IR 05000255/1986030
| ML20212P869 | |
| Person / Time | |
|---|---|
| Site: | Palisades  |
| Issue date: | 03/05/1987 |
| From: | Burgess B NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML20212P832 | List: |
| References | |
| 50-255-86-30, NUDOCS 8703160300 | |
| Preceding documents: |
|
| Download: ML20212P869 (17) | |
Text
.....
..
...
.-
-_.
_
_. _ _ _ _. _ _
...
. _..
.
,
.
_
.. ;. :
-
-
.
,
U. S. NUCLEAR REGULATORY COMMISSION
'
REGION III-
' Report No. 50-255/86030(DRP)
,
Docket.No. 50-255 License No. DPR-20 i
Licensee: Consumers Power Company
~
212 West Michigan Avenue Jackson, MI 49201-
Facility Name: Palisades Nuclear Generating Plant
-
g Inspection At: Palisades Site, Covert, Michigan
Inspection Conducted: July 24 through December 31, 1986 r
.
Inspectors:
E. R.-.Swanson C. D. Anderson
-
.
-Approved By:
8. L. BurgesV, Chief Reactor-Projects Section 2A Date 4^
Inspection Summary
.
Inspection on July 24 through' December 31, 1986 (Report No. 50-255/86030(DRP))
l Areas Inspected: Special inspection of the containment escape airlock-E equalizing valve interlock problem; the low pressure safety injection pump flow problem; the low pressure safety injection pump discharge valve throttling;
.o the component cooling water flow problem that affected shutdown cooling heat
'
exchanger operability; the containment air coolers; and the service water system.
"
!.-
Results:. Violations of Technical Specification requirements were identified
.
-in each of the six areas inspected. One Notice of Violation for the escape E
airlock accompanies this report. Violations of 10 CFR 50, Appendix B were also identified in each area: Paragraph 4 details design control, test control,
!
!
corrective action and Technical Specification violations for the low pressure
safety injection pump; Paragraph 5 details a Technical Specification violation j-for the low pressure safety injection flow control valve; Paragraph 6 details
'
~ violations of test control and Technical Specifications for the component
-
cooling water system; Paragraph 7 details violations of test control, n
z
- corrective actions and Technical Specifications for the containment air coolers; Paragraph 8 details violations of design control, test control and Technical
'
' Specifications for the service water system.
The violations described in
!
Paragraphs 4 through 8 are being considered for escalated enforcement.
8703160300 870309
"~
PDR ADOCK 05000255 G
<-
_.
.
.
..,
.
..
DETAILS
,
1.
Persons Contacted
- J. F. Firlit, General Manager
- J. G. Lewis, Plant Technical Director
- R. D..Orosz, Engineering and Maintenance Manager W..L..Beckman, Radiological Services Manager C. E. Axtell, Health Physics Superintendent
- R. M. Rice,-Plant Operations Manager
- R. A.-Fenech, Plant Operations Superintendent
- H. M. Esch, Plant Administrative Manager
- S. C. Cote, Plant Property Protection Supervisor
- K. E. Osborne, Technical Engineer
.
- G. W. Ford, Plant Safety Engineering
- R. P. Margol, Quality Assurance Administrator
- T. J. Palmisano, Plant Projects Superintendent J. D. Alderink, Mechanical Engineering Superintendent.
- E..A. Dziedzic, Plant Facility and Material Services Superintendent
- Denotes those present at the Management Interviews on September 8 and October 21, 1986.
2.
Introduction On May 19,- 1986, the Palisades Nuclear Plant tripped on Pressurizer High Pressure caused by a loss of the turbine Electro-Hydralic Control (EHC)
system.' As a result of multiple equipment failures which occurred during the trip and a concern with the overall material condition of the plant, the NRC issued a Confirmatory Action Letter (CAL) dated May 21, 1986.
The CAL required investigation and repair of the failed equipment,
,
investigation of the material condition of the plant and Regional Administrator approval of plant restart.
In response to the CAL, the licensee initiated a Material Condition Task Force which identified
,
numerous items requiring maintenance activities (repair, modification or l
testing), several of which are discussed in the following paragraphs of this report.
During investigation of specific Material Condition Task Force items related to safety system operation, concerns regarding the ability of safety related pumps to meet design flow and head requirements were identified.
In an attempt to resolve these concerns the licensee performed a testing program for safety related pumps. Results of pump testing in conjunction with the evaluation of Material Condition Task l
Force findings identified additional concerns regarding safety system performance. To verify that critical functions of safety-related L
components were adequately demonstrated by a test or supporting
}
documentation, a System Functional Evaluation was performed. The result l
of these combined efforts will be a verification of safety systems to l
perform as designed and improved surveillance and maintenance programs.
l l
.
,
l i
.
.
..
.
To provide additional assurance that the plant "as-built condition" is consistent with design documents and Technical Specifications, the licensee has committed to the implementation of an extensive Configuration Control Program. The increased use of vendors to supplement the lack of plant experience and improved training of plant personnel are steps recently' completed by the licensee to improve overall plant material condition.
3.
Containment Escape Airlock Background The Palisades containment thirty-inch escape airlock was manufactured and installed in accordance with W. J. Wooley Company drawings. Design features of the lock utilized a single operating shaft to open either door and simultaneously operate an equalizing valve which equalizes the air pressure across the door being opened.
Testing during preoperational phases (1968-1971) confirmed proper operation of the interlock to open the equalizing valve and its associated door and showed that only one door at a time would open.
The licensee did not test to insure that only the appropriate equalizing valve opened each time a single airlock door was operated.
During plant operation, performance of the surveillance test on the personnel air lock (S0-4A) required entrance through the escape airlock.
This test was completed at least every six months and is documented to have occurred on several occasions when the reactor had been critical. A recent example based on records reviewed occurred on February 15, 1985.
On July 24, 1986, during the use of the containment escape airlock following preventive maintenance, a licensee employee observed the outer door equalizing valve operating when the inner door was opened.
Subsequent investigation revealed that the event was due to a design defect and in addition, the as-built condition had been accurately reflected on the drawings. Based on this design, each use of the airlock breached containment integrity (for the brief time each door and the opposite one inch equalizing valve were open) and created a direct leakage path out of containment.
Containment integrity is defined in Technical Specification (TS) 1.4 which states, in part, that each airlock door must be properly closed and sealed and all non-automatic containment isolation valves be closed.
TS 3.6.1.a prohibits breaching containment integrity unless the reactor is in cold shutdown.
Each time an escape airlock door was operated with the reactor in other than cold shutdown, TS 3.6.1.a was violated.
,
.'~.y
.,
~ Corrective Action
,
On Augu't 8, 1986, the licensee completed a field modification s
to.a roller cam which caused the mis-operation of the opposite door equalizing _ valve and tested it for proper operation. A verbal 10 CFR
'Part 21 report was also made by the. licensee to the NRC Regional.
Administrator on this date and a written followup report was submitted on August 13,.1986. This report outlined the design defect and listed other_ utilities which are likely affected.
Licensee Event Report No. 86023 was submitted on August 22, 1986, regarding the above and evaluated the event's' significance. The licensee concluded that while the use of the escape airlock would have caused a
'
.brief-venting of the containment, more containment air is released during a normal exit through either the escape air lock or the personnel airlock which.is the normal access path to containment.
Regarding the improper' design verification and inadequate testing,
~
the licensee's actions as described in Paragraph 2 relating to the
'
Material Condition Task Force and System Functional Evaluations are considered adequate to address these concerns and prevent recurrence.
-At th'e February 20, 1987 enforcement. conference, the licensee presented an additional evaluation showing that with the containment at design
' pressure (55 psia) following a design basis accident, a door and opposite door equalizing valve could remain open for over 25 seconds before offsite radiation dose limits were exceeded.
- Summary As designed,- the actuating mechanism for the escape airlock caused a i
vent path from the inside of containment to the outside' environment each time an escape airlock door was opened. This resulted in violation of containment integrity during the brief time a door was open during operation of the plant above 210 degrees. TS 3.6.1.a does not permit violation of containment integrity unless the reactor is in cold shutdown. TS 1.4 defines Containment Integrity as including at least one door in each airlock being properly closed and sealed, and all
,
'
nonautomatic ~ containment isolation valves being closed.
Each opening of an escape airlock door,- with the reactor in other than cold shutdown
,
during the period April 1971 until May 1986, violated containment integrity.
I Numerous violations of this requirement occurred during surveillance performance and during access to and exit from containment.
Each breach
!
of containment which occurred due to the equalizing valve design deficiency is considered a violation of TS 3.6.1.a (Violation No. 50-255/86030-01).
!
Causes that contributed to this violation include design control failures
.
and inadequate testing.
Since the airlock was found to be built as designed, the design was evidently not properly verified. A proper
'
!
testing program which checked for desired functions as well as the absence of undesired functions would have identified.the defect early
l
- f
..--~-. -. _ _,.__._.-_. _._ __ _ - _._. _ _ _. _ _ _ _... - _, _ _ _ -
_.,_.~., _ _
.._,
.
.- -.-
.-.
.-
-. -. -
. _ ~...
.
..
"
.
-in plant life. No citation will be issued for these concerns since
-
they occurred prior to the imposition of _10 CFR 50, Appendix B, quality -
cassurance criteria. Corrective actions,taken by the licensee are considered adequate to correct the cause and prevent recurrence of the=
Technical Specification violation and the Appendix B concerns.
One violation and no deviations were identified.in this area.
4.
Low Pressure Safety Injection Pump
,
Background Due to excessive leakage and decreasing pump performance, the licensee decided to rebuild P-67B during.the Material Condition Task Force Review c
established in May'1986 (Item ESS-17).
Following rebuilding and testing, it was found that the pump only delivered 2700 gpm of the required 3000 gpm at 350 feet of head.
Since no changes were known that might-
have affected the pump's performance, it was concluded that the pump had-been in this condition since July 1984. During discussion with the vendor, Ingersol-Rand,.it was determined that the impeller installed in p
1983 had not been modified as those which had been originally installed i
and pre-operationally tested.
I Licensee' investigation of this event determined that during preoperational testing in 1970, a modification was made to the LPSI pumps by the pump
-
vendor, Ingersol-Rand. This modification increased tolerances of the wear ring and backfiled the impeller resulting in higher head and f_ low
characteristics. This was done on both installed LPSI pumps, but not on I
the spare impeller. The licensee noted that the early phase of plant
'
construction and testing was not governed by a quality assurance program,
and, as a result, the modification was not reflected in plant records.
i During a refueling outage in 1983, LPSI pump P-67B-impeller was damaged when maintenance was performed to replace the pump's wear
,
l rings. The impeller was replaced with an unmodified in-stock spare.
'
!
Testing apparently did not ensure conformance to system design requirements in that the test procedures did not include specific design requirements such as minimum design head and flow values as specified in the Final Safety Analysis Report (FSAR).
The plant returned to service after this outage on July 24, 1984.
ASME Section XI testing conducted during 1985 revealed several values of pump differential pressure in the alert range. As a result of an NRC inspection in this area, the problem of reduced pump performance was highlighted to the licensee. Because maintenance had been performed,Section XI permitted adjustment of the acceptance band and the pump i
subsequently, based on this new range, performed acceptably.
'
In February of 1986, a memorandum from the Section XI Test Engineer noted that the P-67B pump performance did not compare with its previousSection XI testing results and requested the system engineer to
.
., -
., m._._. ---
.
. - - -
.-
.-
-.-...
.
.
.
.
.
.
,
,,
-
,
- -
.
.
l
..
investigate the reduced performance and takt-any required corrective i action. No action was taken to identify this condition to management,
'
perform additional testing, or investigate the matter at that time.
A review of plant operations records for the period of July 198Fchrough May of 1986 disclosed changes in operating modes, i.e., cold shutdown through plant operations, for sever, periods, all in excess of 24 hobrs.
Though the test program identified the degraded pump performance, this was not recognized as a condition of.inoperability and plant operation was allowed with a technically inoperable pump in violation of
'
TS 3.3.1.c and 3.3.2.b.
,
,
Corrective Actions The licensee had the vendor modify a spare impeller. The modified impeller was installed and the' pump was tested satisfactorily at -
3000 gpm and 377 feet of head.
-
In September of 1986, an evaluation performad by Exxon demonstrated that the pump's reduced flow would not have caused the plant to exceed fuel peak centerline temperature (PCT). Both LPSI pumps were tested and found to meet their design requirements and plant drawings were corrected to reflect the impeller modification.
Long term corrective actions include post-maintenance testing
-
improvements, an augmentad testing program, the establishment of a configuration management ~ program and the implementation o.f spare parts reconciliation.
Summary The use of an unmodified impeller as a replacement in the P-67B LPSI
,
pump rendered it incapable of perforsing its design function of providing 3000 gpm at 350 feet during seven periods of critical reactor operation commencing July 1984 and ending May 1986. The reduced flow which rendered LFSI pum's P-678 inoperable during critical reactor operation constitutes a violation of TS 3.3.1.c and 3.3.2.b (50-255/86030-02a).
The original design records maintained for these pumps were never changed to reflect the vendor modified condition made in 1970. This lack of documentation resulted in the installation of an unsuitabia LPSI pump impeller in 1983.
Criterion III of 10 CFR 50, Appendix B states that measures will be established to assure that applicable design bases are.
correctly translated into specifications, drawings, procedures and
instructions. Failure to translate the backfiling of the original
,
impeller into appropriate plant document is considered a violation of )
'
10 CFR 50, Appendix B, Criterion III, 0.ign Control (50-255/86030iO2b).
Operability testing, following impeller replacement in 1983, did not detect a significant degradation in pump performance, in part due to the failure to incorporate design requirements and acceptance limits
.
.
M
,
y
.
j,
';
e
..
.
'
~
N
,j
<
j-into pump test procedures, in this case the minimum design flow and head p
values stated in the FSAR.
Failure to incorporate these design
'
requirements is considered a violation of Criteria XI, Test Control (50-255/86030-020).
The licensee also failed to take timely corrective action when the Section XI Test Enginee identified degraded pump performance in February
^ '6f 1986. Criterion XVI of 10 CFR Appendix B requires that conditions
cadverse to quality be promptly corrected. The adverse changes need in the Section XI testing subsequent to the 1983 P-678 pump impel hi replacement should have indicated to the licensee a potential change in pump operating characteristics.
Failure to take prompt corrective action is considered a ytolation of Criterion XVI, Corrective Action (50-86030-02d).
-
'
The safety signiffcance of the violations identified above is evaluated
- below in conjunction with the related flow control valve problem (see Paragraph 5 Summary).
.
The 'iden't"Ification of the above four violations resolves Unresolved Item'No. 50-25S/96023-05.
5.
Low Pressure Safety Injection Flow Control Valve
,
Background
'
-
The subject valve, CV-3006, located in a common flow path for both Low Pressur4 Safety Injection (LPSI) pumps, is required by Technical Specification (TS) 3.3.1.h to be both fully open and disabled by isolation of its air supply during critical reactor operation to prevent spurious closure.
From September through November of 1983, during the refueling outage, records indicate that several Maintenance Orders (MO) were issued requesting that the subject valve be inspected and repaired. The MOs identified two areas of concern regarding the valve performance, specifically, lack of repeatability of valve actuation (83-ESS-01961) and actuator binding (83-ESS-0224).
Final acceptance testing for the repairs consisted only of stroking the valve, and did not include specific criterion, such as smoothness of operation and timing.
!
l After completion of the 1983 refueling outage, critical reactor operation l
commenced in July cf 1984, and continued until August. Valve lineups,
'
which had been done prior to the startup in July 1984, were again
!
conmletea 3n August in prsparation for the next startup. Based on valve
'
lineup ch t klists of August 23 and 25, 1984, the position of the subject
,
valve was, verified three times, at least twice by visual observation of
local ind; cation by expeaienced operators.
l t
!
l l
m
-
- +,
,
--m m
.
' w m
..
- e
,
f On September.2,1984, at about 1:40 p.m. the reactor was taken critical.
The next day an auxiliary operator found the valve not to be fully open (3/8 closed). The unit had been starting up with the reactor in Hot Standby. At 5:29 p.m.,_on September 3, 1984, the valve was manually returned by the auxiliary operator to the fully open position.
Subsequent maintenance performed on the valve (September 9,1984)
determined that it did not fully stroke open due to tight packing and binding of the stem.
It was also found that the local valve indicator was malfunctioning in that it read approximately 100% full open when the actual valve position was 3/8 closed and manual opening of the valve to a full open position would have caused the indicator to go beyond the full e
open indic: tion point.
In analyzing the event, the licensee concluded that the approximate indication of 100% open was "close enough" for an k
operatcr to interpret the valve as actually being 100% open. This interpretation of local valve indication is suspected to be the reason that the position of CV-3006 was not discovered during the aforementioned valve lineups.
A second revision to the-LER (No. 84017, Revision 2) for the September 2-3, 1984 event was submitted on September 13, 1985, in which the licensee outlined an analysis which concluded "... that peak clad temperature during a loss of Coolant Accident (LOCA) would not have been
/
f affected by the reduced flow'! caused by the partially closed valve.
i
Corrective Action y
L On September 3,1984, within several minutes after CV-3006 was v
found3/81 closed,itwasmanuallyJackedopen. The inadequate local valve position indication was modified. A review of similar valves did not identify analogous conditions.
In its initial analysis, the licensee calculated that a valve position of 3/8 closed would result in a 15% reduction in flow below that required by the LOCA analysis of the Final Safety Analysis Report (FSAR)... Specifically, maximum flow rate degradation would occur at a reactor pressure of 8 psig when only 3824 gpm would be realized versus the 4500 gpm requirement.
At the enforcement conference held or M cua 20, 1987, the licensee presented the results of an analysis W iet
. sed that a 10 degree increate in fuel peak clad temperature (PCT) would result from the combined effect of the partly closed valve and the LPSI pump impeller problem discussed in Paragraph 4.
This increase in PCT causes a 50% reduction of the safety margin and is ten degrees below the 2200 degree li:it of 10 CFR 50 Appendix K.
)
d
.
..
-
_
.
.
.
,
,..
b Long' term corrective actions include a valve improvement program, improved' post-maintenance testing, operator training and implementation
- of a. system engineer concept.
.
Based on the licensee's corrective action, LER No. 255/84017 is closed.
Summary
- During a 28 hour3.240741e-4 days <br />0.00778 hours <br />4.62963e-5 weeks <br />1.0654e-5 months <br /> period on September 2-3,'1984, while the reactor was' critical and operated above 2% power, Valve No. CV-3006 was found partially closed due to tight. packing and binding of the stem.'
In addition, post-maintenance testing failed to identify improper valve operation and indication. This constitutes a violation of TS 3.3.1.h and 3.0.3 which require-that Valve No.~ CV-3006 be fully open and disabled
prior to taking the-reactor. critical and during power operation
'
(Violation No. 50-255/86030-03).
The safety significance of the violation was exacerbated by the reduction in safeguards flow caused by the degraded LPSI pump flow (Paragraph 4).
One violation and no deviations were identified in this area.
- 6.
Component Cooling Water Background The component = cooling water (CCW) system consists in part of a heat exchanger (HX) ~and supplies, aside from its normal loads, the shell side of the shutdown cooling heat exchanger (SDCHX) that is used to remove heat from the discharge of the containment sump following a Loss of
' Coolant Accident (LOCA). Cooling of the CCWHX is provided by the Service Water.(SW)' system. Technical Specification (TS) 3.3.1.e requires that both CCWHXS and SDCHXs be operable during critical reactor operations.
TS 3.3.2.d allows one SDCHX to be inoperable for up to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> during power operations.
-
During the initial construction phase in 1969, Component Cooling Water i
Heat Exchangers (CCWHX) were procured by the Architect / Engineer (A/E).
The purchase specification called for two 5700 gpm heat exchangers and the A/E specified a flow capability of 5700'gpm through each of two CCWHX
- shells. The vendor supplied two heat exchangers capable of 5700 gpm per unit. The A/E and the licensee misunderstood the vendor's definition in that a " unit," as referred to $y the vendor, consisted of two heat
,
exchangers. The half capacity of these heat exchangers was the cause for the damage experienced by them during and after the preoperational testing.
- Subsequent to preoperational testing of the CCW system, it was found that high flows through the CCWHX were causing tube vibration and failure.
Per the manufacturer's recommendation, flow through the heat exchanger
'
i
9
-
,_.. _. _ - _. _.. _... _ _ _ _. _. _ _. _ _. _ __
_,. _... _.. _ _ _.._ _ ___ _
.
.
..
.
was throttled and differential pressure, on an administrative basis, was maintained between 10 and 15 psi.
Plant Review Committee (PRC) meeting minutes, dated June 9,1971, indicate that the CCW Heat Exchangers were experiencing tube degradation and that flow must be limited to prevent further tube damage, yet also be sufficient to maintain heat transfer requirements. The PRC meeting minutes additionally indicate that an investigation was to be conducted to determine the extent of the flow problem; however, there is no evidence that the licensee initiated an investigation of the CCW flow requirements at that time.
During a routine surveillance on May 19, 1986, (date of a plant trip),
a low CCW flow alarm to the safeguards pumps was received in the control room. The licensee's Material Condition Task Force addressed the low CCW flow alarm.
Investigation of the alarm disclosed that the safeguards pumps' bearings and seals were not being supplied with required FSAR design flow. Though the FSAR specifies a 140 gpm CCW flow, it is not clearly identified as a nominal value or an operability requirement.
Discussions with the seal manufacturer and pump vendors determined that flows of only 5 or 15 gpm per pump are. required for operability.
Further investigation into causes for low CCW flow raised concerns for CCW system performance. This led to performance of a flow test in late August.
Results of the testing showed that in the 15 psid throttled condition, flow to the SDCHXs was 5,700 gpm.
Unthrottled the flow was 7,950 gpm. The FSAR Table 9-7 lists 8,000 gpm as the design flow.
Investigation of CCW system performance disclosed the system's inadequate flow capability to the SDCHXs whereas the routine operability tests run on the CCW pumps, though in conformance with Technical Specifications and AMSE Code Section XI, failed to detect the system degradation caused by throttling the CCW flow downstream of the CCWHX.
An additional problem regarding the CCW system was identified during the review of the Service Water (SW) flow problem (Paragraph 8). During that review, it was identified that each SW flow control valve from the CCWHX received an open signal only from its associated protective channel.
In the event that the valve was closed as a result of maintenance and an open signal from its associated protective channel was not received, heat transfer capability of the CCW system would be cut in half. Although the valves are normally open during operation, this lack of redundancy l
exacerbates the operability question. This design was consistent with
!
the original intent that each CCWHX was to be a redundant component.
l
'
Corrective Action
!
An analysis by Combustion Engineering of the containment temperature and pressure during a postulated LOCA demonstrated that reduced CCW flow l
has a marginal effect on the second peak in containment temperature. The i
profile was sufficiently different, however, to raise questions about the environmental qualification of some equipment inside containment. These
'
concerns were resolved by shortening the qualified lifetime of affected components.
-
-
--
.
,
.
..
.
Consideration was also given to the impact that the reduced SW flow and higher SW system temperatures discussed in Paragraph 8 would have on the containment analysis. The SW system was found to be inoperable which required procedural and system modifications to restore operability.
When returned to operable status prior to startup, the SW problems identified will have no effect on the containment analysis.
A modification to provide redundancy to the service water isolation valve from the CCWHX is planned to be completed prior to plant startup.
Additional evaluation regarding the CCWHX by a contractor (YUBA)
determined that the heat exchangers could individually handle 4500 gpm without vortexing and vibration. The licensee implemented a conservative value of 2500 gpm or 5000 gpm total to the SDCHX. This was found acceptable due to the low sensitivity of the containment analysis to CCW flow rate. The effect was only a few degrees F increase in peak containment temperature when CCW flow was reduced from the FSAR design value of 8000 gpm to 5000 gpm.
In order to maintain the required flow to the SDCHX, the Spent Fuel Pool cooling will be throttled during plant operation. Additional action by the licensee will be required on this issue before startup after the next refueling.
Summary The Component Cooling Water system was incapable of meeting its design performance requirement from initial plant startup to present. The licensee does not know what the basis is regarding the 8000 gpm FSAR flow requirement for the SDCHXs. The original analysis could not be located.
The new analysis shows that the CCW system will not require modification aside from providing redundant protective signals to
,
actuate the SW flow control outlet valves.
An evaluation of the safety significance of this degraded condition presented by the licensee at the February 20, 1987 Enforcement Conference concluded that the combined effects of the existing CCW system configuration, and Service Water (Paragraph 8) and Containment Air Cooler (Paragraph 7) degradation, were that the peak containment pressure reached would be only a half a pound less than the design pressure and thereafter reduced more slowly.
Due to the undersized CCWHX exchangers, a violation of TS 3.3.1.e and 3.3.2.d which require both SDCHXs to be operable was identified (50-255/86030-04a). Causal factors in this event include failures of the test control and safety evaluation programs.
The licensee also failed to assure that the correct heat exchangers were installed and properly tested during both original construction and subsequent plant operation. This constitutes a violation of 10 CFR 50,
!
Appendix B, Criterion XI, Test Control (Violation 50-255/86030-04b). This resolves Unresolved Item 50-255/86023-03.
Two violations and no deviations were identified in this area.
i I
'
_ _ _. _ _
- _.
_. _._ _,-___
-
..
-
.
-
.
--
,
.-
..
7.
Containment Air Coolers i
Background Palisades has a total of four Containment Air Coolers (CACs). Three are in one ESF train along with one containment spray pump. The other train consists of two containment spray pumps and one CAC. TS 3.4.1.a and 3.4.1.c require operability of all four CACs while TS 3.4.3 and 3.4.4 allow two components in any one train to be inoperable for up to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
In 1983, all four safeguards fan / motor units in the CACs were replaced with environmentally qualified units built by Buffalo Forge and General Electric.
At the end of the last refueling outage (March 1986), a service water flow test to the CACs was conducted and a flow variance was noted between the CACs. Further investigation was deemed necessary and was included in the Material Condition Task Force Review.
The test was run to demonstrate valve position by a "go-no go" flow check through the heat exchanger in each CAC.
Since results of the test did not meet design valves, the licensee argued that test was not to determine flow rate and therefore the results obtained did not infer operability or inoperability. However, this same data was later used by the licensee for heat removal rate determinations in their Licensee Event Report (LER) submitted regarding the test results.
Further testing and analysis in March 1986 would have determined that there was degraded service water flow to the CACs.
As a followup to the above test results, additional investigation was done under the Material Condition Task Force.
The licensee found that the service water outlet valve (CV-0873) from the heat exchanger VHX-3 in CAC 3 was not operating correctly. At that time the licensee believed that the valve closed on a Safety Injection Signal instead of going fully open as designed. A 10 CFR 50.72 notification was made on that day for inoperability of CAC 3.
Further investigation determined the above not to be the case.
Instead, the valve actuator had been adjusted such that the valve did not fully open nor fully close by approximately 22 percent of full travel. This condition was also found to exist on the service water outlet valve CV-0864 from heat exchanger VHX-2 in CAC 2, which is in the same ESF train.
A licensee review of maintenance history information regarding the valves and their actuators disclosed no occasion where an aoJustment would have been performed. Consequently it is presumed that the condition has existed since original plant installation. This reduction in flow to the heat exchanger is considered sufficient to render the two CACs inoperable.
.
.
..
Another investigation into the operability of the CACs resulted subsequent to the failure of a fan motor (V-1A) in July 1986 in CAC 1.
On August 8, 1986, it was discovered that an access plate to the ducting between the CAC 3 fan (V-3A) discharge and heat exchanger VHX-3 was not in place. After an engineering analysis was completed, a 10 CFR 50.72 notification was made on August 25, 1986, reporting that the V-3A fan would not have met its design air flow requirement. The licensee has not conclusively determined when the access cover was removed. The bypassing of air flow through the access cover rendered the CAC inoperable for an indeterminate period of time. The licensee believes that the access cover was on when they started up from the last refueling outage, however they have no supporting documentation.
In addition, the licensee believes that the pre-startup walkdowns of containment or maintenance performed on the fans (lubrication) would have detected the access cover or loose bolts on the floor. The length of time that CAC 3 was inoperable due to low air flow cannot be conclusively determined, though it is postulated to have occurred during recent reactor operation.
(Note: CAC 3 was also inoperable due to inadequate service water flow as discussed above). The licensee also believes that bearing deterioration and excessive shaft movement of the V-3A fan motor, discovered during an inspection, may have been of sufficient magnitude to loosen the access cover belts.
It is noted that the excessive shaft movement resulted in stator damage due its contact with the rotor.
In the licensee's evaluation of CAC system performance, CAC 3, due to fan V-3A, was considered inoperable, while CAC 1, CAC 2, and Containment Spray Pump P-54A were considered operable.
Utilizing the suspect service water flow values to CAC 1 and CAC 2 obtained in March 1986, the licensee evaluated that CAC 1 represented a heat removal rate of 7.63 E7 BTU /hr, CAC 2 represented a heat removal rate of 4.28 E7 BTU /hr, and CAC 3 is not credited with removing any heat. The operable CACs therefore, represented a total heat removal capability of 1.19 E8 BTU /hr. Assuming that Centainment Spray Pump P-54A was operable, an additional 1.17 E8 BTU /hr heat removal capability was available. The resulting 2.36 E8 BTU /hr heat removal capability exceeded the FSAR requirement of 2.29 E8 BTV/hr.
Since operability of CAC 1 is assumed in their evaluations, it is noted that fan motor V-1A failed during non-accident conditions in July of 1986, which raises the question whether it would have performed very long during a LOCA environment.
In addition, V-4A and V-3A were also found to have bad bearings and were likely candidates for premature failure.
Corrective Action The licensee has adjusted both service water outlet Valves, No. CV-0873 and No. CV-0864, to provide proper opening and closing of the valves.
System and component service water flow tests were performed prior to declaring the CACs operable.
Improvements to the post-maintenance testing program are planned.
_
..
.
.
An evaluation of the fan motor bearing failure mode by GE concluded that
!
the root cause was over-lubrication which caused skidding (instead of rolling) of the bearings in the race. The lubrication had been conducted in accordance with GE's recommendations. GE also suspects that the
'
quality of the bearings utilized (Romanian Manufacture) had an impact on the failures. The damaged motor and the bearings in the other three fans have been replaced.
Preventive maintenance practices for greasing and vibration have also been revised.
All dampers and inspection doors were inspected and most dampers were weld repaired. The licensee plans to develop preventive maintenance to determine air flow.
-
During the Enforcement Conference held on February 20, 1987, the licensee evaluated the effect of the inoperable CACs on the containment pressure and temperature response during a design basis accident. The results of this evaluation showed that the worst case situation was acceptable, although by a significantly reduced margin.
Summary TS 3.4.1.a and 3.4.1.c require operability of the CAC's while TS 3.4.3 and 3.4.4 allow two components in the same train to be inoperable for up to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Two CAC heat exchangers VHX-2 and VHX-3, were inoperable due to degraded service water flow from initial. plant startup through operations in May 1986, exceeding the conditions allowed by Technical Specifications.
(The Palisades reactor has been critical for nearly 70,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />.) In addition, CAC 3 was inoperable due to degraded air flow through the cooler for an indeterminate period of time.
Finally, CAC 1 was made inoperable. in July,1986 due to the failure of the fan motor bearings (Violation 50-255/86030-5a).
10 CFR 50, Appendix B, Criteria XI requires that a test program be established to assure that all testing required to demonstrate that systems perform satisfactorily in service is identified and performed in ac.ccrdance with procedures incorporating requirements and acceptance limits contained in applicable design documents. The licensee performed no routine CAC system surveillances (Violation 50-255/86030-5b).
The service water flow testing conducted in March 1986 should have alerted the licensee to the degraded service water condition and inoperability of the CACs.
Failure to promptly identify and correct conditions adverse to quality is a violation of 10 CFR 50, Appendix B,
,
Criterion XVI (Violation 50-255/86030-Sc).
These determinations resolve Unresolved Item 50-255/86023-02.
Two violations and no deviations were identified.
__
.
..
8.
Service Water System Background Palisades has three Service Water (SW) pumps, two in one Engineered Safety Features (ESF) train and a single pump in the other train.
Technical Specification (TS) 3.4.1 requires all SW pumps to be operable when the reactor is critical. TS 3.4.2, 3.4.3 and 3.4.4 specify SW system'operabi.lity requirements during power operation.
Service Water (SW) pump impe11ers were replaced on all three pumps:
P-74 in 1980, P-78 in 1983, and P-72 in 1982. The original pump impellers had been backfiled but the replacement impellers were not. Testing at this time was not adequate to reveal the degraded pump performance.
In August of 1982, and as reported in LER 255/82-24, SEP Topic IX-3,
" Station Service and Cooling Water Systems" was evaluated and it was concluded that the existing SW system, with two of three pumps having the same power supply and two required to meet analyzed cooling loads, was unacceptable. The licensee evaluation also concluded that a LOCA with loss of offsite power would result in loss of instrument air, failure of the SW flow control outlet valves from the Component Cooling Water Heat Exchanger (CCWHX) to the full open position and the subsequent starving of SW to the Diesel Generators (DGs) resulting in DG failure.
Resolution of these concerns included installing hard stops on the SW flow control outlet valves and revising procedures to require alignment of fire water to the service water header if the diesel generator supplying the two service water pumps fails.
SW system pressures and computer modeling were used to conclude that there would be adequate SW flow. The computer model assumed original pump curves which were no longer valid due to impeller replacements.
As a corrective action following other ESF pump operability problems, the licer.see tested the SW pumps on September 25 and 30, 1986. The FSAR required flow of 8000 gpm at 140 feet of head was not met for any of the three pumps.
P7A was 7330 gpm at 135 feet, P78 was 7323 gpm at 134 feet, and P7C was 7503 gpm at 136 feet. The pumps were declared inoperable at 2:30 a.m. on September 30, 1986, and a four-hour non-emergency report was made at 3:02 a.m. on the same date.
Discussions between operators and the system engineer during Service Water System (SWS) testing identified that the design SW inlet temperature had been exceeded during periods of summertime operation. On at least one occasion (July 19 through August 5, 1983) the design value of 75 degrees F was exceeded by up to 9 degrees. This will require further evaluation by the licensee to determine the impact on the LOCA analysis of containment and system operability. While SW temperature had been monitored, the significance of elevated temperature relative to the operability of critical loads had not been previously recognized or addressed.
e
..
Corrective Actions In order to ensure sufficient SWS flow so that the plant.is not limited in operation by lake water temperature, the licensee modified the pump impellers to restore the original pump flow characteristics, modified the CCWHX bypass valves to close on a Recirculation Actuation Signal, changed CAC heat exchanger VHX-4 SW outlet valve to isolate on a Safety Injection Signal actuation, and tested and balanced the system. ~ Provisions were also made in the Emergency Operating Procedures to align fire water pumps to supplement SWS, and the licensee will establish a new SWS limiting temperature value until fully redundant SW flow capability can be achieved. Component flow requirements were verified or generated and a flow balance of the system was performed to optimize flow distribution.
As previously discussed in the CCW and CAC paragraphs on corrective action, licensee evaluation of the impact of the deviations from design requirements was favorable, though significantly reduced margins for containment temperature and pressure were identified.
Additionally, the ESF pump room cooler was evaluated for accident conditions using the elevated service water temperature and reduced SW flow.
It was found that the temperature exceeded the design conditions for the room.
Summary The service water pumps have been inoperable since impeller replacement in 1980-1983.
Critical operation since that time was a violation of TS 3.0.3 because operation continued in configurations in excess of those addressed in TS 3.4.1, 3.4.2, 3.4.3 and 3.4.4.
(Violation No. 50-255/86030-6a).
When the impellers were replaced the licensee failed to maintain adequate control of the original design.
10 CFR 50, Appendix B, Criterion III, which requires that measures be established to assure that applicable regulatory requirements and the design basis are correctly translated into specifications, drawings, procedures, and instructions, and that measures shall be established for the selection and review for suitability of parts that are essential to the safety-related functions of systems and components. Measures are also required to verify and check the adequacy of design (Violation No. 50-255/86030-6b).
10 CFR 50, Appendix B, Criterion XI requires that a test program be established to assure that all testing required to demonstrate that systems and components will perform satisfactorily in service is identified and performed in accordance with written test procedures which incorporate the requirements and acceptance limits contained in applicable design documents. The licensee's testing failed to show the degraded pump performance for five years (Violation No. 50-0255/86030-6c).
These determinations are the resolution to Unresolved Item No. 50-255/86027-02.
.
Three violations and no deviations were identified in this area.
r l
l
l l
i
-
7-
-
o
..
9.
Management Interview Management interviews were conducted as indicated in Paragraph 1.
The scope and findings of the inspection were discussed. The inspectors also
'
discussed the likely informational content of the report with regard to documents or processes reviewed by the inspectors during the inspection.
The licensee did not identify any such documents / processes as proprietary.
l
.
i i
i
!
l l
l l
.
- - -
. _ -. _ -
.-_ __
_. - _ _ _ - _., _.
. _
.
.
.
......
..