IR 05000352/1988024
| ML20248J203 | |
| Person / Time | |
|---|---|
| Site: | Limerick  |
| Issue date: | 03/31/1989 |
| From: | Eugene Kelly NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20248J185 | List: |
| References | |
| 50-352-88-24, NUDOCS 8904140352 | |
| Download: ML20248J203 (18) | |
Text
q7p 4
'
Q{gy
'.
Q)
L
.
.
J.
'/
U.S. NUCLEAR REGULATORY COMMISSION REGION.I.
,
N
- Report:No.
88-24.
Docket No.
50-352
,
. License No.
-NPF-39
,
v Licensee:
' Philadelphia. Electric Company Correspondence Control Desk P. 0. Box 7520
Philadelphia, PA. 19101~
Facility Name: Limerick' Generating Station, Unit 1
'
Inspection Period:
November.7 - December 18, 1988
- Inspectors:
E. Kelly, Chief, Technical Support Section E,' Conner, Reactor Licensing / Risk Engineer
'L. Schol Reside Ins e Approved by:
.
[
E.: Kelly,pief,TechnicalSpportSection
' Dat:e Summaryi A'PRA: risk-based. inspection of balance of-plant systems involving 78 hours9.027778e-4 days <br />0.0217 hours <br />1.289683e-4 weeks <br />2.9679e-5 months <br />
-
on site, found consistent and close attention to 'non-safety systems, particularly
- c for instrument air No apparent vulnerabilities in the balance of plant were
- found which have, to date, contributed to reactor challenges such as recurrent
.
Several strong. initiatives, although still in an early stage, were scrams.
noted,.that include ISEG' involvement in balance of plant issues, a newly formed
- group dedicated to. performance monitoring, and PRA/ work coordination efforts.
Actual unavailabilities as much'as one order of magnitude higher than assumed by. the PRA for. safety systems such as diesel generator, HPCI.and RCIC, and low pressure. injection systems experienced during the first two cycles of Unit 1 operation were found as a result of PECo efforts to use the Limerick PRA for,
. among other applications, preventive maintenance decisions.
The updated PRA
- and the draft-Risk and Reliability Program Plan were found to be comprehensive and useful safety initiatives.
.
8964140352 890331
~
ADOCK0500g2
'
- r-DR 4' _ - _ - _ _.
.
l-l
.
..
L l
DETAILS
,
'
.
1.0 Introduction I
l 1.1 Risk-Based Inspection Guide (RIG)
The NRC has developed Risk-Baseo Inspection Guides (RIGS) for a number of plants in order to make insights from PRA studies more readily available to inspectors.
RIGS combine risk information from a plant's PRA with other plant-specific information (system design,-Technical Specifications, and operating procedures) in a form useful for planning and conducting inspections. A RIG for Limerick was-issued in March 1987, and has been subsequently used in several prior NRC inspections
,
at Unit 1.
The Limerick RIG contains a brief introduction.providing l
perspective on the dominant accident sequences and design vulnerabil-j ities for Unit 1.
It includes a list of plant systems that are most j
significant to risk, rank ordered by their contributions to core damage frequency.
1.2. Balance of Plant Inspection
'
'
This balance of plant (B0P) inspection was performed using the guidancr_
provided in NRC Inspection Procedure 71500, to review preventive t%
corrective maintenance, modifications, operating procedures, a".d manage-ment attention to BOP systems important to safety. The systrims selected for this review and coupled to the above described RIG methodology were:
1) Condensate and Feedwater; 2) Turbine and Bypass Control; 3) Normal Service Water; and, 4) Instrument Air.
2.0 PRA Initiatives The Probabilistic Risk Assessment (PRA) for Limerick I was first performed in 1980, during the licensing process, in response to a request by the NRC. A commitment was made to update the PRA and use it in support of engineering and operations. The licensee first updated the PRA in 1986, and has maintained it as a "living" document since that time.
This is accomplished by changing the PRA to reflect modifications installed in the plant. The most recently updated PRA contains fault tree analyses for some systems that were given only cursory treatment in the original PRA, sr.,h as instrument and service air.
2.1 Plant Use of PRA The inspector was provided a listing of 46 applications of PRA at the Peach Bottom and Limerick facilities.
For Limerick, significant recommendations or conclusions have included:
--
Evaluation of tornado missiles on the ultimate heat sink;
--
Out-of-service times for the third standby liquid control pump; i
_. _ _ - _ _. _ _ - - - - - _ - _ _ _ _ _ _ _ -. _ - _ _ _ - _ _ _ - -
_-.
_
. _ _ _
_-_
_ _ _ _ _ _ _ _ _ _ _ - _ - _ _ - _ _ _ _ _ - _ - _ _ _ _ _ _ _
L
,
..
,
,
J
_
)
,
\\
j.
.
,
,
l
--
Application of NRC Information Notice 86-11, service water protection;
--
Low probability of a safe shutdown earthquake occurring during re-placement of a. freeze plug in the emergency service water system; Low failure probability of core spray or low pressure coolant
--
injection systems, necessary while standby gas treatment (SGTS)
was not tied-into the refuel floor unavailable during first refueling outage; Bypassing certain diesel protective trips following a loss of
--
off-site power was evaluated to reduce engine reliability;
--
Modification or replacement of main steam system angle valves to reduce cavitation failures; Concluded no cost beneficial justification for the steam-condensing
--
mode of RHR exists; and,
--
Recommended that the feedwater control system be adjusted to ensure system availability without operator action following a turbine trip.
Also, licensed operator reliability data, collected at the Limerick
simulator for evaluating human failure probabilities, are still being l
analyzed..The inspector concluded that the licensee has made con-l siderable use of,PRA methodology in supporting engineering and opera-tion activities at Limerick.
2.2 Updated PRA The most recently updated PRA combines both plant specific failure rates and generic failure rates for pumps and valves, and the latter
,
data correlate well for Limerick.
In order to improve PRA usage, the licensee has formed a PRA Task Force consisting of members f-om each organization of the Nuclear Group. The Tark Force developed a draft Risk and Reliability Program Plan and will <
inue to support the program after the plan is finalized. The i n's objectives include:
--
Suppor risk evaluations requested by plant management;
--
Evaluate a reliability centered preventive maintenance program;
-
Manage outages for systems important to safety;
--
Maintain a computer risk model in a living format;
--
Support the NRC's request for an Individual Plant Evaluation (IPE) in accordance with NRC Generic Letter 88-20;
__
_ _ _ _ _ _ -
_
__
-.
i
.
-
y
.
.
. e)
e r
a
'
4-
]
' Respond to risk-related questions from the PECo Nuclear Group;
--
Develop a' method for evaluation-of proposed _ Technic'al Specifications;
--
Investigate PRA'use in development.of realistic emergency scenarios;
-,
Use.PRA insights to program quality assurance audits;.
--
Promulgate a summary list of PRA applications; and,
--
--
Maintain a PRA familiarization training program for the Nuclear Group.
The inspector' concluded that the licensee's draft Risk and Reliability-Program plan is comprehensive, and that utilization of PRA in this proposed manner should improve overall Limerick operating safety.
Th'e inspector reviewe'd the contribution to Core Damage Frequency as a function of initiating events from the updated PRA.
B0P systems, primarily turbine trips and condensate /feedwater upsets, remain important considerations for the transient, ATWS, and special.initia-tors. Although these core damage probabilities are relatively low (average about 2 E-06), they are in;the same range as the loss of
~ off-site power. initiators for Limerick. Thus 80P inspections, such
.
as this..one, potentially provide important insights to reactor' safety.
2.3 Safety System Unavailability The' inspector discussed the progress of the licensee's application of their.PRA to Unit 1 operations, specifically regarding preventive main-tenance decisions, with engineering personnel.
The inspector found that preliminary estimates for total safety system unavailability during Unit 1 Cycle 1 operation which began in February 1986 and continued until April 1987 were, in most cases, greater than assumed in the original PRA.
The estimates of actual unavailability of safety trains and systems during this time frame ranged as much as an order of magnitude higher than originally assumed for the Limerick PRA.
Those estimates, listed below, were derived conservatively from oper-ating logs, surveillance records, and maintenance history.
The esti-mates included preventive maintenance activities which, because pre-vious Peach Bottom experience was used in determining the PRA unavail-abilities for t.imerick, were a primary cause of the higher than expected Cycle 1 unavailability history.
The Peach Bottom Technical Specifica-tions impose significant test surveillance requirements on ECCS equipment when a single train or system is taken out of service, which serves
..
_.hm_m__m.2__.-_-4_________a_.-___.__m______._m. - _. _. _ _
__m__-_._____m_
__.__ _. _ _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _. _ _ -. _ _. _ _., _ _ _. - _ _.. _ _ _..__.
. _ _ _ _ _ -
_ _ _ _ _.
.-
_____2 a
___a
,
-
_ -.
..
_
m,
_
<
-
,/
'
"
,
,.
- s
.
5 l
L
.
"
~as a potential'disincentive to preventive maintenance.
Contrasted
,
,
!
'
=with these requirements are the. Standard Technical Specifications
.such as at' Limerick which do not impose: additional extensive ECCS:
testing when.one. train or system is removed (or forced) from service.
Also', there are generally more surveillance tests performed at a newer near-term' operating license (NTOL)' plant such as Limerick.
The' estimates of actual Limerick Cycle 1 unavailability were conserva-tive. This was because the total time assumed out of service was,
.for testing, purposes, calculated from log entries when test personnel signed-on and off of a test. These times bound the' actual time'when equipment'is truly unavailable due to test (e.g., by as much as a factor'of 3-5. times).. The same conservatism was applied to estimates of actual unavailability'due to maintenance,' again bounded by the logged. removal' from service period,'which is generally longer than true unavailable time while disassembled.
At the conclusion of this inspection, the licensee was in the process of calculating actual unavailabilities for Cycle 2 operation during the. period beginning in September 1987 through December 1988.
Pre-liminary estimates available after this inspection indicate smaller actual-unavailabilities although, in some cases, still in excess of original PRA assumptions.
Improved trends were experienced.towards the last half'of Cycle 2 operation, specifically after the higher
-
unavailabilities were identified by the licensee. The overall effect of increasing core damage frequency was not calculated by the licen-see for Cycle 1 but was estimated to be within the PRA analysis cal-
- culational uncertainty band (i.e., plus/minus factor of 5).
Based upon the above findings, the licensee's Risk Branch sf their corporate Nuclear Engineering Division began the development of a methodology to:
a) begin tracking system total unavailabilities; b) evaluate on an ongoing basis the effects on total core damage frequency (CDF) of increased unavailability; and c) establish and support site PRA expertise and coordination. These efforts are described in the following Section 2.3.
.
.
_
_
_ _ _ _ _ _ - _ _ - _ _ _ - _ - _ _ _ _ _
..-
,
-
.
.
6-l TOTAL SYSTEM UNAVAILABILITY UNIT 1 CYCLE 1 (1/86 THRU 4/87)
PRA Cycle-1 Ratio of System Assumed Actual Actual / Assumed HPCI 1.0(-2)
6.4(-2)'
6.4 RCIC 1.1(-2)
4.5(-2)
4.1 Batteries:
--
Division 1 1.4(-3)
Division 2 1.4(-3)
'
--
Division'3 1.4(-3)
4.7(-4)
0.3
--
Division 4 1.4(-3)
5.6(-4)
0.4
!
--
- . :
. Diesel Generators:
'
--
D11 1.0(-3)
1.6(-2)
16.0 D12 1.0(-3)
8.4(-3)
6.8
--
--
D13 1.0(-3)
1.3(-2)
13.0 D14 1.0(-3)
1.4(-2)
14.0
--
Service Water:
ESW Loop A 2.0(-3)
4.2(-3)
2.1
--
' ESW Loop B 2.0(-3)
1.2(-3)
0.6 l
--
RHRSW Loop A 1.0(-3)
1.0(-2)
10.0
--
RHRSW Loop B 1.0(-3)
1.25(-2)
12.5
--
LPCI:
Loop A 2.0(-3)
1.5(-2)
7.5
--
--
Loop B 2.0(-3)
1.2(-2)
6.0 Loop C 2.0(-3)
2.1(-2)
10.5
--
--
Loop D 2.0(-3)
1.8(-2)
9.0 Core Spray:
--
Loop A 2.0(-3)
3.4(-2)
17.0
--
Loop B 2.0(-3)
1.1(-2)
5.5 Standby Liquid Control:
--
Pump A 1.1(-2)
1.6(-2)
1.5 Pump B 1.1(-2)
2.6(-2)
2.4
--
Pump C 1.1(-2)
1.5(-2)
1.4
--
!
l l'
_ _ _ _ - _ _ _ _ - _ _ _ _ _ _
_ _ _
_ _ _ _.
_
_
_
_ _ _ _ - _
- _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - _ - _
.
.
.
.
.
t.
>
2.4 PRA Applications to Maintenance The licensee's administrative controls and procedures for performing maintenance (A-26) require a site PRA coordinator to prescribe guidance concerning working prioritization, planning, and scheduling based upon
' risk concepts derived from the Limerick PRA.
Qualitative and general
!
guidelines were developed by the licensee's Nuclear Engineering Depart-
'
ment and were being incorporated into two Administrative Guidelines (AG-
!
42 and 43) at the time of this inspection.
The Guidelines were issued on January 30, 1989, and provided to the inspector subsequent to this
,
inspection. A brief over view and progress with respect to these I
initiatives is provided below. Since Unit 1 ended Cycle 2 operation and shut down in January 1989, the use of PRA in planning system outages and prioritizing maintenance is designed to become fully implemented starting with Cycle 3 operation expected to begin in the
-
~
Spring of 1989.
l The Guidelines identify four categories of systems with respect to risk j
importance, based upon PRA calculations regarding the impact of system unavailability.
Further amplification of specific equipment importance within a given system is not explicitly contained in the Guidelines, but is intended to be provided by the site PRA Coordinator via the corporate PRA Branch of the licensee's Nuclear Engineering Department (NED) ~. Actual work is coordinated via a site Work Coordination Branch, and monthly system unavailability statistics are provided by a newly formed site Performance Monitoring Group. Also, beginning in January 1989, the calculated and reported total unavailabilities for safety systems (INP0 performance indicators) will include planned preventive maintenance as well as unavailability due to testing.
A running 12-month calculation of the differential (or delta) core damage frequency (CDF) contribution due to system unavailability is made using a correlation or set of curves developed by the PRA Branch of NED. The results of PRA/ maintenance decisions for either large outages or smaller individual system outages are used as a planning tool to regularly review where, at a given point during the operating cycle, the plant is with respect to CDF, and how much margin exists to the currently calculated CDF of 6.7E (-06). The details of these Guidelines, and actual imple-mentation during Cycle 3 operation, will be evaluated in future inspec-tion.
This program is an excellent and well-concerned initiative to manage risk, promote wider understanding of the ramifications of the Limerick PRA, and insure safe reactor operation.
2.0 Maintenance / Surveillance To evaluate the maintenance and surveillance programs for the BOP systems selected for inspection, the inspector interviewed responsible plant staff members, reviewed maintenance and surveillance procedures, and observed
portions of a surveillance test in progress.
- _---
,
.
. _ _ _
_.
.-.
..
.
.
'
3.1 - Equipment Failure Analyses The licensee has a problem identification tracking (PIT)' system to keep
. track of'such items in accordance with Administrative Guideline AG-21.
The. licensee'.makes.no distinction between the' method of handling PITS
'for BOP' systems versus.that for safety systems, indicating a_ccasistent treatment.of plant safety problems.which may derive from either safety or B0P systems.
The inspector _ reviewed the following PITS related to the 80P:
~
PIT 88-24, Feedwater Speed Change an'd Oscillations: This March 23, 1988 event was simi.lar to; oscillations that occurred during the performance of reactor feed pump; turbine (RFPT) lube oil operability test. These oscillations of the'"B"'RFPT could be eliminated by reducing.the motor c
gear unit (MGU) output manually until dampening occurred and then re-
'
turning to the original level without perturbation. -The corrective ac-tion was to complete a:previously' identified maintenan_ce request (MRF 8708154) to replace two bad bearings, a worn drive gear, and a bent shaft on the "B" RFPT speed controller.
PIT 88-56, Condensate Pump Trip Due to High Bearing Temperature Resulting from Lower Sight Glass Breakage: This October:7, 1988 event resulted in the. stopping of "C" condensate pump and limited reactor level upset
_
without a reactor trip. Although vibration was first considered the
'
cause of the failure, the most-probable cause was overtighting the sight l_
glass,at some earlier time.
In addition, the licensee is planning to
plumb.the condensate pump to reduce the vibration.
This event was thoroughly investigated.
The inspector also reviewed the September 19, 1987 electro-hydraulic control (EHC)-weld failure evaltation (Modification 5731). This EHC failure resulted in a turbine trip / reactor trip. The evaluation included a metallurgical laboratory report conr;luding that the' failure resulted from lack of weld root penetration, an engineering memorandum stating
'
that vibration aggravated the weak weld, and recommending vibration monitoring to prevent future failures. The inspector concluded that a good engineering evaluation was pe" formed on this BOP failure.
3:
3.2 Preventive Maintenance (PM) Program The inspector discussed the PM program with the maintenance staff. About 400 to 500 PMs are performed during refueling outages. Maintenance performs approximately 350 Technical Specification (TS)-required F
surveillance tests, and from 80 to 100 PMs per month during plant
,
. operations.
These are performed under individual maintenance request forms (MRFs) for job control. The inspector reviewed five recent MRFs; no' problems were identified.
,
u
__ ____
_ _ _.. _ _ _ _ - _, _.. _ _. _ _ _ _ _ _ _ _.. _ _ _ _ _ _ _ _. _ _ _ - _ _ _ _. _ _ _ _.
_ _ _,
..____m____
_. _ _ _ _ _ _ _ _
_ _ _. _ _. _ _ _ _. _ _ _
_ _ _ _ _. _
.__
._
_ _ _
- - - - - - - -
-
,
.
y 7.y
w
,x a
.
'. "
.
.
f
,
I 9-
,
The licensee stated that they use technician training, procedures and j
-
- supervision to_ prevent multiple surveillance or PM mistakes onllike
,
. eq ui pment'.
In addi+1on, there are 10 maintenance system engineers as-t
. signed.to work with.the technicians on the jobs.
The NRC RIG (described in Section 1.1.of this' report) identified eight j
Jremotely. operated valves'in the condensate /feedwater system and
.
'
isolation and check valves..in the service water system whose failure
-
toLremain in (or:to operate to) the designed accident position could be a condition that can lead to core damage.
In addition, the-inspector questioned differences in implementation of the PM program for' safety systems versus B0P systems..The.. licensee provided computer'
y printouts.of maintenance. request forms- (MRFs) and the PM baseline. for
!
- the' selected BOP valves.
These valvo are listed as critical components, and are treated as' safety components with the exception _of PM frequency.
!The frequency.is based on failure history instead of TechnicalLSpecifi-cation: requirements.
The inspector reviewed the same data for the
- main turbine' bypass valves.
The results were consistent.
No indicationLof inadequate preventive maintenance was identified.
-The RIG identified rupture of the main steam high pressure header'to the w
> :feedwater pump turbines, or ruptures at any of'seven sections of the-condensate /feedwater piping, as important initiating events.
Since
.
f erosion / corrosion of steam and condensate /feedwater piping has been a problem.at BWRs, the' inspector reviewed the licensee's Erosion / Corrosion Detection Program :The ultrasonic detection of pipe thinning is
'
controlled by ISI-SP-3. Rev. O.
This procedure is used to detect and document wall loss resulting from single-phase erosion / corrosion of elbows, reducers, and tees. A two phase flow program is under development. During the first refueling outage as well as at an April 1988 mini-outage, fifteen pipe fittings were inspected under the single-phase erosion / corrosion' program.
The minimum wall thickness readings were sufficiently above the allowable minimum wall thickness.
The licensee. plans to measure 18 single phase and 26 two phase locations during the upcoming outage begun shortly after this inspection.
The inspector questioned whether PRA methodology was used in the selec-tion of piping to be measured.
The licensee's program is based on system flow characteristics and piping configuration; however, there has been no evidence of erosion / corrosion in systems where fluid temperature is below 200 degrees F.
This relegates those feedwater lines from the discharge of the number one feedwater heater to the vessel the most susceptible single phase, erosion / corrosion area in
'.
-the plant.
The piping selected for the two phase program includes moisture separator lines, heater drains, extraction steam piping, condensate drains, and reactor water cleanup piping. Considered in the selection process are pipe materials (e.g., drains are normally lower grade piping than major lines), geometry, experience from other power plants, and ALARA considerations.
__:__a_---_-__--
-
__ --
_
__
_
_. _ _ _
- -
,
.
<3.
. -
.
.
- .
10-
~
The licensee is presently ~using the " Band" ultrasonic method of measure-
~
ment. Much of the' industry utilizes the " Grid" method of testing because measurement data are.more easily compared in future examinations.,The licensee is considering such a change in-their program.
FL or;the present,.
~they have a measurement target marking system so' measurements can be compared.with past measurements. The inspector concluded thatLthe,lic-ensee's ultrasonic measurement' program is structured such that erosion /
corrosion does not occur undetected.
3.3 ;0perating Experience Assessment
'PECO's Operating Experience Assessment Program (0EAP) is responsible to evaluate information from the nuclear industry, including Peach Bottom.
andl Limerick, and determine those actions;needed to address potentially similar events. A Nuclear Group Standard was issued in June 1988 to.
control these activities. A Nuclear Group Administrative Procedure is responsible for identification of new or revised commitments, Implementation' of commitments, tracking commitment status, and
. maintaining; commitment data' base records.
-Administrative Guideline AG-40,, Control of LGS Operating Experience-Assessment Program Items and Commitments, providas guidance to site-personnel: including disposition and closure. A Regulatory Support Enginee'r' tracks the. status of all OEAP items, but the assigned respons.ible organization is responsible for task completion.
In order to evaluate the tracking system effectiveness, the inspector. reviewed maintenance and instrument and control group systems.
The inspector obtained a three page Maintenance Section Conformance Tracking System printout, dated December 5, 1988. The listing included
. licensee-identified issues, vendor safety information letters and Nuclear Network or INP0 issues, as well as NRC Bulletins, Information.
Notices, Generic Letters, and other correspondence.
I&C Correspondence Control Guidelines were developed from which the inspector chose a number of NRC items and requested their status.
Except for Information Notice 86-110, Anomalous Behavior of Recir-culation Loop Flow in Jet Pump BWR Plants, all requested information was available in I&C files.
Data related to Notice 86-110 was provided to the inspector subsequent to the inspection, and resolved the recirculation flow issue for Limerick. This was based on lack of observation of bistable flow oscillations at Unit 1, and GE's evaluation concluding that plant operational safety and equipment /
component service duty are not compromised by bistable flow.
The inspector concluded that tracking of issues assigned to I&C by the OEAP is adequate.
l f
'
_
_
.-
_-
- _
-.
_ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - - _ _
_
Yn.,.
.
s
,
z.
q:
..
,
'll 13.4 Maintenance Procedures
'
The inspector discussed the handling of B0P equipment with the mainten-ance staff..The exact same. types of calibrations are performed on BOP and
.
safety equipment; the.. sole difference is frequency of' testing.
For safety systems,' calibration frequency is set by TS while for 80P systems the frequency is based on. operating and failure experience.
As an ex' ample of handling.a plant problem, the inspector reviewed the recent vibration problem with the "C" condensate pump. The technical.
staff maintenance engineer issued an engineering work request (EWR) to machine the pump pedestal' gasket surface'to align the' pump / motor assembly properly.
A' test engineer issued an EWR_to install a dynamic absorber as a-temporary fix while the pump / motor realignment was being investi-gated.' The-lower motor bearing'was'to be redesigned for long term re-liability improvement, indicating positive action on a 80P system that should improve plant reliability.
The inspector observed a portion of surveillance testing of the B Loop ESW (Pump,. Valve and Flow Test ST-6-011-232-0)..The purpose of this test is to prove ESW pump operability, exercise and stroke time ESW loop valves, and. insure that ESW check valves operate correctly,
'
. including those that isolate the normal service water.
The procedure was well written and was_being followed by test personnel staff.
The test was started near the end of day shift and was suspended for the next operating shift to complete.
No safety concerns were identified.
3.5 Replacement Parts
.The inspectors. reviewed.the controls which are in place for the control of spare parts.. Administrative procedures A-27.11, Coding of Spare Parts for Use at Limerick Generating Station, and A-27.12, Changing Coded Spare Parts Procurement Information or Stocking Levels, include instructions on how to establish the proper procurement method for spare parts and how to make changes to spare parts. procurement information. This program ensures that any replacement part including those for balance of plant systems, are either identical parts to the original or any substituted part receives a technical review for its acceptability.
These reviews are documented in accordance with the above procedures.
3.6 Measurement and Test Equipment The inspector talked to maintenance craft supervision and' reviewed main-
'
tenance department administrative procedure MA-6, Rev. 7 (Calibration and Control of Maintenance Department Measuring and Test Equipment).
Maintenance' craft personnel are responsible for assuring that test equipment is within its calibration period and properly returned to
.
. _ _ - - = _ _
_ - - - _ _ -
..- _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ - - _ _ - _ _ _ -
_ _ _ - _ _ _ _ - - _ -
_ _ _ - - _ - _ _ _ _
_
_ _ _ _ _ _ _ _ _ _ - _ - _ _ - _ _ _ _ _ _ _ - _ - _ _ - _ _ _ _ - _ - _ _ - - _ - _.
- _ - _
fl
<
w
.
.s.
w, ;.
,
l
'
r
- -
,
...
storage'after use. The' tool attendants maintain a measuring and test equipment; usage logiand ensure the general condition of all test equip-ment.
The. licensee has a calibration group that is responsible for the repair and/or calibration' of all test equipment.. There is no distinction
-
~between the test equipment, either hot or cold radioactively, used for B0P or safety-l system components. -No safety issues-were identified.
'3.7.' Maintenance Traininj The' inspector: discussed maintenance training.with craft supervision.
A new hire receives a two-week orientation and then~is.a maintenance helper, rotating through the different crafts (electrical, mechanical, and' instrument and_ control), for 24 to 33 months. When a-third class craft opening occurs, the helper can bidion the position and take written and practica11 tests.
Passing the. third class test leads.to six-to-eight weeks craft training at the licensee's Barbadoes training facility.
.
~
I After a minimum of six months as a third class craftsman,'any second class craft position may be bid upon and another test taken.
The minimum time to. bid on an open first class craft position is eight months. Again, a comprehensive written and practical test is used..The written test administered require about eight hours to complete and the practical test require.up to 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />. After the first class craft position is reached,
,
L it takes about five years to complete the on-the-job training-required g
by.the licensee.
The licensee is moving towards the use of a." tradesman" who will be trained to perform about 80 percent of the mechanical,. electrical, and instrument and control work on an integrated basis. All tradesmen who receive training after January 1,1986 are sut iected to this integrated training in accordance with MA-9, Procedure for Training and Testing of Maintenance Department Tradesmen.
The NRC inspection program will con-tinue to' review the implementation of the tradesmen concept.
The. training requirements to work on a B0P system are no less restrictive than for work on a safety system.
Routine tests are performed on non-TS systems (i.e., feedwater turbines) in much the same manner and frequency
>
<
'
as on safety systems. No plant safety issues were identified.
4.0 Modifications The inspector requested a printout of B0P modifications in process at Limerick.
H-
~A list of modifications for the condensate - feedwater systems was provided.
The e modifications are planned to improve system reliability and minimize the etentit i for plant transients.
The inspector reviewed the following plac modifications (mods), scheduled for the next refueling outage, in detail.
--
__
__
. - - _
_.
.
-. -
,
,
-
.
Mod 5007: Replaces the existing Kieley-Muller feedwater heater level control-1ers with Fisher controllers.
The reasons for the change are that the exist-ing controllers sometimes exhibited erratic operation and replacement parts were difficult to procure. By improving the controller reliability, the potential for the-loss of feedwater heating and the associated reactivity addition transient is reduced.
Mod 5515: Changes the type of coupling between the reactor feed pump and the feed pump turbines.
The purpose of the change is to eliminate the potential for feed pump vibration problems or coupling failures which can be caused by inconsistent lubrication.
However, periodic lubrication is not practical, since all three feed pumps are required to suoport full power nperation.
Lubrication also results in additional radiation exposure to plant personnel.
i Thus, modification 5515 eliminates a condition (the inability to periodically i
lubricate the couplings) which is highlighted as a critical coupling vendor recommendation. Improving feedwater system reliability thereby reduces the possibility of a loss of feedwater transient which could challenge plant safety systems.
ihe inspectors reviewed the temporary circuit alteration (TCA) program I
which is implemented in accordance with administrative procedure A-42.
This procedure controls the installation and removal of temporary system jumpers on both safety systems and balance of plant systems.
The following observations were made:
--
Safety evaluations are not routinely performed or procedurally required on TCAs implemented on balance of plant systems. However, a significant revision to procedure A-42 is being written which will require a written safety evaluation for all TCAs, and also approval by the Plant Operations Review Committee (PORC) prior to any TCA installation.
Some TCAs have been installed for up to four years while awaiting
--
permanent plant modifications.
This concern was previously identified in Inspection Report 50-352/87-19 and is being addressed by the licensee via higher engineering priority to those TCA's requiring permanent plant modifications.
--
TCAs are utilized in the same manner on 80P systens as they are in safety-related systems, resulting in an accurate accounting of all system jumpers and temporary alterations.
The licensee has been involved for over a year in a program to reduce the number of TCAs, presently numbering about 120.
To accomplish this goal, the licensee has assigned co-ordination of TCA reduction to a project engineer, and expects the number of TCA's remaining after the present outage to be approximately 20.
!
!
_ _ _ _ - _ - _
. - -.
.
,
.
,
,
.....
-
.
p 5.0 Operations The inspectors reviewed various plant procedures and other documents related to feedwater system operation.
Included in this review were:
Piping and Instrumentation Drawing M-06.- Feedwater System
--
Procedure S06.1.A - Startup of a Single Reactor Feed Pump
--
Procedure S06.1.C - Placing a Standby Reactor Feed Pump in Service
--
Procedure S06.2.0 - Removing the Reactor Feed Pumps from Service to a
--
Standby Conditior
--. Procedure S06.7A - Reactor Feed Pump Seal Injection System Operating Modes
--
Procedure S06.8. A - Reactor Feed Pump Warm-up
--
Shift Night Orders Control Room Operator Logs
--
.
--'
Equipment Operator Round Data Sheets Plant Incident Tracking Reports (PITS)
--
--
Upset Reports 5.1 Operating Procedures Plant operating procedures were found to be consistent with the as-built drawings and appeared to be sufficiently detailed as to ensure proper operations of the systems. The amount of attention given to the development and maintenance of the B0P system procedures appeared to be commensurate with that of safety related system procedures.
The plant operators are knowledgeable in the content and use of the procedures.
Prior to changing the operating alignment of B0P systems, operators evaluate the potential effects on overall plant operation and the possi-bility of causing an unexpected operational transient.
5.2 Conduct of Operations In-the past the operators have demonstrated a thorough knowledge of feedwater system operations through their ability to rapidly respond to
. feed system perturbations, thus preventing plant trip _ _ _ _
. - _ _ _ _ _ _
.__
_ _ _ _ _ _ _ _ _ _ _
_ _ _ - - _ _ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ _
___
fs E
,, '..
..
s'
.,
i'
. Pre-shift briefings are thorough and include a discussion of any proce-dural or system changes which have the potential to affect operations, i
This information is also provided in written form via the shift night
!
orders.
Thus, the operators are constantly kept abreast of information.
which may enable them to avoid unnecessary plant transients.
Day-to-day activities such as troubleshooting and testing of balance of plant systems are well contr'olled and performed in accordance with the established administrative controls.
!
5.3 Corrective Action Programs System problems which cannot be immediately corrected are documented and tracked in the plant incident tracking system (PITL). This system pro-vides a means for organizing and documenting sequences of events, troubleshooting results and any necessary temporary or permanent correc-tive actions.
Before the incident can be resolved, it must be reviewed by the PORC to obtain concurrence that actions taken were appropriate.
.The PIT system is used extensively and is applied to B0P system prob-lems as consistently as for safety related system problems.
The plant staff also utilizes an Upset Report to document detailed reviews of plant operational transients. These reports are useful in ensuring a thorough analysis as well as appropriate corrective actions.
Upset reports are not limited to those types of events which are required to be reported to the NRC via a Licensee Event Report.
5.4 Feedwater System Walkdowns Portions of the feedwater, feedwater heating and feed pump lube oil sys-tems were inspected. The systems were generally found to be in good condition. One problem which was_noted is that two of the three sixth stage feedwater heaters have cracked or broken level indicator sight glasses which are isolated to prevent steam leakage.
The glasses cannot be repaired during operation since shutting an additional isolation valve,
'for personnel safety concerns during maintenance, would also isolate the automatic level control instrumentation.
This condition prevents the operator from being able to verify heater level to monitor proper auto-matic level control. The controlling of feedwater heater level in the proper band is important so that inadvertent positive reactivity addition transients caused by the loss of feedwater heating can be avoided.
The inspector noted that in addition to the feedwater heater level sight i
glasses differential pressure gauges are also installed.
These gauges i
have an indication of -5 to +5 inches with zero inches being the normal desired feedwater heater level.
These gauges are used primarily for level monitoring during level controller tuning however through some additional operator training these gauges could become more useful indi-cation to back up a broken sight glass.
_ _ _ ___.___
-
.
.
.
- - _ _ - _ - _ - _ - _ _ _ - _ - _
-
//
-...
.a r
.n :
k:
e
'
Another condition which was noted during the walkdown was that the feed pump seal injection differential. temperatures would gradually cycle above and below their' normal.setpoint of 20 F.. Differential temperatures were observed:between 10 and 35.F.
The significance of these temperature swings was,not immediately apparent; however, the loss of a feed pump
>
seal and subsequently the associated feed pump could lead to an undesired L
plant transient.
Licensee personnel subsequently tuned the pneumatic controllers for the seal injection control; valves, which in turn
.significantly reduced the temperature variations.
At the time.of this inspection, the 'C' reactor feed pump was inoperable after being shut down when it was found to.be making unusual noises.
The cause of the noise has not yet been determined and the investigation has been suspended'until the upcoming refueling outage shutdown due to the inability' to obtain satisfactory isolation for further maintenance.
Since plant power is reduced due to fuel cladding' degradation, the third pump is'not required and plant operation could be continued with only one feed' pump-in operation. However, licensec 'anagement made the.de-cision that the plant would be shut down ir event that a second pump fails.
This operating strategy displays a conservative safety perspec-n tive by not-risking a loss of feedwater transient which could potentially
'
o
'
be the result of a common mode pump failure.
5.5 Operations Summary:
In summary, the control and operations of balance of plant systems re-ceives the necessary attention by operators and site management to minimize plant transients and resultant challenges to safety systems.
- 6.0 Instrument Air Systems The inspectors reviewed the status of maintenance, test, modifications, and operating history for the instrument air system. Air system issues were dis-cussed with the responsible system / test engineer, and system procedures were reviewed.
6.1 Operating History
,
-The Unit 1 instrument air (IA) system has recently had an extensive multidisciplinary review in response to NRC Information Notice 85-35..
l-INP0 SOER 88-1, and NUREG-1275 (Volume 2). A number of recommendations were made that were in various stages of implementation as of the time of this inspection._ These included revisions to operating procedures, alarm response cards, operator training programs, preventive maintenance programs, and potential design changes. The inspector reviewed licensee evaluations issued in November 1988 and discussed air system modifica-tions with cognizant personnel.
.
I,;
I
._ - - - _-___ _ ---_---_ ___
. _
- -, -. - - - - _ - - - - - - -. - - - - - - - - - _ - - - - -- - - - - _ _ - - - - - - - _ - - - - - - - - - - - - - - -
--
k (Y:I
'
..
-
.
...,
.
.
<
{
)
I Modification Number 5561, underway over the past year, involves a com-plete walkdown-of the IA system to identify all air supply block valves to each pneumatic user which are not currently identified on the system P&ID M-15.
The walkdown, eventual numbering and tagging, and procedure revisions were in response to a Unit 1 scram on September 7, 1987, and a number of reportable events in 1987 involving reactor enclosure isola-tions that were caused by incomplete or inadequate IA block valve iden-tification.
This effort is a significant step towards the reliability of instrument air to all unit I systems (and eventually for Unit 2 as well), particularly those important to risk, including BOP systems.
6.2 Maintenance and Testing The licensee's preventive maintenance program, as applied to IA systems, was found to be well developed and thorough, primarily as a result of
!
the previously mentioned industry information and generic concerns for air systems (i.e., NRC Generic Letter 88-14).
Several examples cited are: the compressor air capacity checks (RT-6-015-310), which have for the past three years consistently shown adequate system performance; changeout of critical air filters and dryer maintenance, and an FSAR commitment to monitor moisture and contaminants in the IA system at six randomly sampled points in the system every refueling outage via test procedure ST-2-015-800-1.
Knowledgeable system engineers in test, maintenance, and engineering organizations are aware routinely of IA system performance and regularly communicate with each other to assure the reliability of the system. This includes potential modifi-cations to the system such as Engineering Assistance Request No.88-004 initiated in February 1988 to evaluate increased IA capacity in response to single failure concerns; and, positive seating check valves at strategic locations in the system to enhance the redundancy of IA Loops A and B.
Recommendations from the initial study were being implemented at the conclusion of this inspection.
6.3 Operations and Risk Evaluations The review of alarm response cards and off-normal procedure ON-119, Loss of Instrument Air, indicate that air system contingencies and expected design response have been effectively integrated into plant operating procedures.
Finally, a preliminary risk evaluation recently completed for the IA system concluded that random or common cause failure of the system is a low frequency event; and, that support, frontline BOP and safety systems are not noticeably affected by the loss of instrument air.
Therefore, the loss of IA initiator is considered to be, from a risk perspective, one associated with very low core damage frequency sequences (i.e., in the range of 1.0E -08 to -09).
- _ _ _ _ _ _ _
- - _ - -
C..
.
<
..
I-'a
7.0 Management Meetings The inspectors discussed the issues in this report with members of licensee management and staff throughout the inspection period.
The findings were summarized at an exit meeting held with the Plant Manager, Limerick Generating Stat' ion, on December 5, 1988. No written inspection material was provided to licensee representatives during the inspection period.
!
l
l
_ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _