Inspection Report - Trojan - 1987031 |
|---|
|
|
Text
.
.-
.
U. S. NUCLEAR REGULATORY COMMISSION REGION Y Report No:
50-344/87-31 Docket No.
50-344 License No.
NPF-1 Licensee:
Portland General Electric Company 121 S. W. Salmon Street Portland, Oregon 97204 Facility Name:
Trojan Nuclear Plant Inspection at:
Rainier, Oregon Inspection conducted:
September 21 - October 2, 1987 Inspectors:
/l)
/74-#7
,
W.' P. Ang, Team L'eader Date Signed b YYM
/z// 7/8 *7 J. F. Melfi, Reactor Inspector Date Signed-by YN*g N
f'.
y f r-/4-gy J. E. Tatum, Resident Inspector Date Signed
&
h
/2/N/7 l
L. M. Padoven, Resident Inspe tor Dat6 Signed-Y&
,0
/2 //C/97 J. R Ball,Ae ide t Insp'ec or Date. Signed f
k
/2.//7/f 7 V
R. J.~Jolliffe, Inspectefr, Office for the Dafe Signe'd Analysis and Evaluation of Operational Data Consultant:
B. L. Collins, G&G Idaho, Inc.
Mr
/2//c/j7 Approved by:
-
S. A. Richardt, Chief. Eng}oeering Section Date Signed Summary:
Inspection on September 21 - October 2,1987'(Report 50-344/87-31)
Areas Inspected: Routine, announced team inspection of the Trojan Nuclear Power Plant. The inspection focused on the ability of the plant to learn from-previous experiences in the nuclear industry and from. experiences at Trojan.
.
'8712310209 871217'
PDR ADOCK 05000344 G
DCD
I
.
.
-2-The Instrument Air System, the Emergency Diesel Generator Air Start System and the Main Feedwater System were selected as the sample systems for inspections.
This inspection was performed by six NRC inspectors and one NRC consultant.
Inspection procedures 99021B, 25578, 25583, 30703, 37700, 41700, 41701, 62700, 62702, 62705, 71707, 71710 and 92700 were used.
,
Results:
In the areas inspected, six violations of NRC requirements and one deviation from UFSAR commitments were identified.
i i
-_.--______._.._________________a
- - - _
.
.'
-
.
DETAILS 1.
Persons Contacted Portland General Electric
- W. Lindblad, President, Generating Division
- D. Cockfield, Vice-President, Nuclear
- C. Olmstead, Trojan Plant General Manager
- L. Erickson, Manager, Nuclear QA Department (Acting)
- A. Roller, Manager, Nuclear Plant Engineering i
!
R. Schmitt, Manager, Operations and Maintenance D. Swan, Manager, Technical Services j
'
- D. Wheeler, QC Supervisor
- E. Davis, Manager, Electrical Engineering Branch
- D. Bennett, Maintenance Supervisor
- P. Morton, Plant Engineering Manager
!
State of Oregon
- H. Mooney, Oregon Department of Energy NRC Resident Inspector
- R. Barr, Senior Resident Inspector G. Suh, Resident Inspector In addition to the individuals identified above, various other engineering, quality assurance, maintenance, and operations personnel and other members of the licensee's staff were contacted by the team and attended the team exit meeting.
- Attended Exit Meeting on October 2, 1987.
2.
Instrument Air (IA) System The inspection of the IA system was divided into four areas; design,
~i testing, maintenance and operations. The review performed and resultant findings are described below.
,
A.
Design The team evaluated the adequacy of the IA system design by assessing the design basis, design calculations, and design modifications made to the system. Although the air system is not classified as a safety related system, past industry experience has shown that air system failure or degradation can negatively affect other systems
,
important to safety. Section 9.3.1, " Compressed Air System," of the
!
Trojan Updated Final Safety Analysis Report (UFSAR) end piping and i
instrument diagram (P&ID) M-223 " Instrument and Service Air System"
'
were reviewed for consistency, and to provide the inspectors with an overview of the Trojan plant IA and Service Air (SA) systems.
Plant
.
,-
.
walkdowns of-the IA and SA systems were performed to familiarize the inspectors with the installed plant. equipment, and to assess the as installed IA.and SA configuration and condition. Additionally, discussions were held with the system engineer and Nuclear Plant Engineering (NPE) representatives from the licensee's Portland headquarters office.
(1) Design Basis For the Instrument and Service Air systems at the Trojan facility, no design basis was found to exist. No documented
'
analyses were available to 1) verify that the air supply had adequate capacity to supply all of the loads demanded of the system or 2) determine what' effects future air loads would have on the air system capability. The only. documentation the licensee had.regarding system capacity consisted of comparisons of air supply capacities with other plants in the industry. At the team exit meeting with licensee management, the licensee stated that air system design basis documents were scheduled to be assembled in 1989. The completion of this action will be followed as part of the NRC review of design basis documents.
(2) UFSAR Commitments Page 9.3-4 of the UFSAR stated that " Required air cleanliness is maintained by the following features:"
"oilless compressor cylinder"
" filters installed in all lines to instruments and valves."
Contrary to the above UFSAR description, the team identified the following:
(a) A Sullair rotary screw compressor had been added to the
'
system. An air / oil mixture is used to. lubricate the compressor screw.
(b) The use of in line filters was recommended in some valve'
vendor technical manuals reviewed by the team and in AE0D/C701 " Case Study Report--Air Systems Problems at' U.S.
Light Water Reactors." The licensee indicated air lines
-
in which pressure controllers were installed "might" have
_,
integral filters in the controller bodies.
However, air lines to the main steam isolation valves (MSIVs) and the air lines to the isolation valves supplying main steam to the' auxiliary feedwater (AFW) pump Terry turbine were without controllers ~and did not have separate in-line-filters. This was identified as Deviation 87-31-0.
.m
,
,
(3) P&ID's The P&ID drawings for the air system contained errors and were incomplete. Air receiver tanks T-118 A, B, and C were shown on the P&ID to have capacities of 550 cubic feet each, instead of.
.
the installed (and UFSAR described) individual tank' capacities of 57 cubic feet. The " temporary air dryer" was not installed in the field as shown on the P&ID.
Furthermore, P&ID diagrams for about 22 of 32 IA and SA. lines downstream of the air headers existed only in draft form and consequently were not available to plant operators and the engineering staffs.
'I.t.is unclear when these drawings would be available for use.
(4) Design Modifications The control and documentation of the Instrument and Service Air System design activities at Trojan were considered to be inadequate. Design modification packages were reviewed and found to be' incomplete and incorrect. Design requirements i
could not be provided, calculations had not been performed, and drawings were not updated to'as-built conditions.
The design package (RDC-76-291) supplementing the Joy air l
compressors with a Su11 air screw type compressor, as..the primary source of instrument and service air, was reviewed.
According to the system engineer responsible for the air systems, this design change also included 1).the installation of a coalescent and particulate filter for removing particulate and aerosols (oil) from the Sullair supplied air,
'
2) the installation of a temporary air dryer allowing the main dryers to be bypassed, and 3) modifications to ' instrumentation monitoring the IA..The package was difficult to obtain data from as it consisted of several folders in a cardboard storage
'
box and appeared to have little organization. Most of.the design information dealt with the physical installation of-the equipment and indicated that Bechtel was responsible for the design of the system. However, there were-no specifications for the design requirements, no calculations'for supporting the design and hardware chosen, no material ~ requirements, and no testing instructions.
The design modification package contained no information about the installation of the " temporary air dryer." The design should have addressed the temporary dryer 'since it appears 'on-the P&ID (drawing No. M-223)'for the air system and is.
l considered to be.a permanent installation.
However, the dryer skid is not anchored to.the floor; is connected to the air system using undocumented flexible rubber air hose loosely,
draped over existing plant piping; and is' powered' from an
'
ordinary wall outlet using an extension cord.. A flange-connecting the. flexible hose-to valve SA-046 was missing three of its eight bolts. The temporary dryer'only had the' drying capacity of approximately 1/2 the capacity of_the main drying system. Therefore, a bypass'line around the main dryer was-
. - _. - -
. _ _ _ - - _ -
.
.-
-
.
..
required to be partially open for operation of the temporary dryer. The untreated air from the bypass was combined with the
-
air from the temporary dryer and could compromise the quality l
of the air supply. The above concerns regarding design,.
installation and controls for use of the temporary air dryer are considered to be an open item, pending licensee review and evaluation of the adequacy of the installation.
(0penItem 87-31-02). -
(5) Design Conclusion The NRC' inspection team concluded that'PGE does not fully understand the design of the IA or SA-systems at Trojan.
Design engineering personnel did not appear.to have open communications nor adequate interface with the plant. _Also,
-
the system engineer did not appear.to have an in depth-understanding of the system. -Additionally, several areas'were-a noted were engineering involvement in modifications to,the IA system was poor.
,
Lack of rigorous engineering was evidenced in the weak j
- documentation of system design an.d modifications. No IA design bases could be provided, few calculations were"available, and system drawings did not accurately represent'as-built'
conditions. The Trojan UFSAR was utilized as the' design,
documentation, yet differences existed between the UFSAR, vendor technical manuals, and the hardware installations.
B.
Testing I
USNRC Regulatory Guide 1.68.3, "Preoperational Testing of Instrument '
and Control: Air Systems," April 1982, specified that plants which undergo major modifications to the IA system, or portions thereof, should perform both the rapid loss of IA supply.. test-and the gradual i
loss of IA supply test prior to plant restart. Additionally, the
'
Guide recommended the loss-of-air supply tests should be conducted on all branches of the IA system simultaneously, if practicable.'
NRC'Information Notice Numbers 82-25,.85-35,.86-50, and AE0D/C701 also discussed the importance of performing gradual loss-of-air testing.-
In'1979, the Trojan IA system was modified to add a fourth air-(Sullair) compressor (C-?16) to increase ~ plant air capacity. The gradual ~1oss of IA system test was never performed on the Trojan Plant. As safety related backup accumulator testing'has been-accomplished.for only a few selected accumulators, and testing did not include a simultaneous slow loss of air test, there is no solid evidence that backup accumulators at Trojan will perform their -
intended function...The. licensee stated that Trojan has not committed to the requirements of Regulatory Guide 1.68 3 ' April-1982,' and that they did not feel _ it necessary to perforra this ' test.
However, the licensee further stated that they.would reevaluate
.their position regarding.this' matter.
(0penItem 87-31-03).
,
-_
_ _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ - _ _
.
.
.
.
. Periodic Engineering Test.14-2, " Seismic I Accumulator Leakage Test," was performed to check for leakage from the backup air accumulators on AFW Terry turbine steam supply valves CVs 1451, 1452, 1453, and 1454. The test involved having the accumulators-hold the steam supply valves openLfor a two hour period. All.four steam supply valves failed to pass the test. Repairs to the valve-operators and IA system were performed and all valves subsequently-passed.' However, the. procedure was revised to require a hold time of only 20 minutes. An NRC Safety' Evaluation Report dated October, 1980,:specified that AFW flow would be ~available for a minimum of two hours after a total loss of all AC power. A licensee telephone
~
memorandum dated July 29, 1987, documented a telephone conversation between the licensee's D. Swanson, K. Bohlander, and B. Sherwood on this subject. The memo indicated that once the steam supply' valves were opened. steam pre'ssure would hold the valves open._ However,'no objective evidence was available to support this conclusion.
..
Consequently, pending 11.;ensee evaluation and determination of the.
adequacy 'of testing of the AFW steam turbine supply valves with'a hold period of only 20 minutes, this is considered an unresolved item..(UnresolvedItem 87-31-04).
TrbjanplantTechnicalSpecificationSurveillanceRequirement(SR)
4.7.1.2.1, in part, requires verification that the AFW steam turbine-driven pump started and developed pressure in accordance with SR 4.0.5.
SR 4.0.5 requires inservice testing (IST) of the pump and its associated valves (including CVs 1451 through 1454)
every 92' days in accordance with Section XI' of the ASME B&PV. Code.
ASME B&PVC, Section XI-1983 " Rules for Inservice Inspection of Nuclear Power Plant Components," Division:1, Article IWV-3415
.
" Fail-Safe Valves" specifies "When practical, valves with fail-safe actuators shall be tested by observing the operation of the valves upon loss of actuator power." CVs 1451 through'1454 have their actuator power (IA) backed by safety _ related air accumulators.' The licensee indicated IST of CVs 1451 through 1454 was being performed with IA available to the valve operators, rather than relying on the air accumulators as required by the code.
In addition,'the AFW'
Terry turbine has not been verified to start by operation of the steam admission valves using only accumulator air. -This omission is an apparent violation of the Technical Specifications.
(Violation 87-31-05).
Additionally, required IST of. Quality Group 2 check valves
,i separating the back-up air accumulators from the IA system (such as
!
MS 226 and 222) has not been performed. These omissions from the licensee's IST Program (Topical Report PGE-1048) are an apparent violation' of 10 CFR 50, Section 50.55 a(g). -(Violation 87-31-06).
The licensee should re-evaluate the adequacy of IST performed on all accumulator. backed valves.-
During routine surveillance: testing (calibration)'of IA'
q instrumentation, testing of remote functions has not verified.
For
,
example, when calibration of a pressure switch was performed, only.
the retpoint of the pressure switch was verified and its function to i
,
___-_x_____-_
__
_ _ _ _ _ _
_ _ _ _ _
~*
t
'*
(
6-i
.
.
/
initiate an alarm in the control room or to activate a-soletioid valve in the SA system was not verified. Separate system testing was not done to verify these functions.
PGE has no verification-that entire: instrumentation circuits on the IA system are integrally
~
tested. Thus, instrumentation could detect unacceptable IA conditions without. compensatory-actions being taken, which could.
result-in plant transients.
Pending licensee Integral testing of IA
,-
'
instrumentation circuits, this was identified as an open item.
,
(0penItem 87-31-07).
q The above described items illustrate PGE's lack of a complete understanding.of the.IA system and inadequate follow-up of j
documented industry problems.
Furthermore, testing of the'IA system
,
did not appear to adequately verify system functionality.
i
,
C.
IA Maintenance
!
The inspection team evaluated. administrative aspects of the maintenance program, focusing.primarily on those elements that would
>
,
tend.to identify and implement lessons learned from industry ar.d l
s previous Trojan experiences. The inspection team also evaluated actions' taken by Trojan to preclude recurrence of selected problems previously: experienced either at Trojan or elsewhere in the nuclear industry.
Based on this inspection, it appears that significant c-
"
'
improvements can be made.in the maintenance program to take advantage of' industry experiences and previous Trojan experiences.'
In this regard, the inspection team identified weaknesses in the use
>i
'
I of procedures and maintenance requests, implementation of vendor
'
j recommendations, and plant trending.
In addition, for selected examples,.it' appeared that Trojan had not implemented thorouch
'
corrective actions to preclude recurrence of problems previously
-
experienced either at Trojan or within the nuclear industry.
Paragraph 4, Maintenance, of this inspection report, contain examples noted.
During~ the'IA system walkdown, the. team identified a spring can supporting the safety related' portion of the IA line to the MSIV-
-actuator in main steam support cubicle A that was disconnected.
,
Trojan nuclear plant drawing number HBD-80-50 delineated the installation of the H 834 support. Quality Assurance Criterion V
" Instructions,- Procedures, and Drawings" of.10 CFR 50, Appendix B,
stipulates " Activities affecting, quality shall be prescribed.
l by... drawings,...and shall be accomplished in accordance with -
these... drawings." Accordingly, the uncompensated disconnection of the IA line spring can while at power represents an apparent violation of 10 CFR 50, Appendix B (Violation 87-31-08).
,
g
'
The plant " instrument.index" listing IA instrumentation setpoints'
'
q
~
.
was found to be incorrect. < Twelve IA instrument setpoints on the.
index were-reviewed for accuracy, with the resulting finding that:
the following nine were incorrect. relative tq the actual calibrated condition. '
-
'
..r
,
s
,
.
m)
j { ] %.
- '
f} Q]
~~
., 1p 3,
j e
.
,
.,
,
,{
i 7-
"
'
,.
.
.y i.'.* [7h
,','
P
"M H
.
l'
.
"
N\\ ;
,
,
Instrument No.
Instrhnint Index SetpoiNt h,Ac,tual Setpoint,
.
"
10 PSI,0 10 PSID
'
'
PDIS 44t0, 44 ~ k]. -
%,a
25 PSIk
105 PSIG'
')
PS
,,.
44? -
"I
'
PS 4463
_f 1% PSIG 100 PSIG 103 PSIG
PS v
o
. $ PSID (
%'
S 10 PSID PDIS 4468
-
i
,
i '
L TS 4473 400'F 450 F-t.
,1
.
TS 4474 400 F 450 F
'
.,
[
TS 4478A J
/ 350 F 375 F-r 350 F 375 F-TS 44788
-
.,
y-
_
s 3.
Thitefore,qMMaduld review all instrument index setpoints for the IA/SA systea.and correctly document existing system instrumentation yApoints.
(0penItem 87-31-09).
,
,
~
_t
,
- boroge of un$psured prefabricated piping sections in the
^
l
'lemporary
was noted by tih team dwin the inspection. These prefabricated
- pipingassembliesincludedk,fivefootsectionofoneinchpiping
,.N l formed in the shape of a tuning fork, a' four foot length of angle /,
s
,
Durh.9 a y, y @" :tel, and a straight eipt foot section of one inch pipe.ehmic even
+
'
s
' guerator E 2010) could p mpacted by this unsecured piping. T0is x
piipn;(waseventuallyrespvedwhenquestionedbyan:NRCinspector.
s s
In
ummitment on dewpoint was thus ~40 degrees C, the
'
'
do<pthnt al.Pm on the IA system at Trojan was set significantly less i
j
-
co.sprvativtey aty.6 degrees C.
No basis for 61s setpoint was.
'
'
C E ntified, Trending of the IA dewpoint by the sy4 tem engineer l4andom strip chart from the dew-point recorder, the in3pector:indicated A
'
f
,% nbserved the dewpoint monitor went into the alarm cnndition 3 times'
.
s in a 48 hour5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> period. Orf this basis, it appeared to tbt term that.
'y
'
the state of IA quality was not well monitored by site engineering l',
N
,
or the operaticr Q tsff.
,
,
.L
<>
On May 7-12, 1987, Trojan personnel performed the first blowdown '
test of the entireLIA/SA system since initiating commercial
>
oper3tions in May 1976. A review of the test results (REF:
.
yH'
Mainuncrde Request (MR) 87-2482) irldjcatedthatonlyapproxim ely (3
[N
s
%
A
.
T (
j Q
%
))
\\,.
qf f
_p
'
2d.
r
~
m
_
,
-
,
.
_B'
s
>
,
,
.
77% of the IA/SA system valves had been cycled during tn'is test.
'
The remainder of the valves had not been cycled because they were
,"
located in high radiation areas er in areas where otWr work was in progress. During the test, about two cups of water were collected
-
from one valve, and particles of rust and a spot of oil were collected on a rag from other valves. The licensee stated that because of the good results of the first blowdown test, they had not decided when the next test would be performed, but it probably would not be performed on an annual cycle. However, the ifwqsee agreed during the inspection to reevaluate their position on hva often they plan to perform this test. Selected components are blown down each week by plant operators. However, the blowdowns are performed without procedural guidance as to'.what components are to be blown down, and without documentation of the results other than a control room log entry that shift routines were performed.
In contrast, a vendor technical manual for one filter recommended daily blowdowns.
The issues of an annual IA/SA system blowdown, including the need to
..plowdown M 1 high radiation area valves, and procedures for weekly
ulowdowns will be followed by the NRC.
(0penItem 87-31-10).
The team noted that licensee _ procedures allow the IA dryers and filters to be totally bypassed without evaluation of the effects of moisture and particulate on the IA system and air ~ actuated i
components. ' This action place 6 the IA system in an indeterminate '
air quality condition without compensatory measures being instituted, such3as additional air quality verifications or limits on the duration of time the system was in the degraded condition.
Lack of compensatory measures and time limitations.are an open item.
'
(0penItem87-314i1).
-
<
Discrepancies were also observed with licensed odr'ator training on
,
I the IA system. The " License Training Program t%ciment - Instrument
and Service Air System" contained the following incorrect V
information:
.
i o
incorrect! tank capacities for T-118 A, B, C, and T-160
'
i o
incorrect delta P setpoints o
missing particulate coalescent filter diagrade In summary, the above design, testing, and operations inconsistencies, coupled with omissions and inadelp;acies, demonstrated'that Trojan plant and industry IA experiences have not been properly factored into the
'
control of the IA/SA system at Trojan. The team also concluded that the system was not being operated in an optimum manner.
3.
Emergency Diesel Generator (EDG) Air Start System
The EDG Air Start Systow,is described in the Trbjan UFSAR section 9.5.6 and illustrated in Figures 9.5-10 and 9.5-11 of the UFSAR. AE0D/C701 J
Case Study Report on Air Systems Ptoblems at U.S.' Light Water Reactors noted that many events have been reported which involved a failure of an EDG to start as a result of poor _ starting system air quality. An l
,
,
.i g
.
s-
. i.
,.
_ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _
~
,
.
.
inspection of the Trojan EDG Air Start System was performed to determine if it was susceptible to similar problems identified in the AE0D report.
The inspection was also performed to determine what licensee action had
'
been taken with regard to the EDG Air Start System based on industry and Trojan experience.
A.
Design The reconstituted design basis document (DBD) for the EDG's and supporting systems had been issued prior to the inspection. The DBD for the EDG Air Start System was reviewed and used during the inspection. Calculations for system capacity and other parameters were noted as an open item in the DBD and were still to be obtained or re-performed. The licensee stated that DBD open items were being considered for inclusion in the licensee's commitment tracking list but had not yet been included during the inspection.
In addition, completion dates for the open items were still being developed.
An inspection of the installed EDG Air Start System was performed using,the following documents:
(1) Trojan UFSAR Section 9.5.6 (2) The reconstituted DBD for the EDG's (3) Vender Technical Manual M16-90, Revision 6 EDG and Accessories (4) Bruce GM Diesel Inc. Drawing A071F07001, Revision 9. Schematic Diagram - Air Start System for Tandem Engine Unit The following discrepancies'were identified between the above noted docume;.ts and the as-found/as-built' EDG Air Start System:
(1) The EDG Air Start System schematic diagram, the vendor technical manual, and the DBD, require pressure-reducing valves to reduce system air pressure from 250 psig to 200 psig. UFSAR Figure 9.5-10 required the same valves to reduce the system air pressure from 300 psig to 200 psig. The as-found pressure reducing valves (8 valves) were found to be set at downstream pressures ranging from 180 psig to 190 psig. The licensee was unable to provide instructions, procedures or records for the reduced setting or technical justification of the reduced setting.
There also was no 10 CFR 50.59 evaluation record for the reduced setting. This appears to be a violation of 10 CFR 50.59 in that changes to the system operation as described in the UFSAR did not appear to have been evaluated for a potential unreviewed safety question. This is identified hs violation 87-31-12.
(2) The locations of the eight lubricators and eight relay air valves were found to be interchanged from that shown on UFSAR Figure 9-5-10, the air start system schematic diagram, the vendor manual and the DBD. This appears to be a violation of
.
-
-
.
.
10 CFR 50 Appendix "B" Criterion V, and is identified as violation 87-31-13.
(3) One of the eight relay air valves had a pressure gage connected to it that was not shown on any of the above noted documents.
Similarly, a component that appeared to be a pulsation dampner was found to be installed in one of the eight air start systems on the tubing to the governor air boost. These were identified as additional examples for violation 87-31-13.
(4) UFSAR Figure 9-5-10 requires pressure reducing valve pressure gages to be 0-200 psig gages. The air start system schematic diagram required the same pressure gages to be 0-300 psig.
Seven of eight as-found gages were 0-300 psig gages and the eighth gage was a 0-200 psig gage. The 0-300 psig gages appeared to be more appropriate than the 0-200 psig pressure gage since pressure is required to be set at 200 psig. At that normal pressure, it would not be locally possible to tell if a gage reading of 200 psig indicated a correct reading or if the gage was pegged high and actual pressure exceeded 200 psig with 0-200 psig gages. The gages could be required for setting the pressure reducing valves. This was identified as an additional example of violation 87-31-13.
No periodic monitoring of EDG Air Start System air dew point appeared to be performed.
Furthermore, there was no readily apparent rationale as to whether or not dew point monitoring was required. This was identified as inspector follow-up item 87-31-14.
B.
Testing and Maintenance A review of periodic testing and maintenance of the EDG Air Start System was performed and discussed with the system engineer and licensee maintenance personnel. Maintenance Procedure (MP)-12-7, Revision 16, Emergency Diesel Generatnr Plant, provided the periodic inspection and maintenance requirements for the EDG's, including its air start system. A PGE letter dated August 29, 1986, to NRR submitted the second 10 year IST program for Trojan. MP-12-15 provided instructions for inservice testing of valves at Trojan.
The EDG Air Start System air receiver relief valves (8) were not included in the above noted maintenance and inservice testing
,
documents. PGE Nuclear Quality Assurance Program, Supplement 1, commits to ANSI N 18.7-1976. ANSI N 18.7-1976, paragraph 5.2.8, requires that a surveillance testing and inspection program shall be prescribed to insure that safety-related. structures, systems and components will continue to operate, keeping parameters within normal bound, or will act to put the plant in a safe condition, if they exceed normal bounds. The licensee had no record that the EDG Air Start System air receivers relief valves (8) had ever been tested, inspected or set since original construction. This appears to be a violation of the QA program commitment to ANSI N 18.7-1976 and was identified as violation 87-31-1.
l
.
EDG Vendor Technical Manual M16-90 provided maintenance instructions l
for the EDG air start lubricator. The manual required that the lubricator needle valve be adjusted to permit three drops of oil per minute to be deposited in the air line and required further adjustment of the needle valve if the exhaust air shows excessive oil. The licensee had no record, instructions or procedures to provide for the lubricator needle valve setting or adjustment. This was identified as inspector follow-up item 87-31-16.
MP-3-5, Revision 9, Cleanliness Control, provides the Trojan cleanliness and foreign material exclusion control requirements.
It was noted by the team that the procedure did not apply to the IA and EDG Air Start Systems.
In addition, the team noted that teflon tape was used extensively in both the IA and the EDG Air Start Systems.
The licensee had no instructions or procedure for use of teflon tape. Specifically, no precautions were provided to preclude introduction of teflon tape into the IA and EDG Air Start System, thereby potentially causing problems with equipment operability or j
blockage. This was identified as inspector follow-up item 87-31-17.
C.
Operations The operability of the EDG Air Start System was discussed with on shift operations personnel.
The system had low pressure alarms for both the air receivers and the pressure reducer downstream air pressure. During these discussions, a shift supervisor stated that if attempts to correct the low pressure conditions were unsuccessful, the condition would be referred to engineering for an operability determination. The inspector noted that this operability evaluation would be very difficult, if not impossible, without the system design basis calculations that had been previously noted in this section as not being currently available.
The team concluded that the licensee should prescribe what ccastitutes operability for the Air Start system, such as is done for the EDG fuel oil system.
The EDG Air Start System includes an air compressor and four air receivers for each EDG.
UFSAR Section 9.5.6 states that one of the compressors is equipped with a dual-drive 480 V AC motor and diesel engine with a hand crankshaft system. An operations shift supervisor stated that efforts to correct system low pressure conditions could require using the diesel engine drive for one of the air compressors. The inspector noted that the only procedure available for starting the air compressor diesel engine was a hand written one page procedure posted next to the compressor. An auxiliary operator, however, was readily able to start the diesel engine.
Further discussion with both the shift supervisor and the auxiliary operator indicated that maintenance personnel would have to be called to change over the belt from the AC motor to the diesel engine. The licensee had no procedure for the belt changeover and tightening requirements of the vendor technical manual.
In addition, it appeared that the belt changeover would be difficult since the compressor was bolted to its mounts, the diesel engine exhaust appeared to be hard piped to the room over head, and the
- _ - - _ _ _ _ _ _ _
i
.
i
.
vendor technical manual did not take this into account. The licensee had no record of the belt changeover ever being accomplished or the diesel engine being started by hand with the belt attached to the compressor. This was identified as inspector follow-up item 87-31-18.
The EDG vendor technical manual specified that the air receivers be blown down weekly and the wye strainer blowdown valves be blown down on a routine schedule.
Blow downs of the air receivers were
accomplished by the licensee as part of the Saturday midshift routines. However, the only record of these blow downs was annotation on the control room log that shift routines were accomplished. No record was kept regarding the results of the blow down; ie, water, oil, rust. Blow downs of the wye strainer were included as part of the MP-12-7 six month inspection of the EDGs, along with annual cleaning of the air start in-line strainers. The apparent lack of specificity of the blow down results were identified as inspector follow-up item 87-31-19.
4.
MAINTENANCE PROGRAM The inspection team reviewed the licensee's program for administrative control of preventive and corrective maintenance of the IA system, main feedwater system, and the EDG air starting system. This review included maintenance procedures, operating procedures, test procedures, vendor technical manuals, nonconformance reports, maintenance requirements, maintenance schedules, maintenance requests, maintenance records, and the existing material condition of the selected systems.
Interviews and discussions were conducted with personnel from plant engineering, system engineering, maintenance, maintenance planning, quality control, quality assurance, and the administration departments, as well as various craft personnel, a.
Plant Procedures and Maintenance Requests A review of various maintenance department procedures indicated that maintenance procedures lacked detail.
For example, MP 3-5 does not appear to adequately address the criteria for assigning quality control cleanliness inspection points, especially on Class
"C" and
"D" Systems (which include quality related systems).-
Maintenance work on the Main Feedwater System and the Auxiliary Feedwater System is required to be performed in accordance with the Class "C" cleanliness requirements specified in MP 3-5, Revision 9.
This procedure states that it is the responsibility of the work group supervisor to determine cleanliness controls and quality control cleanliness inspection requirements for maintenance work on plant systems. The maintenance requests (87-4112, 87-4113, 87-4116 and 87-4117) prepared to rebuild the air actuator cylinders for the
,
Auxiliary Feedwater System steam supply valves (CV-1451, CV-1452, CV-1453, and CV-1454) included a requirement to maintain Class "C" cleanliness per MP 3-5, Revision 9.
However, of the four MRs, only MR 87-4113 (CV-1452) specified a quality control inspection.
Maintenance planners pointed out that only MR 87-4113 included breaking into the system, i.e., disassembling the valve, repairing
_ _ _ _ _ _
_ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _
h
.
.'
'
.
.
.
the-internals,'and reassembling the valve, whereas the other three MRs did not include breaking into the system. While this. method of-operation appeared to be reasonable, cleanliness procedure MP 3-5
.did not appear to distinguish this situation. Another example: MR l
85-2346 was prepared to-rebuild four emergency diesel' generator air starting valves.in the maintenance shop and return them to-the'
warehouse for stock. MR 85-2346 specified a Class "A" cleanliness level in accordance with MP 3-5.
When asked why a quality control cleanliness inspection was.not'specified on the MR, a maintenance.
planner stated that a QC cleanliness inspection was. not necessary because the work was done in the maintenance shop. While this approach may be appropriate, this situation is. not addressed in MP 3-5.
'A review of various maintenance requests also indicated that the maintenance requests were lacking in specificity. The following maintenance requests were typical of those examined by the inspectors:
(1) MR 83-2707 was written to address a failure of 14"'feedwater
' isolation valve, M0-29710, to open. The MR provided the following work instructions:
"T/S system per vendor's manual.and repair as necessary. Any further instructions needed contact foreman."
(2) MR 84-3748 was written to address seat leakage on 14" feedwater
,
regulating valve FCV-530. The MR. provided the following work
' instructions:
1.
Check-to see that controller is functioning. properly and allowing valve to.close.
2.
If disassembly.is required, refer to "ENA" section of
';
Fisher Book.
3.
Maintain Class "C" cleanliness per MP 3-5.'
.I 4.
Check stroke of valve.
'
Without any additional work instructions, the mechanic dissembled the valve, replaced the valve internals and reassembled the valve.
(3).MR 85-3037 was written to address failure of the clutch to engage on feedwater pump discharge <stop check valve MO 2997B',
such that the Limitorque operator could not. operate the valve.
The MR provided the following instructions-
" Adjust clutch for: proper' operation of hand and motor drive of valve."
i
_-_-
__ ___ _ __ -_ __ _Y__
_.
.
-
.
.
Although the valve technical manual was not referenced, the inspector observed that the technical manual did not provide instructions for making this adjustment.
The inspector discussed the apparent lack of detail and specificity provided by the procedures and maintenance requests with the licensee. The licensee _was aware of this weakness, and stated that a contract had been issued to develop a new procedure writers' guide and to write two new procedures for each department. Upon completion of this effort, the scope of enhancing all of the procedures will be evaluated.
In discussing this issue further, the inspector learned that it has taken approximately a year to allot funding and bid the contract for this work. Given the consequences that could result from having inadequate procedures, it appeared that the licensee's actions have not been timely to resolve this concern.
In addition, it appeared that little attention had been focused on improving the quality of maintenance requests.
b.
Vendor Technical Manuals and Prints While attempting to review vendor technical manuals and vendor prints of the IA and Main Feedwater Systems, team members and plant engineers encountered difficulty in determining the vendor technical manual numbers and vendor print numbers.
In the case of manually operated valves and check valves, an index did not exist that could provide a cross-reference to the vendor technical manual numbers and vendor print numbers.
For other components, the Trojan Instrument Index, J-900 Numeric Sort, Revision 29, was used as R source of these numbers.
However, the vendor print numbers listed in the Instrument Index often contained inaccuracies, c.
Vendor Recommendations Review of vendor technical manuals indicated that the licensee does not rigorously implement the vendor recommendations. Several examples are as follows:
(1) Technical Manual M123-9, Instruction Manual Main Steam Isolation Stop and Main Steam Check Valves, November, 1969, stated on page 24 that an oiler, filter, water separator and pressure regulator should be provided in the air line to the MSIV air cylinder. This equipment was not provided, and the licensee had no documented evaluation or rationale for not complying with the vendor's recommendations.
(2) ASCO Bulletin 8316 stated that filters should be installed in-line with the solenoid valve to protect it from damage. As previously mentioned in this inspection report, the inspectors observed that filters were not installed in-line with the MSIV air solenoid valves, and the licensee had no documented evaluation or rationale for not complying with this
.)
vendor recommendation.
(3) Technical Manual M35-8, Instruments and i
Service Air Dryers and Filters, Revision 3, stated that plug valves-
'
(IA-14 & IA-015) should be lubricated once every week, and should be cleaned and relubricated with fresh grease once every year. The licensee did not accomplish this maintenance and had no documented evaluation or rationale for not complying with this vendor
.
l*
I
+
.
,
recommendation. This was identified as inspector fsilow-up item 87-31-20.
l d.
Plant Trending Interviews with licensee personnel indicated that trending of plant l
problems was being conducted informally by individuals and that no formal documented effective plant trending program to identify
'
significant trends had been established for the Trojan Plant. A i
review of plant maintenance requests indicated that MRs generally lacked specific details that are needed to trend plant problems.
The licensee stated that a formal program was currently being I
developed whereby trending of equipment problems would be accomplished by a preventive maintenance engineer, and additional attention would be focused in this area.
l l
e.
Industry and Trojan Experiences l
The inspectors focused on selected deficiencies that have been experienced on the main feedwater system and the air systems, either at Trojan or within the nuclear industry. The purpose of the review was to determine if the licensee had taken steps to preclude recurrence of these deficiencies at Trojan.
(1) Air Systems IE Notice 86-57, Operating Problems with Solenoid Operated Valves at Nuclear Power Plants, July 11, 1986, addressed the problem of hydrocarbon contaminants in air systems causing degradation of solenoid valve seats and seals made of ethylene and propyler:e materials. As noted in section 5 of this inspection report 0AR 86-64 reviewed the IE notice and noted no problems at Trojan with regard to hydrocarbon contaminants.
During this inspection, the inspectors observed that oil was present in the exhaust port s.1 sulenoid valve SV-1451, associated with one of the steam supply valves for the steam driven auxiliary feedwater pump.. The model number stamped on SV-1451 indicated that it contained valve seats and seals made i
of ethylene propylene material.
Follow-up Item 87-21.
In.eviewing the licensee's administrative controls for maintaining air system reliability, the following weaknesses were noted.
,
i Procedures did not exist which specified cleanliness standards, in process controls or cleanliness verifications for air system maintenance activities.
Procedures did not exist for verifying proper Operation of
,
air system oilers.
!
Procedures did not exist for specifying inspection or replacement requirements for in-line air filters.
i e
_--
- - - - - _ _ _ _ _ _ _
.
..
A specific listing did not exist of equipment which use air pressure regulators, oilers or in-line air filters to assist in preventive maintenance of these components.
The above noted procedural weaknesses will be followed-up in future inspections.
Follow-Up Item 87-31-22.
(2) Main Feedwater (FW) System A review of the FW systen was performed to determine if problems similar to those noted in NUREG 1190 existed at Trojan. NUREG 1190, Loss of Power and Water Hammer Event at San Onofre, Unit 1, on November 21, 1985, identified several instances where seat leakage associated with feedwater system valves exacerbated the water hammer event. The team noted the following procedural weaknesses.
MR 84-3748 identified a problem with seat leakage on feedwater regulating valve FCV-530. The maintenance request indicated that replacement of the valve internals was required. The licensee could not identify specific preventive maintenance requirements that would preclude recurrence of this condition.
MR 85-3037 identified that the clutch would not engage on feedwater pump discharge stop check valve MO-29978 such-that the Limitorque operator could not operate the. valve.
The licensee could not identify specific preventive ~.
maintenance requirements that would preclude recurrence of this condition.
'
I MR 85-5000 identified a problem with feedwater regulating l
valve diaphragm leakage. The licensee could not identify specific preventive maintenance requirements that would-ensure that diaphragm integrity was maintained.
The above noted preventative maintenance procedural weaknesses will be followed-up in future' inspections as part of follow-up item 87-31-22.
5.
Offsite Event Evaluation i
l The evaluation of industry events by the licensee is described in the i
licensee's Topical Report PGE-1044, " Operating Experience Review Program." These evaluations of industry events includes information obtained from NRC Information Notices (ins), Inspection and Enforcement Bulletins (IEBs), Institute for Nuclear Power Operations (INP0)
Significant Operating Experience Reports (SOERs) and INP0 Significant Event Reports (SERs). The licensee's program has the following major.
!
,
phases:
(1) input of operating experience information, (2) screening and i
'
evaluation of this information, and (3) output of recommended corrective
!
action.
In order to accomplish these phases, administrative controls are
)
established which include appropriate follow-up and verification, status l
j
.
~
.
tracking, and documentation control.
The licensee used a format called an Operational Assessment Review (OARS) to document their results.
The licensee's review of the following NRC ins associated with the IA and FW Systems were selected for inspection by the team.
IN 82-25 " Failures of Hiller Actuators upon Gradual loss of Air Pressure" IN 82-25 discussed problems experienced with Hiller actuators at the Grand Gulf Nuclear Station. During preoperational testing a large number l
of valves with this type actuator failed to go to their fail-safe position upon a gradual loss of system pressure.
The IN was reviewed by the licensee and the review documented in OAR 82-114. The licensee noted that there were no Hiller actuators at Trojan.
In a memo (CPY-815-82) dated October 19, 1982, the licensee stated that the air operated steam supply valves to the turbine driven Auxiliary Feedwater Pump and the MSIVs were valves that could be applicable to the IN. The licensee also recognized in reviewing this IN l
that a slow loss of air pressure test had never been performed at Trojan.
The licensee concluded that conduct of a test or other action was not warranted due to the design of the valves and their associated air systems.
!
During a recent test on these valves, the licensee discovered that the accumulators associated with the AFW valves did not hold the required air I
pressure for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> needed. The system arrangement for these valves either lead to leakage through the isolation check valves or through the
,
air operator. The licensee repaired the operators and replaced the check
!
valves. The system passed a second test, however the licensee did not i
performed a gradual loss of system air pressure test on these valves.
Paragraph 2.B of this report identifies an open item regarding the gradual loss of air test.
.
IN 85-35, " Failure of Air Check Valves to Seat" This information nctice provided information on Parker-Hannifin eir check valves used at Byron Unit I that lead to failures of the MSIVs or the l
Feedwater Isolation Valves. The event noted that leaking air check j
valves had allowed the accumulators to bleed down gradually, which in
]
turn did not allow the valves to close fully. The notice also mentioned
<
that preoperationt' testing at Byron did not require slow depressurization.
]
This notice was reviewed by the licensee and documented in OAR 85-65.
In their review, the licensee looked at air check valves for the Component'
Cooling Water surge tank, for the Pressurizer Power Operated Relief Valves (PORVs), and for the AFWP turbine. The licensee noted that the i
check valves for the AFWP and PORVs were not' vulnerable since they had soft seats. The licensee apparently did not consider conducting a gradual loss of air pressure test as a result of this IN.
l
j
_ - _ - _ _ -.
_ _ _ _ _ _ _ _ _ _
..
.
.
IN 85-84, " Inadequate Inservice Testing of Main Steam Isolation Valves" IN 85-84 discussed incorrect inservice testing of the MSIVs at H. B.
Robinson Unit 2 and Turkey Point Units 3 and 4.
The MSIVs were tested without isolating the non-safety-related air supply to the actuators.
The objective of the test was to verify operational readiness under
.
t conditions that reasonably duplicate the design basis conditions.
Further testing at Turkey Point demonstrated that MSIV closure was only l
l assured when IA was powering the actuator.
This IN was reviewed by the licensee and documented in 0AR 85-127. The licensee revised procedures (P0T 5-1 and POT 1-5) for the AFWP turbine steam supply valves and pressurizer PORVs to be sure that the valves are stroked with the normal air supply isolated.
In a memo dated March 27, 1987 (GJK-003-87), the licensee noted that Periodic Engineering Test (PET) 14-2 will be used to verify backup air accumulator supply integrity-for the steam supply valves and the PORVs in lieu of P0T 5-1.
The licensee's testing of the valves is performed with the IA isolated.
However, the testing performed did not provide for T/S 4.7.1.2.1 surveillance requirements as noted in paragraph 2.B of this report.
IN 86-50, " Inadequate Testing to Detect Failures of Safety-Related Pneumatic Components or Systems" IN 86-50 was a compilation of four infonnation notices that discussed a history of problems with the IA system.
The licensee determined that they were not vulnerable to the failure mechanisms mentioned, based on their previous reviews of the ins.
Although not specifically mentioned by the ins, the problem regarding gradual loss of IA system air pressure was again not addressed.
IN 86-57, " Operating Problems with Solenoid Operator Valves at Nuclear Power Plants" This notice provided information on problems in the industry with solenoid operated valves (S0Vs). This notice stated that the problems occurred mainly when the S0V served as a pilot valve. The causes for failures of the S0Vs were identified as:
(1) long term operation in a high temperature environment with the S0Vs energized, (2) Hydrocarbon contaminants in the IA system, (3) Chloride contaminants causing open circuits in the coils of the S0Vs, (4) An inadequate PM program for replacingelastomersandothershort-livedsubcomponentsand(5)
excessive lubrication.
This notice was reviewed by the licensee and documented in OAR 86-64.
The licensee stated that there are approximately 600-700 SOV's with approximately 200 of these 50V's used in the IA system. The licensee
)
documented some problems with these 50V's in the previous 3 years, but did not discern any trend. The licensee reviewed the problem of high temperature and did not note any problems in their high temperature applications, or in having hydrocarbon contaminants. Excessive lubrication was considered not to be a problem. Paragraph 4.e.(1)of this inspection report identified the presence of oil in the exhaust port
'
.,
..
.
of solenoid operated valve SV-1451.
For maintenance programs, it was suggested that maintenance procedure MP 12-4 be revised to provide improved instruction for inspections. The maintenance procedure was revised.
IN 87-28, " Air System problems at U. S. Light Water Reactors" This notice provide information of a report issued by the Office for Analysis and Evaluation of Operational Data (AE00), AE0D/C701 documenting i
and discussing the problems with IA systems in the industry.
This notice was being reviewed by the licensee at tnt time of the inspection and was not complete by the end of the inspection. This
,
l notice is being tracked and will be documented in 0AR 87-44.
IN 85-50, "Com)lete Loss of Main and Auxil_i,ary Feedwater at a PWR Designed by Ba) cock and Wilcox" This notice provided information on the Davis-Besse event, which was a loss of main and auxiliary feedwater. This significant event started after a Main Feecvater pump tripped on overspeed. The reactor tripped, and a spurious signal indicated low water level on steam generator 2. The Steam and Feedwater Rupture Control System (SFRCS) caused the MSIVs to i
close. The closure of the MSIVs isolated the main feedwater pump from its source of steam. Three additional problems were noted in the Information Notice as contributing to this event:
(1) An operator made an error on the SFRCS, and both steam generators were also isolated from both auxiliary feedwater pumps.
(2) After SFRCS was properly reset, the auxiliary feedwater isolation valves did not properly open.
(3) After the AFW isolation valves were manually opened, both auxiliary feedwater pumps accelerated, and tripped on overspeed.
This notice was reviewed by the licensee and documented in 0AR 85-080.
In the OAR, the licensee included the INPO report and the Westinghouse Owners Group (WOG) report on the Davis-Besse event. The licensee appeared to be following the WOG recommendations in this OAR. The WOG-identified 11 potentially generic issues from this event, which could potentially be applicable to Westinghouse plants. The operator error, which involved pressing one button on the SFRCS control system which isolated both steam generators, was deemed not to be possible at WOG plants. The motor operated valves which would not open is being addressed with IEB 85-03.
The overspeed trips on the AFWP turbine were deemed to have a minimal impact on reliability concerns. The WOG stated, in part, that the reliability of the AFWP turbines was not affected by IA problems.
The resetting of the overspeed trip mechanisms was not specifically addressed in the OAR.
In discussions with the licensee, the inspector was informed that INP0 SOER 86-1, as evaluated in 0AR 86-24, included a recommendation on training the operators on'the AFWP turbin '
,
.
..
The inspector concluded that the licensee does review industry information for applicability to their facility, however their actions based on these reviews could be more thorough. A specific example is that a number of ins indicated that a gradual loss of air test would be prudent to perform, however the licensee did not follow them on this potential deficiency.
6.
Onsite Event Evaluation The evaluation of onsite events by the licensee is accomplished by several organizations, including Quality Assurance (QA) and the Performance Monitoring / Event Analysis (PM/EA) group.
The onsite QA group is responsible for assessing the significance and tracking deficiencies in materials, parts or components, and deficiencies in plant activities. The deficiencies in equipment are noted in Non-Conformance Reports (NCRs) and deficiencies in activities are noted in Non-Conforming Activity Reports (NCARs).
The licensee also investigates abnormal events that happen at the Trojan site, and documents them in Event Reports (ERs). These event reports are evaluated by the Plant Review Board for deportability as Licensee Event Reports (LERs).
For approximately 1 year, since the PM/EA group was implemented, the PM/EA group has also been involved in the assessment of these events.
LER Evaluations The following licensee evaluations associated with LER reports on the feedwater system were reviewed by the inspector.
LER ER NUMBER NUMBER TITLE 84-04 84-018 Reactor Trip on Steam Generator Low-Low Level due to Main Feedwater Pump Trip 87-01 C-01 Technician Inadvertently Shorted Feedwater Pump Control Power - Feedwater Pump Trip / Reactor Trip Both of these LERs involve grounding of electrical circuits associated with the main feedwater pumps while performing maintenance. While both assessments addressed the root cause of the event, the difference in the reviews of these feedwater pump trips is substantial.
In the first review, one cause was noted.
In the second review, one cause and several contributing factors were also noted, and a discussion of the event was well documented. The corrective action taken for ER 87-01 appeared to be,
extensive.
The PM/EA group also noticed that this event report had some problems similar in nature to ER 86-131, and some of the conclusions and recommendations of a observation report on I&C activities (PM/EA-0R-002-87) from this previous ER were deemed applicable.
I
I
.
.h
. ;
g
.
Event Reports87-084, 87-091 Event reports ER 87-084 and ER 87-091 deal with an accumulator line break that happened when water was being transferred from the "A" to the "D" accumulator via.a 1 inch "A" accumulator fill line. The line failed at the accumulator nozzle. Following repair of the line, the "A" accumulator was successfully hydrostatically tested. The lineup was tried again, and loud noises were heard. The transfer was immediately.
stopped. The' valve lineup.was checked and transfer restarted, when loud noises were heard again. Once more, the valve lineup was rechecked and operation' restarted. At.this time. the fill line failed for.a second
. time at the nozzle to pipe weld. The banging noises were due to backflow through a Kerotest valve.
The root cause appraisal included several analysis methods. The root causes were attributed to failures in the areas of procedures, communications, and the management system. The specific root causes attributed'were: (1) The Operations staff violated procedure 01-5-2 on transferring water through the accumulator fill line, (2) Inadequate
- ccmunications within the operations department on previous problems with the fill line and inadequate communication between the Engineering and Operations departments with the known problems on this type of valve, 3)
(
the management system use for the OAR review lacked the required depth to evaluate the recommendations of 0AR 85-76 which discussed operational problems with Kerotest valves.
The report to the Plant Review Ecard (PRB) on this event (GJS-024-87)-
contained eight recommendations for improvement.
In the PRB meeting
.m nu es (9/2/87), the PRB agreed with the recommendations except for a i t recommendation on the OAR program. Since problems with the Kerotest valves have been known in the: industry, and had been documented in.the 0AR program (i.e. 0AR 83-14 and 0AR 85-76) the PM/EA group recommended that a review be undertaken of previous.0ARs.for potential problems. The PRB.did not agree because of the magnitude'of the effort required and the difficulty in' knowing what to look for..'The PRB did agree that all future OAR's should include a review of previous related OAR's.
Event Report 87-139
!
Event Report 87-139.noted a problem with the "C" Main Feedwater
'1 Regulating Valve (FWRV). The root cause was attributed to the local-handwheel being out of position. This event was reported to the NRC in i
!
LER 87-23. Approximately 36 ' hours later,' another reactor trip happened due to control problems with the same FWRV (ER 87-140 and LER 87-24).
The event report memorandum to the PRB chairman (JJT-020-87) had three corrective action recommendations. These were (1) a-pre-evolution briefing with personnel assigned SG 1evel control duties, (2) FWRV control problems should be investigated and resolved and (3) additions to Operating Instruction 01-8-1 to verify handwheel position.
l
,
T
.
..
.
.
Conclusion Based on the recent event reports reviewed, and discussions with the licensees PM/EA group, the depth of review of recent event reports appears adequate. The recommendations made in these event reports appear appropriate to address the concerns and the root cause of the events.
7.
Quality Control A previous NRC Region V team inspN tion of Trojan (50-344/85-33)
identified a violation (civil penalty) regarding lack of Quality Control inspection of Maintenance activities. The team followed up on the licensee's response to this experience.
In the licensees' response to the violation, the licensee agreed with the violation and initiated short-term and long-term corrective actions. Furthermore, the licensee acknowledged that independent inspections of electrical maintenance and modification activities had not been implemented at a level required by the QA program.
An inspection was performed to determine the progress of the licensee's implementation of its corrective action. The process of writing maintenance requests (MR) by the maintenance department was examined.
The maintenance department consists of the electrical, mechanical and Instrumentation and Control (I&C) disciplines.
Generation of MRs are provided for by the following procedures:
Procedure Rev.
Date Issued Ti tle A0-3-9
June 30, 1987 Maintenance Requests AP-3-1
Aug. 27, 1987 Procedure for Processing Maintenance Requests Administrative Order A0-3-9 is the site wide procedure for the use of maintenance requests. This Order notes the site wide requirements for initiating, preparing, performing, and documenting work and post-maintenance testing on maintenance requests. Some of the specific
<
requirements include:
'
All maintenance to quality related equipment requires an MR.
The work group supervisor reviews the MR for its effect on plant I
reliability and safety, determines the need for QA/QC involvement, l
and provides for installation checks.
'
QC inspection points are to be identified with an I in the left hand margin of instructions.
The Administrative Procedure AP-3-1 is a lower tier procedure to provide I
more specific instructions in processing MRs for the maintenance department.
Some of these more specific instructions include:
'
All corrective maintenance affecting reactor trip, ESF trip and actuated equipment from the SSPS including all actuated devices and
____ -_____-_____-____ _____ __ _
-
_
I
'
.
equipment shall have either a QC inspection hold point, a QC in
,
process observation, or a peer second verification.
l The assembled work package is routed to obtain the work supervisor's
'
approval for the work instruction; for QC review when the J
maintenance activity requires a QC inspection, and for radiation protection review when the package requires ALARA planning.
j The inspector observed that the QC organization has expanded and the observation of work activities has improved for mechanical and electrical i
maintenance. During the refueling outage in the second quarter of 1987, the amount of QC inspections increased significantly in these two maintenance disciplines.
For the work in the mechanical and electrical maintenance disciplines, QC observation and hold points were noted.
However, the inspector questioned the level of QC observation of I&C activities as follows:
'( A) The inspector noted that a person knowledgeable and experienced in nuclear I&C activities still does not exist in the QC group.
(B)
For work observed in the Instrumentation and Control (I&C)
discipline, inspection hold points were not observed to have been added to the maintenane requests reviewed by the inspector.
(C) The inspector was informed that 180 does not send thek maintenance
,
requests to QC for insertion of hold points or observation points.
i
'
Observations 6ctivities of I&C work are being selected by QC by reviewing MRs for that day in the I&C shop, and choosing the maintenance to observe.
In a graph provided by the licensee, the number of observations of I&C
>
work declined from April 1986 to March 1987. There were no QC observations on I&C work during the second qucrter of 1987 (April, May, June) when the refueling outage was in progress. The licensee apparently recognized this decrease and ensured that QC observation activities were performed during the third quarter.
The recent QC observation reports for the third quarter were examined.
The period from July 1,1987 through September 22, 1987 (Observation Reports 1249 through 1332) was included. The following QC reports were noted to have been made in the I&C area. A total of 57.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> was expended by QC on I&C work observations.
QC Rep.
MR Number Date Requested By Hours 1320 87-5737 9/10/87 Quality Control 6.0 1318 87-5658 9/8/87 Quality Control 3.0 1317 87-5603 9/4/87 Quality Control 2.5 1316 87-5603 9/3/87 Quality Control 1.0 1315 87-5503 8/31/87 Quality Control 4.0 1309 I&C 5 8/18/87 Quality Control 4.0 1308 I&C 5 8/17/87 Quality Control 4.0 1307 I&C 5 8/14/87 Quality Control 1.5
_
-
.
-
.
--
.
--
_ _ _ _
_ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
..
f-
'
e 1303 I&C 4 8/13/87 Workgroup I&C 3.5 1302 I&C 4 8/12/87 Workgroup I&C 3.0 1294 I&C 5 8/6/87 Workgroup I&C 4.0 1271 87-4608 7/21/87 Quality Control 4.0 1259 Several 7/9/87 Workgroup(??)
14.0 1251 87-4136 7/1/87 Workgroup (??)
3.0 The inspector concluded that, although the licensee has been performing QC observations of I&C work, the lack of specific inspection points in maintenance requests, and the lack of QC personnel experienced in nuclear I&C activities, are significant weaknesses. The licensee's efforts to improve in this area will be followed up in a future inspection report (50-344/87-31-22).
8.
Plant Operations As a part of this inspection, the team also reviewed activities affecting overall plant operations. The following types of activities were observed:
Control Manipulations System Alignments Conformance with Technical Specification Limiting Conditions for Operation Control room and shift supervisor log entries Implementation of temporary modifications l
Implementation of clearance and tagging Auxiliary Operator routines l
Plant Housekeeping l
The inspector found that formal procedures or guidelines had been established for control of the observed activities. The inspector did, however, note in particular during a review of normal and abnormal l
operating precedures and alarm response guidelines that to a large extent l
these procedures lack specificity. A similar concern was documented by l
INP0 during their 1985 audit of the facility. The licensee informed the inspector that efforts were indeed now underway to correct these deficiencies and that a contract had been awarded to an outside consultant with work scheduled to begin within the next few weeks. The l
j scope of this contract included the creation of a new procedure writer's l
guide and revision of selected procedures with planned revisions to all I
administrative, operating and maintenance procedures occurring within the next two to three years.
The inspector also observed during this inspection that improvements in the utilization of present procedures by the operating crews could be made if only by the operators more frequently referring to the
,
established operating instructions during performance of routine tasks.
l By doing so, it would appear that some reduction in the number of
personnel errors, which have caused the licensee problems in the past,
!
could be realized.
In the area of plant housekeeping, some problems were noted and brought to the attention to the licensee.
In particular, a number of loose
-
.
[
'
.
objects and materials were found scattered about the main steam support structure.
It was pointed out to the licensee that the identification and correction of such conditions should be one of the primary purposes of routine tours by auxiliary operators and that more attention to detail in this area should be given.
9.
Miscellaneous Observations During the inspection, other examples of less than rigorous engineering i
efforts and failures to learn from previous experiences of both the Trojan facility and others in the industry were noted.
A.
During reactor cooldown, very loud " banging" noises were observed to be coming from one of the MSIVs. The engineers indicated that this was not of concern because the valve was operating in its designed manner.
Previous experiences have shown that slight vibrations in the valve occur and have been determined to be of no concern based on telephone conversations with outside consulting firms. NUREG 1190, " Loss of Power and Water Event at San Onofre, Unit 1, on November 21, 1985." paragraph 6.6, discusses a situation where a banging noise in the feedwater system was not properly resolved. A more rigorous handling of this condition could have prevented the water hammer event.
It is the inspection team's belief that this heavy banging should be addressed by a more detailed analyses for operational, material, and structural degradation.
B.
While operating at reduced reactnr power levels, some feedwater system piping may have been subjected to turbulent flows resulting from throttled valves. No erosion / corrosion analyses have been performed for these operating conditions because the licensee considered that the reduced power conditions are conservatively l
'
bounded by the analysis for full power operations. Also, these reduced operating conditions are considered by the licensee engineers to be temporary short term plant conditions, and therefore
,
no further analyses are deemed necessary. The inspector stated that i
a more rigorous technical evaluation should be completed to address this concern, and referred the licensee to previous operating experience at San Onofre Nuclear Generating Station where substantial wall thinning occurred in the feed and condensate system during reduced power operation. Considering past plant experiences with erosion / corrosion, the engineering staff should be more
>
rigorous in evaluating abnormal operating conditions.
i
'
C.
Several instances of failures of control room indication and system control have been experienced by Trojan in the past due to high temperature conditions during control room ventilation testing.
In 1985, irregular steam generator control was observed due to overheating of LC-0519 level controller. A failure in feedwater
.
'
flow indication was noted in 1986 due to overheating of a square root extractor during control room ventilation testing. Other instances of operational concerns due to high control room temperatures were also noted in systems other than the and a Detailed Construction Package (DCP-11)gn Change (RDC-86-045 feedwater/ steam systems. A Request for Desi have been generated to
._ _ _-______ _ __ - -__ - - _ -
.
-
-
-
-
..
Y 26-o p
.o
!
stop this. overheating' problem by installing.two cooling units-in the control room.
This effort is. scheduled for completion in 1988.
Due to the vulnerability of control room components due to abnormally high temperatures, this design should be implemented'as soon as possible.
D.
Various sections of this report identify. weakness in functions.that~
should have been provided for by the system engineers.
The team recognized that the PGE implementation of a system engineer program.
was a step taken for performance improvement.
The' team also recognized that the system engineers appeared to be as dedicated and knowledgeable as possible.
However, the team noted that system
_
engineers were normally assigned four or five systems.
Consideration should be given to a smaller number of systems assigned to each' system engineer to preclude minimizing the benefits
~
gained from having system engineers..
10.
Unresolved Item Unresolved items are matters about which more information is required to
.
determine whether they are acceptable or may involve. violations or:
{
deviations.
Two new unresolved items were identified during this inspection and are discussed in paragraphs 2.A(2) and 2.B.
i 11.
Exit Interview The inspection scope and findings were summarized on'0ctober 2,1987,'
with those persons indicated in paragraph 1 above.
The inspectors described the areas inspected and discussed in detail the inspection.
i findings.
The following new items were identified during this inspection and discussed with the licensee on December 17, 1987.
'
Report Item Paragraph Number Category Description 2.A (2)
87-31-01 Deviation IA System UFSAR! Commitments-
!
2.A.(4)
87-31-02 Follow-up
' Incomplete Design' Change Package.for i
Temporary Air Dryers'
'
'
2.B.
87-31-03 Follow-up Gradual Loss of Air Test '.
.
.
,
2.B.
87-31-04 Unresolved Adequacy of AFW Steam Admission _ Valve L'
2.B.
87-31-05 ' Violation Failure to Test AFW pump Terry Testing turbine using accumulators only
' 2. B.
87-31-06 Violation L
Failure to Perform IST on Various Air Valves.
2.B.
87.31-07 Follow-up Integral Yesting of IA Instrumentation Circuits-2.C.
87-31-08 Violation Failure to' Follow Drawings,....
Installation'of. Pipe Support H834 2.C.
87-31-09 Follow-up:
Instrument Index'Setpoints 2.0.
87-31-10 Follow-up Controls for Weekly IA Blowdowns1 2. 0.
87-31-11 Follow-up Controls for By-Passing IA Dryers:and Filters
,
-
- - - _ _ - _ - _ _ _ _ - -. - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
___
___ ___
. -.
O
p e
3.A.
87-31-12 Violation Failure to Perform 10 CFR 50.59 Evaluation for Reduced EDG Air Pressure Setting 3.A.
87-31-13 Violation Failure to Follow Drawing, Installation of EDG Air Lubricator 3.A.
87-31-14 Follow-up Dew Point Measuring of EDG Air Start Air
!
3.B.
87-31-15 Violation Failure to Test EDG Air Start System
,
Air Receiver Relief Valves 3.B.
87-31-16 Follow-up EDG Air Start Lubricator Setting 3.B.
87-31-17 Follow-up Controls for Use of Teflon Tape 3.C.
87-31-18 Follow-up EDG Air Compressor Diesel Motor Belt Changeover 3.C.
87-31-19 Follow-up EDG Air Start System Blowdown Controls 4.C.
87-31-20 Follow-up Vendor Manual Recommendations 4.E.(1 87-31-21 Follow-up 011 on Valves w/EPR Material 4.E.(1 87-31-22 Follow-up Maintenance Procedure Weaknesses and (2
87-31-23 Follow-up QC involvement in maintenance activities l
)
l
l
4 j
