were planned which should enhance the program. The Program Manager appeared knowledgeable and proactive.
Rigging Problems During 1R13
The DCISC met with Mike Gibbons, Mechanical Maintenance Manager, to discuss rigging problems during 1R13 at the March 22-23, 2006 Fact-finding Meeting (Volume II, Exhibit D.7).
DCPP experienced a “near miss” during 1R13 when one of the slings failed as the 17,000 pound rotor of Reactor Coolant Pump (RCP) 1-4 was being inverted from a vertical to horizontal alignment. This caused the rotor to swing and strike the Polar Crane before regaining its vertical axis. No one was injured during this event. One DCPP employee and one contract employee were doing the rigging. The reason for the sling failure was that the sling was cut by the sharp edge of the RCP rotor during the lift. The sling should have been protected with sufficient thickness and strength sling protectors on the sharp edges. If it had been protected, then it would not have been cut by the sharp edge while under load.
A safety tailboard was generated by Learning Services and presented to all Maintenance Services personnel during a scheduled 1R13 Outage Safety Meeting. The tailboard described the incident and lessons learned. There have been numerous other similar industry events which were discussed as well as an INPO Significant Event Notification (SEN). The manufacturer’s requirements regarding sling protection use was also discussed.
Initial training lesson guide “Rigging Fundamentals” has been revised to include this event with the lessons learned/human performance error reduction tools. The manufacturer’s requirements regarding synthetic sling protectors and how to choose the correct type for the lift being performed are stressed. There are several industry events found in this lesson guide to help support this topic.
Rigging is a problem in the industry and it appears that Operating Experience (OE) has not been effective in preventing events at DCPP.
Rigging problems on the Low Pressure (LP) Turbine work were identified on an Action Request (AR) as documentation of rigging safety issue LP rotor retrofit project. The AR was routed to the Projects Group for their comments. The entire hood placement was completed, so there are no unresolved industrial safety issues on the 1R13 outage related to this event. The AR stated the need to capture lessons learned so the job is done safely in 2R13.
A rigging event occurred during replacement of Reactor Coolant Pump (RCP) 1-4 because riggers did not use proper equipment during the lift. No one was injured. DCPP appears to have corrected the improper use of rigging equipment through revision of the training guide and annual training. It appears that Operating Experience (OE) has not been effective in preventing events like this at both DCPP and other plants.
Troubleshooting Program
The DCISC met with Andy Kulikowski, Maintenance Manager, I & C Department and owner of Troubleshooting Program, to review the progress made in the Troubleshooting Program at the March 22-23, 2006 Fact-finding Meeting (Volume II, Exhibit D.7, Section 3.6).
A Quality Verification (QV) audit stated that “The organization continues to be reluctant to enter troubleshooting (TS)”. For example, any data gathering (e.g. voltage or current checks) is considered TS in which the TS procedure MA1.DC10 should be implemented. Foremen do not implement MA1.DC10 for all Corrective Maintenance (CM) work. QV believes the threshold for entering MA1.DC10 TS is not well-defined in the procedure and therefore left to individual interpretation. When they enter TS per MA1.DC10, the effort is generally acceptable, but MA1.DC10 does not outline clear guidance when to enter the procedure.
It appears that DCPP has not resolved problems with troubleshooting. A QV audit identified these problems, and DCPP is in the process of implementing corrective actions. DCISC will continue to review this area again.
4.2.3 Conclusions and Recommendations
- Conclusion:
- Overall it appears that the DCPP Maintenance Program is functioning satisfactorily as evidenced by high plant capacity factors, successful risk-based on-line maintenance, successful ten-year in-service inspections, effective Measuring and Test Equipment Program; however, it needs improvements in certain areas, such as troubleshooting, procedure adherence and rigging. DCPP has returned to a functional maintenance organization which DCPP believes will better suit the plant than the previous process-based organization design. The DCISC will continue to concentrate efforts in the Maintenance area.
- Recommendations:
- None