ML20212L042
| ML20212L042 | |
| Person / Time | |
|---|---|
| Site: | Nine Mile Point  |
| Issue date: | 01/22/1987 |
| From: | Collins S, Kane W NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20212L002 | List: |
| References | |
| 50-220-86-17, 50-410-86-61, NUDOCS 8701290294 | |
| Download: ML20212L042 (50) | |
See also: IR 05000220/1986017
Text
.
.
.
U. S. NUCLEAR REGULATORY COMMISSION
REGION I
Report Nos. 50-220/86-17
50-410/86-61
Docket Nos. 50-220
50-410
License Nos. DPR-63
CPPR-112
Licensee: Niagara Mohawk Power Corporation
301 Plainfield Road
Syracuse, NY 13212
Facility: Nine Point Units 1 & 2
-Location: Scriba, New York
Dates: August 25 - 29, 1986
Inspectors: E. Kelly, Senior Resident Inspector, Limerick
S. Kucharski, Resident Inspector, Limerick
C. Marschall, Resident Inspector, NMP-1
G. Napuda, Lead Reactor Engineer, DRS
R. Paolino, Lead Reactor Engineer, DRS
W. Raymond, Senior Resident Inspector, VY
Investigator: R. Matakas, Office of Investigations, RI
Team Leader: MM
'S. Collins, Deputy Director, Division of
IfM/87
' ~Date
Reactor Projects
Approved by: // I2/E7
W. Kane, DirFctor Dite
Division of Reactor Projects
.
6
8701290294 870122
PDR ADOCK 05000220
O PDR
-. .--
. .
-
.
.
.
CONTENTS
Page
1. EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . 1
2. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . 3
3. PURPOSE . . . . . . . . . . . . . . . . . . . . . . . . 7
4. REVIEW 0F ALLEGATIONS . . . . . . . . . . . . . . . . . 8
4.1 CRD Pump Vibration . . . . ............ 8
4.2 Helium Leak Testing _. . . . . . . . . . . . . . . 9
4.3 Feedwater Check Valve - Leak Test . ....... 10
4.4 Feedwater Check Valve - Flow Diversion . . . . . 12
4.5 LPRMs . . . . . . . . .............. 16
4.6 QC Involvement . . . . . . . . . . . . . . . . . 19
4.7 Harassment . . . . . ..............20
4.8 IRMs . . . . . . . . . . . . . . . . . . . . . 20
4.9 Unplanned Exposure . . . . . . . . . . . . . . . 22
4.10 Tool in Reactor Vessel . . . . . . . . . . . . . 22
5. QA PROGRAM REVIEW . . . . . . . . . . . . . . . . . . 23
6. REVIEW 0F NMPC INVESTIGATION . . . . . . . . . . . . . 27
7. CRD MAINTENANCE REVIEW . . . . . . . . . . . . . . . . 29
7.1 Event . . . . . . . . .............. 29
7.2 Review and Findings . . . . . . . . . . . . . . . 32
7.3 Conclusions . . . . . . . . . . . . . . . . . . . 44
I
8. PROGRAMMATIC ISSUES . . . . . . . . . . . . . . . . . 46
9. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . 47
Attachments:
1. Letter W. F. Kane to C. V. Mangan, dated August 11, 1986
2. Letter J. C. Linville, Jr. to Alleger, dated August 18, 1986
3. Cte.bined Inspection Nos. 50-220/86-16; 50-410/86-46 dated September
30, 1986; subject - Radiological Controls
4. Inspection Report No. 50-220/86-13 dated December 18, 1986;
subject - Generic Letter 83-28 Followup Actions
5. Inspection Report No. 50-410/86-52, dated November 19, 1986,
subject - QA Program Effectiveness and Quality First Program (Q1P)
, . _ . - - . .- - _ -
9
.
.
.
1. EXECUTIVE SUMMARY
Background
On July 11, 1986, while observing maintenance on local power range monitor
(LPRM) connectors at Nine Mile Point Unit 1 (NMP-1), the NRC resident
inspector received allegations concerning the connector qualification and
installation techniques from an instrument and control (I&C) technician.
The technician subsequently met with Niagara Mohawk Power Corporation
(NMPC) representatives to convey his concerns. On July 22 the technician
came to the NRC Region I office to discuss his concerns and to provide a
sworn statement, which was transcribed. On August 11 an NRC letter was
sent to NMPC enclosing a summary of the I&C technician's allegations. The
letter acknowledged the ongoing NMPC investigation into the concerns and
requested a written report of the results (See Attachment 1). The tech-
nician was subsequently notified of the NRC actions and advised to contact
the Department of Labor (DOL) to address the concern alleging harassment
by his supervision (See Attachment 2).
In an August 15 letter NMFC outlined its' approach to investigation of
the allegations and provided a summary report of the investigations and
associated conclusions. NMPC. concluded that no activities were found
which would jeopardize the safe operation of the station. A meeting was
subsequently held on Augast 18 to discuss the findings, and in an August
31 letter NMPC set forth the investigation findings relative to the
allegations, including the evaluation methodology, short and long te m
remedial actions, and means to measure the effectiveness of those actions.
Purpose
The primary purpose of the inspection was to assess the impact of the
allegations on the safe operation of NMP-1. Further, the inspection
assessed the effect of the allegations on the pending Region I recommend-
ation on the licensing of NMP-2. To achieve these purposes the inspection
reviewed the validity of some parts of the allegations, reviewed the NMPC
evaluation of the allegations, and assessed the technical significance of
the allegations. Due to limited inspection resources the inspection was
not intended to establish the validity of all aspects of the extensive
'
allegations. Since the technician had presented the allegations to NMPC
on July 15 and the August 11 NRC letter to NMPC had summarized the allega-
tions to ensure that NMPC had received all the allegations, NMPC was
responsible for the review of the complete scope of the allegations.
Inspection
For each allegation the inspection reviewed the allegation, determined
the basic concern, and focused on the root cause of the technical issue
from the perception of the NRC to assess the impact on Unit I and 2
programs. Plant hardware was inspected, independent reviews were
performed, and extensive discussions and interviews were held with NMPC
.
.
-
2
.
.
personnel. The bases for any conclusions were derived from inspection of
the actual item or area, reviews of NMPC records, previous NRC inspection
findings, established engineering knowledge and practice, and quality
assurance records.
Further, the inspection reviewed portions of the ongoing NMPC investiga-
tion of the allegations to assess its effectiveness. Also, an evaluation
of quality assurance programs at Unit I and Unit 2 was performed to
evaluate the ability of these programs to identify and correct the problems
associated with the allegations. Finally, a review of the Unit I forced
outage on August 22 was performed concurrently with the NMPC evaluation to
assess the effectiveness of NMPC's root cause determination and corrective
actions.
Summary and Conclusions
Most of the allegations were found to be factually correct, however their
safety implications were subsequently determined to be minor. The
inspection did not find any significant adverse safety impact from the
allegations on the operation of Unit 1. Hcwever, the quality assurance
program was not as effective as it shoulo be in reviewing operations and
that it is unlikely to identify the type of problems associated with the
allegations. The NMPC investigation was judged to be independent of
adverse management influence and, given the time to complete its inves-
tigation, capable of establishing the facts surrounding the allegations.
Our reviews also indicated that there were programmatic weaknesses evident
in the NMPC management system that allowed these issues to develop and in
some instances spread. This conclusion was based on the results of the
NMPC evaluation of the contributing factors to the allegations and on the
NRC review of the NMP,C evaluation of the August 22 forced shutdown due to
CRD valve maintenance.
This report and related inspections (Attachments 3, 4 and 5) contain
apparent violations which will be the subject of a future enforcement
conference.
.
.
-
.
3
9
.
2. BACKGROUND
A chronological listing of the allegation history is presented in
Table 1.
On July 11, 1986, while conducting an inspection of work in progress
under the reactor vessel at Nine Mile Point Unit 1 (NMP-1), the resident
inspector was approached by an instrument and cor. trol (I&C) technician,
who indicated that there were problems with the local power range monitor
(LPRM) connector replacement work. The inspector verified the validity of
the technician's concerns and notified Niagara Mohawk Power Corporation
(NMPC) management of his findings. The connectors were subsequently
repaired. On July 14 the I&C technician (alleger) again contacted the
inspector with other allegations concerning operations, surveillance,
maintenance, quality program activities, and harassment by his supervisor
and peers. The inspector encouraged the alleger to bring his concerns to
NMPC management. On July 15 he met with the Vice President - Nuclear
Generation, and an investigation was initiated by NMPC.
NRC Region I conducted an allegation panel meeting on July 15, regarding
this allegation (RI-86-A-0080). The alleger and NMPC management were
contacted, and arrangements were made to provide for the I&C technician to
express his concerns directly to the NRC. On July 22 an investigative
interview was conducted at the Region I Office in King of Prussia, PA, and
>
two hundred fif teen pages of testimony were transcribed.
On July 31 a copy of the transcript and the NRC summary of the concerns
were provided to the alleger for comment or correction. He agreed that
the summary of allegations was an accurate listing of his concerns.
Subsequently, on August 11 the summary of allegations was provided to NMPC
(Attachment 1) and formally transmitted to the alleger on August 18
(Attachment 2).
During the period of August 4-7, a radiological controls inspection was
conducted onsite, which incorporated the NRC review of the alleger's
concern in this area (Attachment 3).
In an August 15 letter NMPC provided its approach to the investigation of
the allegations and a summary report of its investigations and associated
conclusions, which indicated that no activities were found which would
jeopardize the safe operation of Unit 1. A meeting was subsequently held
in Region I on August 18 to discuss these findings, and in an August 31
letter NMPC updated the report by submitting revised pages.
On August 22 NMP-1 was shut down as a result of maintenance performed on
the control rod drive (CRD) scram valves.
.
, e m __ m
a
.
-
4
.
. .
During the period of August 25-29, the NRC special inspection was conduc-
ted to review the safety and programmatic implications of the allegations
on the continued operation of Unit 1 and the potential licensing and
operation of Unit 2. During the inspection NMPC made a presentation on
their actions to date to identify and resolve the issues resulting from
their internal investigation.
In an August 31 letter NMPC set forth their conclusions relative to the
allegations and the corrective actions needed, including the background,
the methodology used to evaluate the allegations, the short and long term
remedial actions, and the means to measure the effectiveness of the
corrective actions.
.
_
.
.
5
.
.
Table 1
l
1986 CHRONOLOGY OF NMP-1 ALLEGATIONS
January & -
CRD Pump repaired twice.
February
March -
Allegation: vibration testing of CRD pump stopped to
avoid reading which would require plant shutdown.
March 8 -
Plant shutdown for scheduled refueling outage.
March & April - Allegation: inadequate helium leak test of stack gas
monitoring system.
April -
Allegation: inadequate leak test of feedwater check
valve.
May -
Alleger went under vessel first time to perform LPRM
connector replacements.
Next Day -
Allegation: alleger assigned to other work based on his
LPRM complaints. Alleger went to QC and described the
problem.
June 6 -
Allegation: conversation with supervisor about being
denied the weekend off.
June 16 -
Reactor startup begun. Startup stopped when three
Intermediate Range Monitors (IRMs) would not respond.
June 17 -
Allegation: alleger and work crew replaced connector on
IRM 18 and cleaned connectors on IRMs 13 and 16 without
-
proper paperwork.
-
. Allegation: reactor restart begun using questionable test
records.
July-7 -
Plant shutdown for repair of CRD pumps, LPRMs, and
July 10 -
Allegation: alleger replaced two LPRM connectors using
unapproved connectors which he had purchased.
July 11 -
While the alleger was replacing LPRM connectors with
unapproved connectors, the NRC resident inspector entered
under the reactor vessel to review work. Alleger
identified connector problems to NRC resident inspector.
-
Licensee management notified of issue by NRC; connectors
repaired.
.
-
6
.
.
July 14 -
Alleger contacted NRC resident inspector with other
concerns. NRC resident inspector forwarded issues to
Region I and encouraged alleger to notify NMPC management
of his concerns.
July 15 -
Alleger met with Vice President - Nuclear Generation; NMPC
commenced investigation.
-
NRC Region I Allegation Panel met to document events
concerning allegation 86-A-0080.
July 22 -
Alleger gave sworn testimony in Region I.
July 31 -
Statement transcript provided to alleger.
August 4-7 -
Radiological Controls inspection onsite reviewed alleger's
concern (Attachment 3).
August 11 -
Licensee provided summary of allegations. NRC Region I
requested meeting to discuss NMPC investigation.
August 15 -
NMPC summary report of investigation sent.
August 18 -
Allegation receipt letter provided to alleger with
referral to the Department of Labor (DOL) for resolution
of the supervisory harassment allegation.
-
NMPC/NRC meeting at Region I to discuss NMPC
investigation.
.
,
August 22 -
Unit I shutdown resultin'g from CRD scram valves.
August 25-29 -
Special NRC team inspection conducted at Unit 1.
August 31 -
NMPC letter sent to NRC on allegation findings and
corrective actions.
September 8-12 - Special NRC team inspection conducted at Unit 2
(Attachment 5).
September 10-12 & -
Special NRC team inspection conducted at Unit 1
15-19 (Attachment 4).
.
.
.
-
7
9
.
3. PURPOSE
'
An interdisciplinary.special inspection team was selected to perform the
inspection. The disciplines of the team members correlated directly with
the technical concerns identified in the allegations. Disciplines
represented on the team were operations, surveillance, local leak rate
testing /in-service testing, quality control / quality assurance, investiga-
tions, and instrumentation and control. The team leader was a senior
manager (Deputy Director Division of Reactor Projects) with past Branch
Chief and Section Chief responsibilities for NMP-1 and NMP-2, as well as
. Senior Resident Inspector experience. The Division Director of Reactor
- '
Projects was present during August 28-29 and the exit meeting.
The inspection focused on assessing the impact of the allegations upon the
continued operation of Unit I and the potential licensing and operation of
Unit 2. This was accomplished by reviewing selected parts of the
allegations for validity, assessing the NMPC evaluation of the allega-
tions, and assessing the NMPC determination of the impact of the allega-
tions on site programs.
!
Other aspects associated with the allegations were pursued. To determine
<
the thoroughness and accuracy of the NMPC investigation, an NRC investiga-
tor reviewed the work that the NMPC investigators had completed and
assessed the independence that they had in pursuing the allegations.
Also, a review of the operational quality assurance programs at Unit 1 and
Unit 2 was performed to evaluate their ability to identify and correct
problems within the plant staff's activities. Additionally, the inspec-
tion reviewed an NMPC evaluation, which was proceeding concurrently with
the inspection, of an August 22, 1986 forced shutdown of Unit 1 due to
improper CRD maintenance to assess NMPC's ability to effectively evaluate
identified problems and establish corrective actions.
.
%
?
k
l
.
.. . - . - , . - . - -. .-. . .
.
.
,
8
.
.
4. REVIEW 0F ALLEGATIONS
This section is structured such that the review of each allegation
provides the allegation (s) as written in the summary of allegations sent
to the alleger and NMPC, a synopsis of the NMPC evaluation provided by the
August 15, 1986 NMPC letter to the NRC, details and findings resulting from
the inspection, an assessment of the technical significance and impact on
-nuclear safety, and conclusions based on the facts. The allegation
conclusions are presented according to whether they were substantiated
true as stated or unsubstantiated, including whether any part or all of
the allegation was not borne out by the facts.
4.1 CRD Pump Vibration
Allegation:
"1. In March,1986 after weeks of daily vibration tests of the CRD pump,
testing was suspended when it was apparent that the increasing
vibration would exceed the action limit of the ASME requirements and
a plant shutdown would have been required prior to the scheduled
March 8, 1986 shutdown."
NMPC Evaluation:
NMPC concluded that ASME Code and Technical Specification requirements
were met and that vibration testing and corrective maintenance of the pump
were appropriate. Specifically, CRD Pump No'. 11 experienced vibration
problems in January and was repaired (replaced bearings, bushings, and
balancing disk) between January 17 and 21. The pump continued to have
problems and was repaired again (replaced thrust bearings, installed new
bearing oiler, and checked coupling) between February 8 and 10. After the
repair the baseline data were revised on February 11 to reflect the
repaired pump's condition. The vibration readings referred to were for
information and troubleshooting purposes only, and when the pump was
operating acceptably, the readings were ceased.
Review and Findings:
The inspector reviewed the NMPC evaluation of the allegation and found it
to be accurate based on followup interviews with personnel in In-Service
Test (IST), I&C, Maintenance, and Operations. NMPC did suspend its daily
data collection of the vibration of CRD Pump No.11 after February 11, but
only after the pump was rebuilt and further testing showed the pump's
vibrational data to be within the acceptable range. The inspector did
note, based on his investigation, that the IST program met the require-
ments prescribed by Section XI of the ASME Boiler and Pressure Vessel Code
and the Technical Specifications, but the information accumulated by
the group was not being utilized by the various site organizations. It
was also noted that data taken by I&C at the request of other groups were
I
,
, , , - ,- - - .-- -,m,.
.
-
9
.
.
not supplied to the IST group. This is a potential problem in that the
IST records may not have a complete and accurate history of component per-
formance.
Technical Significance:
There are two CRD pumps, and the Technical Specification requires that if
one CRD pump is inoperable, it be made operable within 7 days or the
reactor must be shut down. During this period the second CRD pump was
operable, and CRD Pump No. 11 was never out of service for 7 days.
Further, as the allegation was unsubstantiated, there was no technical
significance.
Conclusion:
Based on this investigation the allegation was found to be unsubstanti-
ated. The licensee's summiry report presented to the NRC was confirmed.
4.2 Helium Leak Testing
Allegation:
"2. In March, 1986, the chemistry supervisor noted that errors existed in
the procedure for helium leak testing the stack gas system, in that
portions of the system would not be tested. The alleger found the
supervisor's conclusion to be correct. The I&C supervisor assigned
the alleger to review the leak testing procedure and propose changes
to it. After completing this work, the I&C supervisor sat on the
proposed changes and later told the alleger to do the testing with
the old procedure because there was no time to change the procedure
prior to startup. The leak testing was done on April 1."
NMPC Evaluation:
NMPC concluded that the stack gas monitoring system was tested using a
procedure with known deficiencies. The I&C Supervisor was aware of the
leak test procedure deficiencies but elected to perform the testing,
because modifications would have been required to the piping system to
properly test it. It was unlikely that the modifications could have been
done prior to completing the outage.
Review and Findings:
The technical concerns raised by the allegation were determined to be
valid. The procedure used to conduct leak rate testing of the stack gas
monitoring system was known to be inadequate, and the leak rate testing as
accomplished was not a test of the entire system.
.
-
10-
.
.
Discussions with the I&C Supervisor indicated that he intended to complete
leak rate testing of the entire system after modifications necessary for
the revised leak rate test procedure we e complete.
The I&C supervisor appeared to have had an excessive workload, resulting
in his inability to complete the necessary modifications prior to the end
of the outage. However, the I&C supervisor took no action to make his
superior aware of his excessive workload. Despite that, however, the I&C
Supervisor's management was aware of the I&C Supervisor's excessive
workload, nevertheless effective action to alleviate it was not taken.
There was no Quality Assurance / Quality Control involvement in the leak
rate testing of the stack gas monitoring system.
Technical Significance:
No specific safety concern exists, in that there are no Technical Specifi-
cations or regulatory commitments associated with the surveillance, and a
weekly bubble test of system components was performed.
Conclusions:
The allegation was substantiated.
Workload management has been addressed on the Instrument and Control
technician level. A repeat problem resulting from excessive workload is
unlikely to occur to I&C technicians due to subsequent staffing initia-
tives. However, workload management has not been addressed on the super-
intendent/ supervisor level. Problems resulting from excessive supervisor
workload could occur again at both units.
The issue of testing the stack gas monitoring system utilizing a procedure -
with known deficiencies is a further example of procedural deficiencies
and inappropriate procedural adherence attitudes which have been tolerated
by NMPC supervision.
4.3 Feedwater Check Valve - Leak Test
Allegation:
"3. The alleger was instructed to apply 100 psi air to seat the feedwater
check valve after it had failed its initial test. It failed the
second test also. Then the mechanic installing the replacement valve
told the alleger the valve seat was hammered in place. The valve
passed the leak test, but stuck shut during start-up." ,
NMPC Evaluation:
NMPC concluded that although the testing procedure did not allow it, the
leak testing of the feedwater check valves was done by pressurizing the
piping to 100 psi and then bleeding down to 35 psi for the actual testing.
.
-
11
.
.
Regarding the hammering, NMPC concluded that when an initial leak test of
feedwater check valve 31-01 failed, the valve center section was replaced.
During the subsequent leak test the tested portion of piping could not be
pressurized and the valve flange (part of the exterior of the valve) was
hammered to cause the internal valve disc to seat (i.e., to close). This
hammering had no effect on the condition of the valve. The check valve
then passed the leak test. Feedwater check valve 31-01 did stick open
during plant startup, and appeared to stick open during the subsequent
plant shutdown. Following this shutdown check valve 31-01 was disassem-
bled, and the Teflon seat was missing. The previously removed center
section was repaired and reinstalled into valve 31-01, and the valve had
no further problems.
Review and Findings:
The inspector reviewed the Instrument Surveillance Procedure, Procedure
No. N1-ISP-25.7, Feedwater Isolation Check Valves Leak Rate Tests,
March 17, 1986 and held interviews with the I&C supervisor and techni-
cians.
Based on the interview with the technicians, it was found to be common
practice to perform a local leak rate test on the feedwater check valves
in the following manner. Once the system is isolated, the inboard isola-
tion valve, an AC motor operated valve, is opened to allow the head of
water in the feedwater line, because of its configuration, to reverse flow
and seat the feedwater check valve. The inboard isolation valve is then
closed, and the volume between the two valves is press'urized to 100 psi.
The purpose of this pressurization is to quickly drain the water within
the volume and to assure the check valve remains seated. Once the water
is drained from the volume, the pressure is reduced to 35.5 psig in the
volume, and the test is performed.
The local leak rate test procedure specifies that the feedwater isolation
valve is to be cycled and closed. The drain line between the inboard
isolation valve and the outboard check valve is opened to drain the water.
Once the line is drained, the test apparatus is connected and the cavity
is pressurized to 35.5 psig and tested.
Concerning the valve hammering, results of the investigation showed that
during the performance of a preliminary LLRT performed on the valve before
it was installed into the system, the valve was leaking during the pres-
surization. The mechanic struck or hammered the valve flange, the valve
seated, and the test was performed. After the check valve was removed
from the system because of other problems, the valve internals were
inspected by Quality Control on August 11, 1986, and no evidence of
hammering on the disc was found. The details of that inspection are
documented in NMPC QA Inspection Report QCIR 1-86-1271.
..
.
.
12
.
.
,
Technical Significance:
Based on the above method used to leak test check valve 31-01, the NRC was
unable to ascertain the technical acceptability of the test. NMPC was
requested to justify the validity of the test. This item is unresolved
(50-220/86-17-02) pending receipt and review of NMPC's response. However,
we note that there is a remote manual power operated valve in series with
the check valve which is also subjected to local leak rate testing. This
provides assurance that the feedwater penetration meets the leakage
requirements of the Technical Specifications.
Conclusions:
The allegation relating to the use of 100 psi air was substantiated.
Technical Specifications 6.8.1 states, " Written procedures and administra-
tive policies shall be established, implemented and maintained that meet
or exceed the requirements and recommendations of Sections 5.1 and 5.3 of
ANSI N18.7-1972 and Appendix "A" of the USAEC Regulatory Guide 1.33...."
However, NMPC performed the local leak rate test for the feedwater check
valves in a manner not in accordance with approved procedures. Failure to
comply with Technical Specifications is an apparent violation
(50-220/86-17-01).
The second part of the allegation in which the alleger states that the
mechanic hammered the valve seat in place was not substantiated.
4.4 Feedwater Check Valve - Flow Diversion
Allegation:
"4. The Shift Supervisor diverted flow in the feedwater lines to free the
stuck feedwater check valve. There appeared to be no procedure for
this and no management review. Eventually, the valve opened."
NMPC Evaluation:
NMPC concluded that while the reactor was shut down on June 21 operator
actions were taken to free stuck feedwater check valve 31-01. However,
the actions were taken with the full knowledge and consent of management
and were acceptable according to the Technical Specifications, the plant
administrative controls, and operating procedures.
Review and Findings:
The inspector reviewed logs and records and interviewed personnel to
substantiate the facts as presented by the alleger and by the NMPC evalu-
ation. NRC review of the activities on June 21, 1986 to open check valve
31-01 essentially confirmed the information provided by the alleger, with
the exception of the extent of NMPC management involvement in the process.
Specifically, Operations personnel opened the stuck-closed feedwater check
-. . . . . - - . . -
I
.
.
13
.
.
valve by cycling feedwater system valves MOV 31-08 and FCV 31-128. The
activity was conducted as " troubleshooting" and without special proce-
dures. However, based on interviews with the applicable individuals,
management review of the process did occur through the direct involvement
of the Unit Superintendent and the Operations Superintendent, who were at
the site and participated in the evaluation of corrective actions.
,
During this evaluation, the decision was made to shut down the reactor to
reduce pressure downstream of the stuck-closed check valve (31-01) to
allow investigation and repair of the valve. The operators also entered
the Technical Specification 3.1.8 Limiting Condition for Operation (LCO),
based on a loss of redundancy within the High Pressure Coolant Injection
(HPCI) System (loss of one of the two injection paths). As part of the
troubleshooting actions, the operators partially closed the motor-operated
valve (MOV 31-08) in the remaining operable HPCI (feedwater) injection
line. NRC review noted that prior consideration was given to the poten-
tial impact on HPCI operability and the action was taken only after it was
concluded that a flow path capabic of delivering the minimum 3800 gpm
required for HPCI operability was assured. Since the valve was manually
closed three turns, which corresponded to about I inch of the 12 inch full
stroke travel for the valve, the minimum flow requirement was more than
adequately met.
One item was identified during the review of the operations area that
requires further review to assure an acceptable resolution. The item *
concerns the NMPC interpretation of the HPCI system description as
provided in the FSAR and the Technical Specifications. The existing NMPC
position is that redundancy within the HPCI system is restricted to the
active components upstream of common piping headers within the feedwater
system and thereby limited to the redundant condensate pumps, feedwater
booster pumps and feedwater pumps. This results in the two feedwater
lines downstream of the last stage of feedwater heaters being explicitly
excluded from consideration as redundant flow paths, since either one is
sized to deliver the required 3800 gpm flow rate for HPCI operability
requirements.
The inspector questioned NMPC's conclusion and stated that the NRC staff
position is that HPCI operability requires redundancy over the entire flow
paths within the feedwater system. As an example, closure or failure of
the containmant isolation valves (31-01 and -02 or 31-07 and -08)
downstream of the common header at the discharge of the feedwater pumps
constitutes sufficient reason to enter the TS 3.1.8 action statement
based on a loss of redundancy in the HPCI flow path. It was noted that
NMPC took this approach during the incident on June 21, 1986.
It appears that the appropriateness of NMPC's interpretation depends on
whether check valves in feedwate- lines FW 11 and FW 12 are considered
active components, and whether the single failure criterion should apply
to their function. The events on June 21, 1986 would seem to indicate it
is proper to treat the check valves as active components, and for the HPCI
.
~
14
.
.
system to be " single failure proof", both feedwater lines should be fully
operable.to meet the Technical Specification 3.1.8 LC0 for routine power
operations.
Further review of this matter is required between the NRC and NMPC staffs
regarding the plant conditions required to meet the requirements of TS
3.J.8. This item was discussed during the exit meeting on August 29,
1986, and NMPC was requested to submit a statement of its position in
writing to NRC Region I to allow further staff review of the matter. This
item is unresolved and will be tracked through a routine inspection report
(50-220/86-17-03). NMPC agreed to take a conservative approach with
regard to redundancy in the HPCI system flow path pending completion of
the NRC review and resolution of the item.
Technical Significance
The inspector reviewed the effects of manipulating feedwater system valves
to increase the differential pressure across check valve 31-01 by 200 psig
and the potential impact on the plant for the operational condition at the
time of the event. No unsafe conditions were identified, in that the
actions taken were within the limits and controls established by station
administrative procedures, operating procedures and the conditions of the
facility license. The inspector noted that the actions by the crew on
June 21, 1986 were less limiting than declaring both HPCI subsystems
inoperable and continuing with a plant shutdown per the Technical Specifi-
cation action statement. Such a course of action would have been allowed
by the facility license and would have been considered within the author-
ity of the shift supervisor's position to parform under the responsibility
of his Senior Reactor Operator license.
Additional Review
Plant operations areas were reviewed as a followup to the allegation. The
review included an inspection of shift activities during shutdown, startup
and routine power operations. The focus of the review was to determine
whether potential problems or deficiencies highlighted by the allegation
were indicative of broader based concerns or isolated deficiencies in the
operation area.
The inspector reviewed plant operating activities for Unit 1 and 2 to
determine whether concerns raised by the alleger in other areas were also
applicable to plant operations. The review focused primarily on the Unit
1 operating practices, but also included shift activities in the Unit 2
control room. This review noted that the operating procedures and
practices used on Unit I are also employed to a large extent on Unit 2.
The inspector noted, however, that there was a limited basis for comparing
Unit I and Unit 2 operations due to the differences in plant operating
modes at the time of the inspection.
.
-
15
.
.
Operating activities were conducted in accordance with an extensive number
of controls established in the operating and administrative procedures.
The conduct of. operations is prescribed by detailed procedures which
govern a multitude of operating activities, including record keeping
requirements, switching and markup operations, installation of jumpers and
lifted leads, operational surveillance testing, system and component
operations,. shift tours of operating equipment, and operations during
shutdown, startup and power conditions. Based on discussions with plant
operators and a review of a selected sample of procedures for markups,
, independent verification, jumper installation, system operations, repairs,
surveillance testing, and startup operations, the inspector determined
that operating procedures provide an adequate level of detail to direct
the activity and do not contribute to an " informality" in plant operating
activities.
Inspector observations of activities performed in the above areas deter-
mined that the procedures were followed or changed as necessary. It was
notable that changes (or exceptions) to valve lineups prescribed by system
operation procedures were processed as a temporary change _to the appli-
cable procedure. Other instances arose during the inspection where
temporary procedure changes were processed to allow deviation from the
instructions in the governing procedure; examples included the reactor
vessel hydrostatic test procedure used to perform single rod scrams on
August 19-21, 1986 and the changes to the reactor vessel drain line block
valve lineup made on an August 25, 1986 temporary change to the drain line
valve lineup. Based on the above examples, it appeared that supervisory
and support groups-become involved in the resolution of equipment issues.
In addition to the above, the inspector noted the following additional
strengths in the plant operations area: (i) there was good interaction
between supervisors and operators, and between operators and other groups
during the conduct of routine operations and testing; (11) staffing in the
operations area appears to be adequate, and there was good control over
the use of overtime; and, (iii) operator and nonlicensed training programs
appear effective based on inspector interviews with personnel in positions
from Station' Shift Supervisor (SSS) to Auxiliary Operator B (A08).
Conclusions:
This allegation was unsubstantiated in that although the specific opera-
tion took place, the concern relating to no p.ocedure and no management
review was unfounded. Based on the discussions nc+.ed above, the inspecter
determined that NMPC's followup reviews and actions f'r the specific
allegation were appropriate and acceptable. The inspector concurred with
NMPC's conclusion that the crew actions were within the scope of the
authority and responsibility vested in the operator position and exercised
as part of their individual licenses.
i
!
{
l
!
L
.
t .
16
.
.
Based on.this review, the inspection identified no findings which indi-
cated that the concerns raised'in other functional areas were applicable
to either the Unit 1 or Unit 2 operating organizations. Additionally, due
to the separate supervisory and operating shift staffs, the pending
licensing and startup activities of Unit 2 should not significantly impact
(or be impacted by) operation and management of Unit 1.
4.5 LPRMs
Allegations:
"5. During the outage non qualified technicians installed LPRM
connectors in that "A" techs installed them without direct
supervision from "C" techs."
"6. During.the outage and years prior, LPRM connectors were routinely
installed without proper Work Request (WR) paperwork, connectors
replacements were represented on WRs as trouoleshooting, and the
Installation and Test Procedure, LPRM-1, was routinely not used or
filled out af terward."
"7. Since the cable replacement six years ago, the LPRM cables have not
fit properly into the connectors. The cable dielectrics have been
meltad smaller (per LPRM-1) or the connector bores have been
drilled larger to fit them together."
"9. On July 10 a different design connector was installed on some LPRMs
(prior to being discovered by the resident inspector), and no
design change had been submitted for it. In addition, no work
requests or LPRM maintenance procedures were prepared until after
the resident inspector came down to witness this activity at which
time the workers involved took a break to generate the paperwork
and get it approved by the shift supervisor."
Initial NRC Inspection:
On July 11, 1986 the resident inspector observed portions of safety-
related maintenance on the Unit 1 Local Power Range Monitors (LPRM)
connectors. The inspector found two Amphenol Type BNC crimp-on
connectors used in place of the Amphenol Type SMA connectors required
per procedure N1-IMP-LPRM-1. One of the two BNC connectors identified
was installed in the cable supplying a signal to LPRM 28-098.LPRM
28-09B is one of eight safety-related LPRMs which provide input to APRM
channel 18. The inspector observed that no Quality Control personnel
were present during the LPRM connector maintenance.
The inspector discussed the installation of the BNC connectors with the
Lead QC Inspector and the I&C Supervisor. The I&C supervisor stated
that he was aware that unapproved connectors were being installed
t
.
.
-
17
.
.
contrary to the procedure controlling LPRM connector maintenance. The
Lead QC Inspector stated that there was no QC coverage due to ALARA
considerations.
NMPC Evaluation:
NMPC concluded the following:
--
"A" technicians worked only under the direction of "C" technicians
and no technicians other than the alleger were aware of "A"
technicians performing work beyond their qualifications.
--
Connectors were sometimes replaced when the WR only documented the
work as cleaning.
--
Procedure LPRM-1 was not required when the connectors replaced were
considered nonsafety-related, but the post-maintenance testing part ;
of it was routinely performed. The copy of LPRM-1 was filled out
after the work, because the area under the reactor vessel was
highly contaminated. If the copy had been taken there, it would
have been contaminated and required disposal.
--
In the early 1980s NMPC replaced the existing LPRM cable due to
moisture and degraded shield concerns. There have been problems
connecting the SMA connectors onto that cable ever since, because
the new cable dielectric is larger in diameter.
--
Due to connector / cable match-up problems, the cable dielectrics
were melted, which was approved in LPRM-1 and sanctioned by General
Electric (GE), and the connectors were machined, which was not
approved. Later evaluation concluded that the machining was
technically acceptable.
--
The BNC type connectors installed on July 10 were an unapproved
design. The I&C Supervisor intended that the new BNC connectors be
installed on nonsafety-related LPRMs and allowed to function in a
bypassed made to ascertain their durability. Some of the BNC
connectors were installed on safety-related LPRMs. A later
evaluation determined that the BNC connectors were a technically
acceptable replacement.
--
The control and accounting of the spare, uninstalled connectors were
inadequate. .
Review and Findings:
For the allegations associated with LPRMs, the NRC interview of I&C
personnel indicated that "A" technicians (technicians with the lowest
skill level) worked under the direction of "C" technicians (technicians
with journeyman skills). No one interviewed was aware of any instances
in which an "A" technician performed work without appropriate
.
-
18
.
.
supervision. Based on the interviews with technicians and supervisors,
work was performed based on individual skills and experience.
Assignments were made on each individual's scope of training and
experience, as documented in that individual's on-the-job training (0JT)
file.
The requirements for documentation of the work and implementation of the
LPRM-1 procedure were dependent on the activity performed, since not all
LPRMs were considered safety-related, and QC involvement and procedural
compliance were not a common practice for nonsafety-related installations.
The inspector noted that NMPC had previously considered the safety-related
LPRMs to be only those that were used in Average Power Range Monitor
(APRM) circuits. Therefore, approximately one-half of the 120 LPRMs
were considered nonsafety-related. Following the discovery of the
unapproved connector design by the NRC on July 11, NMPC revised the
maintenance of LPRMs, such that all LPRMs are now treated as safety-
related.
Work Requests (WRs) for safety related LPRMs (i.e., those in APRM
averaging circuits) contained procedural deficiencies, in that the
quality classification, procedural requirements, material control, and
extent of maintenance task were not always specified.
Engineering and management awareness of, and involvement in, the ongoing
LPRM cable / connector problem were not evident. I&C supervisors burdened
with heavy work loads apparently relied on technicians to resolve the
issue.
Methods used to complete the LPRM cable / connector assemblies included
the melting down r' the dielectric to fit the connector and the
enlargement of th; .onnector inner diameter by machining. The practice
of melting the cable dielectric to fit the connector is a process
approved by-the General Electric (GE) Company and is sanctioned by NMP-1
procedure LPRM-1. However, the enlargement of the connector inner
diameter by machining is not covered by any procedure. Although the
action was subsequently evaluated as being technically acceptable, the
connector should not have been reworked without the proper technical
review and material controls to verify that modifications did not
compromise system operability.
A review of Work Requests for work performed by I&C on other systems
indicated that on Work Requests (WRs) for major safety-related tasks the
WR package was well documented, containing the necessary QC hold points,
procedural requirements and instructions for performing the task. On
minor safety-related tasks, such as repair or replacement of a switch,
the WR packages lacked definitive instructions and apparently relied on
the skills and experience of the individual assigned to perform the
%ork. QC was not generally involved, but there were procedural
requirements for post maintenance testing to check the function and
operability of the component. QC surveillance was used as a means of
assuring adequate quality. For nonsafety-related maintenance tasks no
4
.
-
19
.
.
documentation (other than the WR) or QC involvement was required. The
task was performed by the individuals assigned, that assignment being
based on the skills and experience noted in each individual's on-the-job
training (0JT) file and the personal knowledge of the supervisor.
Technical Significance
Post-maintenance testing of the LPRM circuit was a general practice and
was performed in accordance with sections of the LPRM-1 procedure. In
addition, subsequent calibration after startup substantiates the
adequacy or failure of LPRM circuits. Failed LPRMs are bypassed in
accordance with the Technical Specifications. There is no safety
concern, since the APRM failure mode is upscale for scram and downscale
for inoperable, both of which are fai1~ safe failure modes.
Conclusions:
These allegations were substantiated, in that for the allegations
associated with the LPRMs there was an apparent programmatic deficiency
in controlling the materials, procedural compliance and QC involvement
to ensure quality.
First line supervision's kr.owledge and participation in the ongoing LPRM
problem was evident. In reviewing I&C performance in other areas, it
appeared that the deficiencies noted in the allegation were limited to
the LPRM and IRM (discussed below) issues.
Technical Specification 6.8.1 requires that procedures be established,
implemented and maintained that meet or exceed the requirements of
Sections 5.1 and 5.3 of ANSI N18.7-1972, which requires procedures
governing plant maintenance. NMP-1 procedure N1-IMP-LPRM-1 requires the
use of Amphenol Type SMA connectors during LPRM connector maintenance.
Contrary to the above, two Amphenol Type BNC connectors were installed
in place of the connectors required by procedure. This is an apparent
violation. (50-220/86-17-04). -
4.6 QC Involvement
Allegation:
"8. QC involvement in the LPRM connector work was improper in that I&C
techs frequently did not inform QC that connetters were being
replaced, and even when aware of the connectos replacements, CC
inspected only paper and never went under the vessel because they
knew the work was unacceptable to specifications."
NMPC Evaluation:
NMPC concluded that no QC hold points for inspection of the connectors
existed in tne procedures. The intended method for assuring the quality
of this work was surveillance by the QA Surveillance Group. Inspections
.
I
4
.
20
.
.
were minimized due to ALARA (As Low As Reasonably Achievable)
considerations of inspector exposure, most LPRMs were classified as
nonsafety-related, and post-maintenance testing was sufficiently
verified by Level 2 qualified people.
Review and Findings:
A review of records, and discussions and interviews with personnel
, indicated that although inspection under the reactor vessel was not a
common practice, as concluded by NMPC above, QC had previously conducted
physical examinations and observations of LPRM connector work during
previous outages between 1984 and 1986. The Unit 1 Quality Control
Inspection Reports (QCIRs) associated with LPRM connector work during
the past two years were reviewed.
Conclusions:
The claim that QC reviewed only " paper" and "never went under the
vessel" was not substantiated.
4.7 Harassment
Allegation:
"10. During the outage the alleger was harassed by fellow workers and
discriminated against by his supervision due to his raising
concerns about the LPRM connector work. The supervisors did little
or nothing to correct his harassment."
Conclusion:
The inspection concluded that it was not an appropriate time to inspect
this allegation. This allegation was being investigated by the NMPC
special investigation discussed in Section 6. NRC investigation of this
allegation, as appropriate, will be done following review of the final
NMPC investigation report and Department of Labor action.
4.8 IRMs
Allegation:
. "11. The connector on IRM 18 was replaced on June 17, 1986, and was not
documented on the WR."
"12. The plant was started up on the morning of June 17, 1986 based on
falsified surveillance test records for the replaced IRM connector.
The I&C techs and assistant supervisor falsified the test record
without performing any of the required surveillance testing."
-
. . . .-
.
.
21 .
.
.
NMPC Evaluation:
On June 17, work was performed on IRMs 13, 16, and 18 under WR 015772.
The WR documented only cleaning of the connectors. The individuals
other than the alleger stated that they were unaware of a connector
replacement. However, the Radiation Work Permit (RWP) indicated that
the IRM 18 connector was replaced. Further, since the alleger claimed
to have replaced the connector himself, NMPC could not conclude that the
connector was not replaced.
On the testing, NMPC found that WR 015772 indicated that post-maintenance
testing should be performed under procedure NI-ISP-NEU-2. However, there
was no procedural requirement to perform any testing following connector
cleaning, nor was there any specific procedure covering testing following
connector replacement. Common practice involved circuit tests, including
Time Domain Reflectometer (TOR) traces, for information. Record of the
traces could not be found, but review of the process computer showed that
IRM 18 was bypassed at 2:59 a.m. and was returned to unbypassed at 3:11
a.m., the approximate time needed for the TDR trace. Based on this NMPC
concluded that TDR traces were done.
Review and Findings:
As the problem of undocumented work on LPRM connectors (covered in
Section 4.5) was already under review, the IRM aspect of this issue was
not pursued during the inspection beyond a review of the applicable
paperwork.
Concerning the testing, the inspector reviewed the documents referenced
in the NMPC evaluation and found no inconsistencies. It was determined
that further review would be appropriate following completion of the NMPC
investigation and the NRC review of its conclusions.
Technical Significance
The IRMs have a fail-safe design and will trip and result in a reactor
scram if a failure causes an inoperative or upscale reading. Further, a
rod block results if a downscale reading occurs. As the IRM calibrations
required by Technical Specifications were performed during the week prior
to the alleged event and following the IRM detector replacements which
occurred later on June 17, this allegation has minor technical
significance.
Conclusions:
The inspector concluded that the conclusions regarding the unauthorized /
undocumented replacement of LPRMs in Section 4.5 also applied to IRMs.
Further, the alleged falsified testing of the IRMs will be reviewed when
the NMPC investigation is completed.
,
n --- --
n -- - , . - - - . - -
O
22
.
.
4.9 Unplanned Exposure
Allegation:
"13. An I&C technician working on an LPRM connector received a dose of
1.25 rem which was in excess of his administrative limit.
Conclusion:
This allegation had been previously reviewed during a radiological
controls inspection during August 4-7, 1986 and is documented in
Attachment 3 - Combined Inspection Nos. 50-220/86-16; 50-410/86-46
dated September 30, 1986. The findings of the inspection will be
discussed as part of the enforcement conference covering the
allegation review.
4.10 Tool in Reactor Vessel
"14. A piece of an aluminum tool about 1 inch by 8 inches was lost
in the reactor vessel during the outage. The tool was used for
installation and removal of feedwater line plugs."
NMPC Evaluation:
NMPC had previously concluded that a six inch piece of a tool used to
install the emergency condenser line plug was lost during the outage
and performed an evaluation since the piece was not retrieved prior to
start-up from the outage. General Electric Report MDE 720586 dated
May 21, 1986 reviewed the event and concluded that the piece would be
oxidized and disintegrate. The piece was judged to present no
potential for chemical reaction, flow blockage, or interference with
control rod motion, and safe reactor operation would not be
compromised by it. The Site Operations Review Committee (SORC)
reviewed the report on June 13, 1986.
Conclusion:
The inspector reviewed the NMPC evaluation and concluded that this
event would be more appropriately reviewed in a routine inspection
report as the event had no direct connection with the alleger,
>
appeared largely unrelated to the other allegations and was the
subject of a previous engineering evaluation. This event will be
tracked as (50-220/86-17-05).
_
.
23
.
5. QA PROGRAM REVIEW
Due to the specific allegation concerning QC involvement in LPRM work
and the apparent lack of QA/QC involvement in the other allegations, the
inspection team concluded that a general review of the quality assurance
program was appropriate.
An onsite (Units 1 and 2) Manager-Nuclear QA Operations and his QA/QC
organization report to the offsite Vice President-Quality Assurance
(VP-QA). Also, an onsite QA Audit group reports to an offsite
supervisor who reports to the VP-QA. The following comprise the onsite
QA/QC presence:
--
A Supervisor-Quality Control (QC) with a separate subgroup of
inspectors assigned to each unit.
--
A separate Supervisor-Quality Engineering with a group of QA Engineers
assigned to each unit.
The QC groups are responsible for first level independent inspection of
work associated with safety-related and other Q list components / systems. Until
recently Unit 1 QC also performed a surveillance function of activities
in a high radiation area such as LPRM repair and replacement.
Currently, a Quality Control Inspection Report (QCIR) is developed for
each procedure or specified activity. This QCIR includes scope of work
information, inspection attributes to be observed or reviewed and space
for documenting results. A similar checklist had been used for the
previous QC surveillances. A QCIR is used each time an inspection is
conducted, and this is normally in conjunction with a Work Request (WR),
which is generated for every work activity, including troubleshooting
and tasks considered to be within the expected skills of the mechanic or
technician. In addition to the QCIRs reviewed on LPRM work, the
inspector reviewed a sample of QCIRs from other types of Unit 1
inspected activities. Those QCIR attributes that are asterisked on the
QCIR form must be addressed by the inspector, and should time be limited
.because of workload or other considerations, the remainder need not be
done. Interviews and discussions were held with the Unit 1 QC
Supervisor and cognizant inspectors about inspection scheduling and
methodology. The following aspects of QC everview were identified based
on the foregoing evaluation:
--
A WR issued for troubleshooting or for work within the expected
skills of the worker is designated as "QC not required" but is
annotated that notification must be made to QC if repair / replacement
of an item is decided upon.
L
--
The balance of safety-related or Q categorized WRs have some level
of inspection done on the work,
s
- -. - . l
.
24
.
.
.
--
Using the same checklist for a given inspection one person may
observe ongoing work and clearly document this fact, while a second
person would do the same type of inspection and paraphrase the
inspection attribute (s) as documentation.
--
A third person would use the same checklist for the inspection as
above and do only a record review documenting this fact clearly, .
while a fourth person would do the same document review and
paraphrase the inspection attribute (s).
--
The prevalent method of documenting inspections was paraphrasing
the inspection checklist attribute (s), and this did not clearly
state whether examinations / observations were done or only records
were reviewed when the attribute was to be " verified".
--
Unasterisked attributes (optional) were addressed in the majority
of QCIRs reviewed.
The Unit 1 QA Surveillance Group is responsible for conducting a second
level monitoring effort of ongoing activities associated with functional
arias such as chemistry, fire protection, maintenance, operations, and
radiation protection. Technical Specification (TS) required plant
surveillances are also monitored and a schedule has been developed that
matches TS line items to applicable calibrations and tests so that each
line item will be addressed at least once during a two year cycle. These
QA surveillances are done using checklists similar to those of QC inspec-
tions. Discussions and interviews with the cognizant supervisor and three
group members indicated that surveillance methodology and interpretation
of the word " verify" was almost identical to that of the QC group. A
review of QA Surveillance Reports identified the same lack of clarity in
documenting what was done as in the Unit 1 QCIRs.
The Unit 2 Program was reviewed. The Unit 2 QA Surveillance Supervisor
was interviewed, and discussions included details of how the groups
responsibility would be implemented. It was determined that little
similarity existed between the two QA surveillance groups' monitoring
ef# orts. The following were some of the major aspects of the Unit 2
program.
--
A methodology and training guide manual for the conduct of QA
surveillance was developed.
--
The various amounts of surveillance of a given function were based
on a Probability Risk Assessment (PRA) type of analysis.
--
The number of observations vs the number of identified deficiencies
allowable in a given functional area are tracked so as to maintain
a 95% confidence level of acceptable performance in a given
functional area.
..
.
-
25
.
.
--
Additional surveillances would be conducted in an area where the
confidence level fell below the established value.
--
Approximately 80% of the anticipated checklists have been developed
using a survey methodology with some already being used for ongoing
activities so as to determine where revisions are needed.
--
The schedule is based on the 18 month refueling cycle, and
notificationlof an impending QA surveillance is planned to be given
just prior to its commencement.
Several QCIR checklists developed for use at Unit 2 were reviewed and
discussed with the cognizant QC Supervisor. The checklists were task
oriented, used the word " verify" sparingly, and had acceptance criteria
noted thereon or referenced the appropriate procedure paragraph. It was
noted that a number of guidelines had been developed for specific
inspections such as cable terminations.
The onsite audit group has been responsible for auditing only Unit 2
activities prior to this year. Discussions with the supervisor of this
group indicate that management's intent is for the onsite group to
assume most, if not all, of the onsite audit responsibilities. However,
the final balance between offsite and onsite audit responsibilities has
not yet been finalized. Audits, by nature, are primarily a review of
programs, procedures, records, and other documents, and a review of
Unit 1 audit packages (e.g. checklists, field notes, corrective action
requests) from the corporate audit group reflected this approach. The
audit packages for this year's three Unit I audits that were conducted
by the onsite audit group documented a greater degree of work observation,
physical examinations, measurements, etc. than the others. A review of
a few Unit 2 audit packages indicated emphasis on this approach to audit-
ing and when observations etc. were done, and that fact was clearly
documented.
The Unit 1 QA Engineering group has been reviewing WRs prior to any work
being accomplished and developing inspection checklists based on
approved procedures. Discussions with the supervisor of the Unit 2
group indicated that the development of their checklists will include a
degree of source document review so as to assure specifications, vendor
technical manual, etc. requirements and recommendations are indeed
reflected in those procedures.
.
Conclusions:
The effectiveness of the QA/QC organization in identifying the type of
practices contained in the allegations is doubtful based on the following
assessment.
.
-
26
.
.
--
Should QC not be notified that a repair or replacement will be done
under.a " troubleshooting" or " worker skills" WR, any unauthorized
or out-of-scope work would not be identified.
--
Should a QC _ inspector only review records, unauthorized, improper,
or out-of-scope work would not be identified.
--
A QA surveillance that consisted of only record reviews would have
no probability of identifying unauthorized, improperly conducted,
or out-of-scope work.
.
--
Audits, by their nature, have a low probability of identifying
unauthorized, improper or out-of-scope work.
--
The leading root cause of differences in the conduct of QC inspec-
tions and QA surveillances is the manner in which the word " verify"
is used in a checklist attribute.
--
Of the three levels of QA/QC overview only QA surveillances that
emphasize unanticipated examination or observation of ongoing work
have a high probability of identifying unauthorized, improper or
out-of-scope work.
In summary, although the QA/QC program is being implemented, it is not
being utilized in a manner that permits it to be an effective management
tool to find and correct problems, of the type disclosed during our
reviews of the subject allegations.
.. .
.
-
27
.
.
6. REVIEW 0F NMPC INVESTIGATION
Review and Findings:
This section details a review of the " Summary Report of Niagara Mohawk
Power Corporation's Investigation of Allegations by NMP-1 Instrument and
Contrcl Technician" by a NRC Region I Investigator.
Following notification of NMPC by the alleger on July 15, 1986, NMPC
conducted interviews with the individual to document his concerns and
were notified by Region I on August 11, 1986 of allegations received by
the NRC. At the time of the special team inspection on August 25-29,
1986 the NMPC investigation remained in progress.
This review was based on interviews of the lead Investigator, a NMPC
Assistant Security Inspector, and his assistant, a NMPC Security
Investigator, relating to the conclusions and basis for those conclusions,
as set forth in both the captioned report and confidential investigation
relating to the same subject matter.
The exhibits relating to the reports were voluminous, and only a select
few were reviewed in depth.
Conclusions
Both investigators appeared to be experienced, competent, and credible;
however, this is their first experience at a complex regulatory
investigation.
Conclusions appeared to be consistent with the investigation; however, it
was readily apparent the investigation was incomplete and numerous logical
leads needed to be followed-up. The investigators acknowledged this fact
and emphasized that their investigation was ongoing.
The investigation appeared to be focused entirely on proving or disproving
the allegations identified by the alleger on a technical basis. Certainly
this needs to be addressed, but more attention should be'given to the
scope, responsible individuals, and circumstances which allowed existing
conditions, e.g., corrective action for falsification of LPRM work
requests was addressed, but investigation has revealed similar allegations
in the I&C Department in other areas indicating a larger possible
programmatic problem.
A lot of work has been completed in a short time, and a review of inter-
views indicated that the investigators have not had sufficient time to
correlate the information obtained in the various interviews. Conse-
quently, individuals were being interviewed two and three times on the
same subject matter.
.
.
28
.
.
Investigators need more technical support in order to identify, provide,
and explain site documents and records in support of the investigative
effort. For example:
^
1. Security computer and RWP records should be compared to LPRM work
requests in order to prove or disprove the allegation that unquali-
fied technicians were changing LPRMs under the reactor vessel;
2. LPRM work requests which state _" inspect and cleaned" should be
compared to I&C Department log entries which indicate " connector
changed";
3. SORC minutes and possible other management and supervisory notes
from routine department meetings should be identified and reviewed
to determine if there was any management concern on decisions
relating to the alleged CRD pump vibration problem.
Interviews need to be more probing and interviewees need to he
challenged to assure that they fully answer the questions.
The allegations relating to various LPRM problems should be reviewed for
management involvement since both the QC Supervisor and I&C Supervisor
have testified to NMPC investigators that they have been aware for years
.that there have been problems mating the LPRM connectors with the cables
under the reactor vessel. Additionally, the I&C Supervisor, who was a
former I&C Technician, admitted that he knew LPRM-1 was being violated
during the installation of LPRMs.
'
Conclusions:
The investigators appeared to be independent of any management influence
relating to both their investigative activity and conclusions. Given
sufficient time and resources, the investigation should be capable of
determining the facts of the allegations.
The NRC will review the final report upon completion of HMPC's activities
(50-220/86-17-06).
s.
.
29
.
.
7. REVIEW OF CRD VALVE MAINTENANCE EVENT
7.1 Event
This maintenance activity was selected for further inspection by the
NRC as a measure of NMPC's ability to: respond to an event; investigate
and identify root causes; institute appropriate corrective action; and
effectively and critically self-evaluate existing programs. Conclusions
reached regarding the above are described later. On September 19, 1986
NMPC reported the circumstances of the Unit 1 shutdown required per
Technical Specifications in LER 86-26.
An unusual event was declared at Nine Mile Point site at 1:00 p.m. on
August 22, 1986, and an orderly shutdown of the Unit I reactor from
99.5% power (1841 MWt) was begun. The shutdown was initiated because of
the indeterminate operability of, at that time, ten control rods. The
control rods in question had experienced inadequately controlled
corrective maintenance during the previous 27 hours3.125e-4 days <br />0.0075 hours <br />4.464286e-5 weeks <br />1.02735e-5 months <br />, in that the packing
had been tightened on the hydraulic control unit (HCU) scram inlet
valves without the approval of licensed shift supervision and also
without appropriate post-mainterfance testing. NMPC commenced an
investigation of the work performed on the ten scram inlet valves in
question on the afternoon of August 22 in parallel with the unit
shut down. NRC Region I was informed of this situation by phone at that
time. However, later that day NMPC's investigation found that the
corrective maintenance had been performed on all 129 control rod drive
(CRD) HCUs on both the scram inlet and outlet valves, thereby potentially
affecting the scram insertion times of all control rods. The reactor was
placed in a shutdown condition with all control rods inserted by 7:28 p.m.
on August 22.
The performance of work on the HCUs which resulted in the reactor
shutdown on August 22 was potentially in violation of Technical Speci-
- fication Limiting Conditions for Operation (LCO) for the control rod
I drive system. Specifically, the Action Statement for LC0 3.1.1.C
requires the reactor to be placed in a hot shutdown condition within 10
hours of exceeding the maximum and average control rod scram insertion
-
times specified therein. The maintenance repairs to the scram inlet and
outlet valves (hereafter referred to as the CRD No. 126 and 127 valves,
respectively) were performed on all HCUs between 10:15 a.m. and 2:00
p.m. on August 21.
The operability of all 129 control rods became (at that point) unknown,
t and the Technical Specifications therefore required the reactor to be
placed in the hot shutdown condition. The indeterminate HCU maintenance
existed for approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> until discovery by a Station Shift
Supervisor at 10:30 a.m. on August 22. Because the reactor was brought
to a hot shutdown condition approximately nine hours after the discovery
.
of the HCU maintenance by plant management, and because the 126 and 127
valve packing adjustments were later determined to have no measurable
l
.
.
30
.
.
effect on control rod scram insertion times (as discussed in Detail
4.1), no violation of Technical Specifications occurred. However, a
violttion associated with administrative controls applied to maintenance
work was identified and is discussed later.
NMPC's investigation of the work performed on the CRD hydraulic control
units'was discussed in a special meeting of the Site Operations Review
Committee (SORC) held onsite on August 23. The 50RC reviewed a memorandum
from the Site Superintendent (SP) of Maintenance who ccnducted the
investigation that described the sequence of events associated with the
work. The SORC also reviewed revisions made to the administrative proce-
dure governing corrective maintenance repairs as a short term corrective
action required prior to startup of Unit 1. A startup was begun on August
24, and all 129 control rods were tested and met their required scram
insertion time intervals as specified in the Technical Specifications.
Power ascension was continued, and the reactor reached full power by
August 26, 1986.
The following sequence of events was obtained from document reviews and
interviews conducted by the inspector and, in some cases, add to or
differ from NMPC's investigation of the actior.s summarized in the August
23, 1986 memorandum (K. Dahlberg to T. Perkins). The HCU scram inlet
and outlet valves are referred to as the CRD No. 126 and 127 valve,
respectively. The work request (WR) under which the maintenance was
performed was WR No. 102775.
DATE/ TIME COMMENTS
Friday, Quality Control (QC) Inspector discovered loose
August 15 packing nuts on the CRD No. 126 valves of 10 HCUs.
Work Request No. 102775 was initiated by a QC inspector and
approved for work by the Assistant Maintenance Supervisor.
Monday, The WR was received and approved by the Maintenance
August 18 Supervisor who specified that no procedure was required to
tighten the packing nuts.
Tuesday, Quality Assurance (QA) review of the WR noted that
August 19 inspection would be required.
Thursday,
August 21
7:30 a.m. The WR was assigned to the Chief Mechanic and a working
crew of three.
.
,
.
-
31
.
.
7:45 a.m. A Radiation Work Permit RWP request was submitted for
"CRD No. 126 inlet scram valves (various) packing
leaks."
8:00 a.m. QC was notified of work.
9:30'- 10:00 a.m. RWP approved by Radiation Protection, Lead Worker
and Assistant Shift Superintendent (also the Shift
Technical Advisor (STA).
10:15 - 11:45 a.m. Work crew (three mechanics and Chief) logged in and
out onto the RWP. The workers tightened the packing
on scram inlet and outlet valves (CRD No. 126 and
No. 127) as well as the packing on the accumulator
charging supply and drain valves (CRD No. 106 and
No. 107) of all 129 HCUs.
1:00 - 1:30 QC inspector observed bolts already tightened on 9
HCUs specified on WR No. 102775, and discussed
options to replace missing nut (one of four) on
HCU 14-47.
3:00 - 3:30 p.m. Chief Mechanic discussed mark-up (i.e., tag-out) of
CRD No. 126 valve with shift supervision so that a
replacement nut could be installed. Shift
supervision was not aware of the extent of repairs
made at this time and decided to postpone proposed
nut replacement on HCU 14-47 until the next day. ,
Friday,
August 22
7:30 - 9:00 a.m. Discussion between. shift supervision, QC and
maintenance personnel on how to replace missing
CRD No.126 valve packing gland flange stud nut on
HCU 14-47. Operations decided that a mark-up was not
feasible at thaE4 time, because of difficulties
associated with the resulting inoperability of control
rod 14-47. A consensus was reached to replace the nut
without a mark-up. Otherwise it would be necessary to
open the 127 valve or disassemble its split stem
9:20 - 9:25 a.m. RWP approved to replace missing nut. The task
description on the RWP stated "take up on scram inlet
valve No.126 packing.. . Reactor Building el. 237
accumulator area".
- - - _ - _ - - _ - - . _ - -
. . ,- _. ..
_ _ _ _ .
_._
3
..
-
32
.
'
.
10:30 - 10:45 a.m. Nut replaced on CRD No. 126 valve of HCU 14-47 which
necessitated the filing off of a small section/ edge of
stem coupling.
The Station Shift Supervisor went to the work site (at
HCUs) to observe nut replacement and discuss job with
Chief Mechanic. -The Station Shift Supervisor
discovered that the packing had been tightened on ten
specific scram inlet valves.
11 a.m. - 12 noon Extent of packing adjustments and effect on control
rod operability discussed between Assistant
Superintendent of Operations and shift supervision.
1:00 p.m. Unusual Event declared based on initiation of plant
. shutdown required by Technical Specification Action
Statement for control rod operability.
1:30 - 1:35 p.m. QC inspection of replacement nut.
2:00 - 8:00 p.m. Licensee investigative meetings.
7:28 p.m. All Unit I control rods fully inserted.
Saturday,
August 23
12:30 - 2:45 p.m. Special SORC meeting to review investigation results
and subsequent plant startup.
Sunday, Control rod scram insertion time testing
August 24 conducted in accordance with Procedure No. N1-ST-R1
for all.129 control rods and completed satisfactorily.
7.2 Review and Findings
Discovery of Loose Nuts
The inspector interviewed the QC inspector who discovered the loose
packing gland flange nuts on nine HCU scram inlet (CRD #126) valves as
well as a missing nut on one of the four studs on the HCU 14-47 scram
inlet valve. The QC inspector had observed a similar situation on one
HCU approximately one year ago. While waiting to inspect an unrelated
filter replacement on August 15, 1986 the QC inspector visually
inspected the scram inlet and outlet valves for all 129 HCUs and
identified the HCUs with loose or missing nuts. The loose nuts were
visually detected (i.e., the QC' inspector did not touch the nuts to
ascertain tightness), were backed off of the gland flange by a few
threads (no more than one-eighth inch), and typically were only one of
the four nuts on each scram inlet valve.
.
- - . - . ._ _,. - - - _ _ _ - __ _. m. _, _ _
.
.
33
.
.
The QC inspector noted the ten affected HCUs, informed control room
supervision of the condition, and because licensed operators were busy
at that time, originated Work Request (WR) No. 102775. The QC inspector
stated that there was no evidence of packing leaks on the ten valves
identified. The responsible work group, Mechanical Maintenance, was
then contacted by the QC Inspector, and the WR was approved.
Maintenance Approval of Work
The approved WR No. 102775 was received and reviewed by the Supervisor
of Mechanical Maintenance on August 18. Since the supervisor considered
the tightening of the nuts (and therefore the packing) to be within the
normal ability of a journeyman mechanic, a written procedure to perform
the won was determined to be not required. The inspector interviewed
the Mechanical Maintenance Supervisor and confirmed that, while he had
also not specified post-maintenance testing at that point, he was aware
of the impact on CRD No. 126 valve stroke time and consequent control
rod operability as a result of tightening the valve packing. However,
he failed to note those considerations on the WR or to discuss the
feasibility of this work with Operations personnel. Further, he did not
investigate the need to perform this work and was unaware that there were
no actual packing leaks.
QA Review
A Quality Assurance review of the proposed work under WR No. 102775 was
conducted, as required by Administrative Procedures AP-5.0 for repair,
and QC inspection was indicated as being required for this job.
However, since there was no applicable procedure to perform this work
and no prepared QC inspection criteria existed, there were no acceptance
criteria or inspection and quality requirements specified by the QA
reviewer. More pointedly, it was unclear whether the scope of the job
was intended for QC observation of work in progress or after the fact QC
inspection of the completed work. In fact, as described later, the
post work QC inspection that was performed failed to identify the lack
of Shift Supervision approval and the maintenance that occurred beyond
the approved work scope.
Work Crew Experience
Because of the apparent simplicity of the job and a failure on the part
of first-line maintenance supervision to recognize the potential effect
of packing adjustment on valve stroke time and control rod operability,
the work under WR No. 102775 was assigned on August 21 with no
additional direction or special instructions other than mentioned above.
The work was assigned to a crew of three workers under the direction of a
Chief Mechanic. The assigned Chief was a journeyman mechanic with signi-
ficant experience of approximately 35 years with the company, over ten of
which have been at the Nine Mile Point site, and the last 2-3 years as a
working foreman or Chief Mechanic. The Assistant Supervisor who assigned
the work on August 21 also has had extensive experience with NMpC of over
.
.
f
-
34 ,
'
-
\
i
- 19 years, working as a Chief Mechanic for 3 years and as an Assistant
Supervisor for the past 2 years. The remaining work crew consisted of
'
.
two less experienced "B" level mechanics and a helper.
Supervisory Direction
'
The inspector interviewed the Chief Mechanic who performed the work
under WR No. 102775. The Chief stated that he was not aware of the
significance of tightening the packing on the scram inlet and outlet
valves, nor was ne knowledgeable of the function of the valves and their
effect upon control rod drive scram insertion capability. The Chief was
cognizant of the safety-related classification of the work, but was
generally unaware of the function of the HCUs. The Chief stated that,
upon assignment of the WR (to tighten packing on the ten identified
valves), he was verbally instructed by the Assistant Supervisor to tighten
any other loose nuts found. No specific criteria for tightness were
provided to the workers, either verbally or in writing, as to either hand-
tightness or a specified torque value. As described later the General
Electric Operating and Maintenance Manual for the HCUs was not consulted
at this point, and no effort was made to inspect or examine the CRD No. 126
valves prior to performing the work.
The inspector could not corroborate the Chief Mechanic's statement that
he had been directed to tighten loose nuts on all HCUs (other than just
the ten scram inlet valves identified on the WR).
Notification of Shift Supervision
Section 5.7 of AP-5.0, the Administrative Procedure for Repair, required
notification of Shift Supervision prior to commencing work on WR No.
102775. Because no procedure was being used and no mark-up (equipment
tagout) was deemed required, the normal interfaces with control room
licensed operators did not occur, whereby a notification would otherwise
be made. This lef t the RWP as the only other means of control room
operator notification. Notification of work via an RWP is not the
method intended by AP-5.0, nor is it an appropriate method as evidenced
by the failure to clearly convey to the Shift Supervisor the extent of
the work under WR No. 102775. Nonetheless, interviews with the Chief
Mechanic and his immediate supervision indicated that they had
considered the RWP approval by the SSS (or in this case his assistant)
to constitute appropriate notification and had in fact filled in (after
the fact) line 26 on WR No. 102775 as providing for such.
The failure to clearly notify Shift Supervision of the intent and scope
of work under WR No. 102775 was a primary cause leading to the failure
to recognize the potential effect of scram valve packing adjustment on
control rod operability.
l
t
!
.
-
35
.
.
Initial Packing Adjustments
Interviews with the Chief Mechanic found that, in addition to the ten
valves specified on the WR, an unspecified number of scram inlet and
outlet valves (CRD No. 126 and 127) were tightened by the work crew on
August 21. The Chief stated that a crescent wrench was used to
" snug-up" the loose nuts; therefore, more than hand-tightening but an
indeterminate amount of torque was applied. Additionally,'the CRD No.
106 and No. 107 valves were tightened. These are manual valves
associated with the HCU accumulator charging fill and drain connections
and do not directly affect control rod operability. The Chief was not
aware of how loose the identified packing gland flange nuts were prior
to tightening, but subsequent discussions held by NMPC with the work
crew members involved confirmed that, typically, one of the four nuts on
the CRD No. 126 valves were backed off of the flange by a few threads.
Also, as stated by the QC inspector who discovered the loose nuts, there
was no packing leakage observed by the work. crew at the HCUs.
Packing Leakage
The inspector evaluated the potential significance and sources of
packing leakage from the scram valves. Packing leakage from the CRD No.
126 valve stem would be CRD system drive cooling water at a pressure
higher tnan reactor pressure and originating from the condensate storage
tank via the drive under piston area. Packing leakage from the CRD No.
127 valve stem would be standing water at static pressure and
originating from the scram discharge volume header. The scram inlet and
outlet valves on each HCU are located such that, due to the radiation
fields present (which were 10-18 mrem /hr at the time of the inspection),
an RWP is required to either visually inspect or work on the valves.
However, there was na packing leakage apparent during the period of
August 15-22, 1986. Moreover, the inspector observed no visible
evidence of previous packing leakage at any HCU scram inlet or outlet
valve, except for one HCU.
The CRD No. 126 valve on HCU 14-47 was also the valve that had one of
four packing flange nuts missing. The inspector examined the stem to
c3
'
body area of that valve and noted discoloration and rust indicative of a
previous packing leak. The inspector reviewed the history of previous
work performed on HCU 14-47 bat could not identify any instances where
the scram inlet valve would have been disassembled (and a nut therefore
misplaced) nor any instance where a packing leak was identified and
corrected.
Replacement of Missing Nut
Following the tightening of packing on the HCUs on August 21, the Chief
Mechanic had discussions with the control room shift supervision on how
to replace the missing nut on HCU 14-47 since, with the valve closed,
the stem coupling presented a slight interference to installing a new
nut. At this time, the Shif t Supervisor was unaware that the mechanics
.
.
36
.
.
had worked on tightening the packing on other HCU valves and had
informed the Chief that a " mark-up" (i.e., equipment tag-out) would be
required to disassemble the valve to move the coupling up and away from
the packing gland flange studs. The_ mark-up would necessitate rendering
the associated 14-47 control rod inoperable and would require an
evaluation by a Reactor Analyst. In an interview conducted by the
inspector, the Station Shift Supervisor (SSS) stated that at that point
he had not seen WR No. 102775 and his discussion with the Chief
concerned how to install the missing nut. The SSS requested the Chief
to return the following morning to obtain appropriate staff reviews.
>
\
Early on the morning of August 22, the Chief again presented various
options to the SSS to replace the missing nut. These options consisted
of: a " markup" to tag-out the valve and disassemble or-move its stem
coupling; shaving the nut or filing the coupling to provide for
necessary clearance; or, cutting off a segment of the stud. The SSS
declined to provide a mark-up because he did not desire to make the
control rod inoperable. The Chief returned to discuss these options with
maintenance supervision and with the QC inspector. The eventual solution
was to shave a corner section off of the coupling and add the fourth nut.
Discovery of Work Scope ,
During the time the addition of the nut on HCU 14-47 was accomplished
(1030-1045 a.m.), the SSS became curious as to this job and went to the
HCU area to observe and discuss the work with the Chief. At that point,
the SSS became aware of the tightening of the ten HCUs described on WR
No. 102775, and he informed Operations Department supervision. The SSS
and Operations management recognized the potential effect of the packing
adjustments on control rod insertion time and operability, which
resulted in NMPC's decision to shut down the reactor. The reco5nition
by Operations was based, in part, on Standing Order No. 36 issued in
December 1985 that identified a listing of ASME valves whose stroke
times would be importtnt and potentially affected by maintenance such as
packing adjustments. Standing Order No. 36 was addressed to Operations
personnel but had not been provided to other Station personnel such as
QC or maintenance.
QC Inspection
The QC inspector assigned to cover WR No. 102775 observed a limited
portion of the work performed on August 21. Interviews with the
inspector and review of the associated RWP found that the QC inspector
observed approximately 30 minutes of work under WR No. 102775 (at the end
of the job) and that he concentrated on the ten HCUs identified in the WR
to verify that the loose scram inlet valve nuts were tightened.
. . _
.
-
37
.
.
The QC inspector became involved in the procurement of a properly quali-
fied replacement nut and participated in discussions with the mechanics
as to how to replace the nut. At one point, the QC inspector consulted
GE Vendor Manual GEI-92807A for the proper nut size, but he did not take
note of written instructions in the Manual regarding assembly of the
packing gland flange, nor did he note the specified tightening criteria
for the nuts.
Therefore, although tightening of loose nuts and packing is a relatively
simple evolution, the QA review of the work scope and limited QC
coverage of WR No. 102775 were ineffective in identifying the following:
--
The breakdown in work control on the job by working beyond the
approved scope of WR No -102775.
--
The potential effects of tightening packing on the CRD scram inlet
and outlet valves including the lack of specified post-maintenance
testing.
--
The lack of proper notification of licensed control room operators.
The inspector reviewed the QC Inspection Report (QCIR 86-1287) that
documented the retightening of packing nuts on the valves identified in
WR No. 102775. The QCIR listed attributes that were to be verified by
the actual inspection. However, while the first QC inspection attribute
was planned to verify that the nuts were tightened properly, the
attributes that followed were added after the QC inspection was
completed and the plant shutdown was in progress. Notations were added
such as "per Manual-hand tight" and that a wrench was used as opposed to
the hand- tightening recommended for reassembly in GE Vendor Manual
GEI-92807A. Although Nonconformance Report No.86-070 was issued by the
QA Department to address the activity beyond the approved scope of work
and the lack of post-maintenance testing, the QCIR and NCR were after
f
the fact and subsequent to direct QA involvement in the work under WR
No. 102775. The QCIR was inappropriately completed and is misleading
since, for example, later review of the document (without knowledge of
,
the event or interview of the QC inspector involved) would indicate that
'
QC was aware of the nut tightening criteria prior to the work and that
QC identified the nonconforming conditions.
NCR 86-070 was resolved based on successful scram time testing of all
129 cnntrol rods. Corrective Action Request (CAR) No. 86-2028 was
issued on August 25 to the Maintenance Division because of the
programmatic nature of the issues involved. A response to the CAR was
,
due by September 12, 1986, and not reviewed as a part of this
! inspection.
.
!
,
e <p - - - - - - - r r- - . - - w r-, --e-,,yw- , ~ e - +---, - -- ~e- - -,<--- - ,r-~~
p.
9
-
38
.
.
Reactor Shutdown
,
Upon discovery by the SSS of the work performed on the ten HCUs, Station
management evaluated the potential ef'ect of the packing adjustments on
control rod operability. Scram insertion times for the ten control rods
.
!
in question were concluded to be indeterminate at that time. Since the
Technical Specification Action Statement requires that the plant be in a
Hot Shutdown condition within 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> following the time at which
control rod insertion times are unknown, a potential violation of
Technical Specifications was identified by NMPC, necessitating the
shutdown.
A plant shutdown was begun at 1:00 p.m. on August 22 and all rods were
fully inserted by 7:28 p.m. that evening. An Unusual Event was declared
at 1:00 p.m. and the NRC was notified via the ENS. A phone call was
also placed to NRC Region I office by NMPC to describe the events that
led to the shutdown. At the time of the Region I conference call, NMPC
was unaware that all 129 HCUs (not just the ten HCUs specified on the WR
No. 102775) had experienced packing adjustments, and that both the scram
inlet and outlet valves had been worked (not just the CRD No. 126
valves).
Licensee Reviews and Immediate Corrective Actions
An investigation was begun on the afternoon of August 22, under
direction of the Site Maintenance Superintendent. The results were
documented in an August 23 memorandum to the General Superintendent of
Nuclear Generation. The results were presented to a special session of
the Station Operating Review Committee (SORC) on August 23.
NMPC's' immediate corrective actions consisted of a revision to the
administrative procedure for corrective maintenance, AP-5.0. The new
features of the revision included:
--
approval of the WR by senior licensed member of the Operations
Department
--
specification of appropriate post-maintenance testing prior to
Shift Supervisor approval
--
approval by the SSS (initialled on Line 26 of the WR) prior to
commencement of work
The inspector discussed the results of NMPC's investigation and special
SORC review of this event with the Superintendents of Maintenance and
Nuclear Generation. While the immediate actions to revise AP-5.0
addressed a primary and obvious root cause, the inspector stated that
other contributing factors were not apparently considered by NMPC, such
as training deficiencies and hardware questions. For example, the
inadequate direction given by first-line maintenance supervision, the
working beyond the scope of the approved job by an experienced Chief
e
.
a
. 39
.
Mechanic, and the lack of recognition of the effect of a packing adjust-
ment on a safety related air-operated valve suggest training
deficiencies.
No effort was apparent on NMPC's part to investigate the cause of the
missing nut on HCU No. 14-47, nor the cause of the loose nuts identified
on the other nine CRD No. 126 valves listed on WR No. 102775. The
inspector also found that the appropriate nut tightness was not
critically evaluated, beyond the hand-tight recommendation. Finally,
NMPC's investigation failed to identify or pursue certain facts associated
with the event, such as:
--
The extension of scope of work performed, which was already beyond
that approved on WR No. 102775, to the scram accumulators (No.106
and 107 valves) on an indeterminate number of HCUs.
--
The absence of actual stem or packing leakage observed from any CRD
No. 126 or 127 scram valve.
--
The extent to which the loose bolts were backed off of the gland
flange, or how the nuts came to be loose or missing.
--
The appropriate torque value for the scram inlet and outlet packing
gland flange nuts.
--
The modification of the stem coupling to allow a replacement nut to
be installed on the scram inlet valve for HCU No. 14-47.
Post-Maintenance Testing
The August 23 SORC meeting reviewed the revisions to AP-5.0 and made
Unit I startup contingent upon those revisions being placed in effect.
Scram insertion time testing was performed for all 129 control rods on
August 24 in accordance with Procedure No. N1-ST-RI. All control rods
met their average and maximum insertion times for various notched
positions as required by Technical Specifications, except for CRD No.
38-11. CRD No. 38-11 exceeded the maximum time to the 5% insertion
point (2.4 notches) by 32 milliseconds, and WR No. 102820 was authorized
to adjust the scram inlet and outlet valve actuator springs. The work
was witnessed by QA as documented on QCIR 86-1294, and subsequent '
acceptable scram insertion times for CRD 38-11 were obtained on August
25. The inspector noted that CRD No. 38-11 was not one of the ten HCUs
identified on WR No. 102775.
The inspector reviewed the WR and QCIR, discussed the work with
personnel involved, and found that appropriate procedures were
referenced, notifications and approvals received, and post-maintenance
testing conducted.
.
-
40
~
.
LER 86-26
An LER was submitted to the NRC by letter dated September 19, 1986, to
describe the shutdowr. required by Technical Specifications as a result
of the packing adjustments to the scram valves. The LER described the
potential safety consequences of overtightening the scram inlet and
outlet valve packing nuts. An inappropriate packing adjustment on a
scram outlet valve was evaluated to be a more significant risk since CRD
No. 127 valve failure to open would prevent control rod insertion.
Therefore, the LER concluded that the scram outlet valve adjustments
presented a potential (although stated as " highly improbable") for an
anticipated transient without scram. The same potential did not exist
for an inlet valve packing adjustment, but rather only a potential for
slow scram times.
The LER failed to address the common mode error associated with random
tightening of CRD valve packing gland nuts without authorization or
documentation committed by the maintenance mechanic. Specifically, the
LER ignored the breakdown in training afforded the Chief Mechanic, and
how the inadequate direction provided by first-line maintenance
supervision contributed to the error. The LER also failed to address
the role of QC in observing but yet failing to recognize the mechanic's
errors, as well as the lack of written instructions and utilization of
existing vendor technical manual information regarding torque requirements
for scram valve packing. Finally, corrective actions described in the LER
addressed only the revision to AP-5.0, making no mention of initiatives in
the areas of training, repair procedures or vendor technical information.
'
Effect of Work on Scram Times
The inspector reviewed the scram insertion time test results for the ten
control rods identified as having loose packing nuts. Times recorded
for the two most recent tests (June 5 and August 24, 1986) were
compared. The ten control rods met Technical Specification limits, and
no appreciable time differences were observed for any rod between the
two sets of data. The largest difference in 90% insertion times was
0.240 seconds for CR0 50-35. However, the slower insertion time for CR0
50-35 was still within acceptable limits and reasonable in light of the
number of cycles and other factors experienced during the 80 day period
between testing.
The comparison of mean 5% and 90% insertion times for the ten HCUs in
question and a comparison of those times (for the last two tests) versus
core average times and Technical Specification limits is as follows:
. 41
.
.
Mean Control Rod Scram Times
5% / 90% Insertion
(in seconds)
June 5 August 24
-Ten HCUs .357/2.89 .362/3.01
Average of All Rods .359/2.89 .365/2.99
Tech Spec Limits .375/5.00 .375/5.00
The inspector concluded that the maintenance performed under WR No.
102775 had no measurable effect upon control rod scram insertion times. e
Therefore, no deterioration of the CRD scram function actually occurred
because of the packing adjustments.
Technical Direction in Vendor Manuals
The inspector reviewed recommendations in the GE Vendor Manual
GEI-92807A for the reassembly of the CRD No. 126 and 127 scram valves.
The manual was not consulted or applied during the work under WR No.
102775, although a note existed on page 5-29 (step f) to hand-tighten
the gland flange nuts, which should be sufficient to prevent excessive
-
stem leakage. Additional tightening was recommended only enough to
prevent stem leakage. The inspector noted that this evolution was
intended for a valve with no pressure initially on the packing (i.e...
a rebuild), and that the directions from GEI-92807A had not been
properly transferred into Mechanical Maintenance Procedure No.
N1-MMP-6.4 for overhaul of the HCUs.
Based upon discussions with the GE Site Representative, the inspector '
concluded that more recent technical information was available onsite
with quantitative torque criteria for the CRD No. 126 and No. 127
valves. Operating and Maintenance Manual GEK-9582A contained more
specific and up-to-date technical information regarding maintenance on
the scram inlet and outlet valves. Section 5-59, Hammel-Dahl Scram
Valve Packing Replacement, describes assembly and lubrication of five
packing rings along with tightening of the gland flange nuts to an initial
value of five inch pounds of torque. A note cautions that, if stem
leakage is experienced, additional pressure up to 15 inch pounds is
acceptable and that torque beyond that value will require post-maintenance
stroke testing of the valve.
A vendor technical manual upgrade project instituted for the Maintenance
Department was begun in June 1986. However, the manual for the CRD HCus
has not yet been reviewed. Based on discussions with Maintenance super-
vision, the technical manual upgrade is an extensive project that will
take several years to complete. The inspector reviewed a list provided
by NMPC's contractor performing the project for manuals currently under
review and identified the initial number of manuals started to be
'
,
.
42
.
.
approximately 100, although the majority were related to electrical
maintenance associated with relays. The inspector noted that no defined
prioritization existed for those manuals receiving initial review and
incorporation into station procedures.
The inspector reviewed a memorandum dated August 26, 1986 from the GE
site' representative to the Supervisor of Technical Support that
addressed the significance of the loose packing nuts. The condition was
evaluated by GE San Jose engineering personnel. No similar BWR
experience with loose nuts on the scram inlet and outlet valves could be
identified. The theoretically loose packing (associated with loose or
missing gland flange nuts) was concluded to present less stem drag force
and result in faster valve stroke open time. NMPC's explanation of the
loose scram inlet valve nuts was a geometrical argument that in
diametrically tightening the four nuts the last nut to be tightened would
tend to loosen another nut.
The inspector concluded that sufficient technical information was
immediately available but never consulted prior to performing the work
under WR No. 102755. Since no packing leaks were evident, the loose
nuts could have been left as-is or appropriately hand-tightened without
affecting control rod operability. Had a more thorough investigation by
NMPC been able to confidently identify which scram inlet and outlet
valves had been tightened and approximately how much torque had been
applied by the mechanics, then the reactor shutdown may not have been
necessary if that torque had not exceeded 15 inch pounds. This was
conceivable since the torque.was applied to pressurized packing.
However, the qualitative direction given and the " snugging-up" described
by the Chief Mechanic probably would have exceeded the recommended
hand-tightness or the 5-15 inch pounds of specification GEK-9582A.
Also, the inspector determined independently that other utilities have
researched the quantitative bounds of " hand-tight" and have determined
these to be between 10-30 inch pounds of torque. Therefore, although
insufficient technical direction was provided to the mechanics (either
verbal, written or procedural instructions) to perform the CRD No. 126
valve packing adjustments, the decision to shut down the reactor was
conservative and appropriate.
Mechanical Maintenance Training
The inspector reviewed the training records and experience of the Chief
Mechanic and Assistant Supervisors involved in the work under WR No.
102775. All personnel had sufficient experience and training such that
the violations of work scope and procedural notifications should not
have occurred. Basic training in procedures, Technical Specifications,
BWR systems, and technical skills such as for packing and seal
maintenance had been received by the personnel involved during the last
3 years.
i
-.
r
-
43
.
.
,
On the other hand, basic concepts and administrative controls generally
applicable to routine performance of maintenance were not recognized or
followed in this case, such as:
--
Working within the approved scope of the WR.
.
--
The effect of packing adjustments on valve operability.
--
Provision of appropriate work instructions (commensurate with the
- . skills of a mechanic) via first-line supervision.
--
Performance of maintenance without a " mark-up".
--
The importance to reactor safety of the HCUs.
9
.
--
Clear notification of and approval by Shift Supervisor prior to
release of safety related equipment for maintenance.
--
Understanding of responsibilities defined in AP-5.0 for corrective
maintenance.
4
--
Appropriately defined post-maintenance testing.
The inspector concluded that, given the previous training and experience
of the individuals involved, in contrast with the above mentioned
concepts, a deficiency in the training provided to mechanics and first
line maintenance supervision exists.
The apparent deficiency lies more within the administrative controls and
operational interfaces involved in work rather than in the skills of the
mechanic to actually do the work. However, the packing nuts were tight-
ened without regard for the operability of the valve in spite of a lack
of technical tightness criteria and even though no packing leaks
actually existed. The inspector considered this conclusion to be of
significance, since NMPC relies upon work performed in many cases (as with
WR No. 102775) with no written procedures, because the activity is judged
to be within the skills normally possessed by qualified maintenance
personnel. This reliance shifts more burden upon the verbal instruction
'
provided by first-line supervision and therefore their experience, as well
as on the availability of clear technical reference material such as in
vendor manuals.
Existing Administrative Controls on Maintenance
As discussed in ANSI Standard N18.7-1972, Section 5.1.6, Administrative
Controls for Maintenance, to which Unit 1 is committed per Technical
i Specification 6.8.1, maintenance that affects the functioning of
safety-related systems should be performed in accordance with written
procedures, documented instructions or drawings appropriate to the
circumstances. The standard addresses the use of appropriate vendor
'
manual information or, in lieu of such, a suitably documented procedure
.-. - __ - - . - _ - - - _= .- ._. _- ,
. -
-. . - ___ _ .
o
.
44
.
.
that provides adequate instruction to assure the quality of the work.
-
The standard recognizes that, for skills normally possessed by qualified
maintenance personnel, a detailed step-by-step delineation in a written
procedure is not necessary (such as a relatively simple packing
adjustment). Nevertheless, the scram valve maintenance performed under
WR No. 102775 was not performed in a manner to assure the quality of the
CRD scram function, nor was a " suitable level of confidence" in the
scram insertion capability attained (as called for in ANSI N18.7) by
either quality inspections or post-maintenance testing. Moveover, ANSI
N18.7 sets forth standards associated with maintenance work controls
such as:
-
Documented permission for equipment release for work by operating
personnel, including length of time out of service.
-
Development of new or existing repair procedures, as experience is ,
gained in the operation of the plant, revising routine maintenance
practices to improve equipment performance and also added to the
compendium of already existing specific repair procedures.
-
Evaluations of the cause of failed equipment.
The administrative controls present in AP-5.0, Precedure for Repair,
were inadequate or in some cases, not followed in performance of the
work under WR No. 102775. The overall lack of control of that
maintenance is an apparent violation (50-220/86-17-06) of Technical
Specification 6.8.1 as it applies to the administrative controls
described in ANSI Standard N18.7-1972 and implemented in Procedure
Attitudes Towards Safety
The inspector noted a proper concern for personnel and reactor safety
among all individuals interviewed. However, a lack of critical
questioning and attention to detail was evident in the activities
.
associated with WR No. 102775. A tendency towards filling out the
paperwork after the fact (e.g., the QCIR and WR) versus a focus on the
potential technical ramifications of the packing adjustments was
observed. Also, the NMPC's investigation and SORC review focused on a
procedural revision and compliance with Technical Specifications while, at
least initially, apparently dismissing hardware and training deficiencies.
<
All personnel contacted were friendly and cooperative, and responded in a
timely and serious manner to the concerns raised by the inspector.
7.3 Conclusions on CR0 Maintenance Event
The tightening of packing on the CRD No. 126 and 127 scram valves was
found to have an insignificant effect upon the scram insertion times of
the control rods. The large spring opening force of the Hammel-Dahl
scram valves is appreciably greater than the stem drag force caused by
tightened packing, and it is doubtful that any change in rod insertion
,
, -, , . _ . . ,_- - . - _ _ , . . _ . ,, _ . , - . . ,y. -
_ _ - . _ , -___-,._-.7 -
m,.
(
f
-
. 45
.
.
times occurred. None was discernible. Because the packing adjustments
were hand-tightening of the nuts under pressure, no deterioration of
the CRD scram function was realized.
A fundamental breakdown in work controls occurred which allowed
activities to proceed beyond the approved scope of work and without
proper Shift Supervision review and approval. Training of the mechanics
and their first-line supervision was ineffective in that the importance
and significance of a basic packing adjustment weren't recognized.
Existing administrative controls for maintenance were ineffective and
were violated. The verbal instructions given by first-line supervision,
in lieu of written instructions or formalized procedures, were
insufficient and proved to be not within the skills and training provided
to the workers. Pertinent technical information was available but never
consulted or translated into appropriate directions to the workers, nor
was that information used after the fact to assess the significance of the
event. Although all control rods were found to be operable during sub-
sequent scram time testing, post-maintenance testing was not properly
considered and defined prior to the work on the WR because of a failure to
involve Operations personnel.
QA review of the scope and subsequent QC involvement and observation of
the work were insufficient in identifying the above mentioned deficien-
cies. Also, QC documentation after the work was performed was
inappropriate and misleading as evidenced by the completed QCIR. It
should be noted that QC inspectors did identify the original loose nut
conditions, and the QC inspector was the first individual to consult
available technical information associated with the scram inlet valves,
although not with the intent of defining proper packing gland nut
tightness.
When eventually informed of the extent of the work performed, Operations
personnel recognized the effect of the packing adjustment on control rod
operability. Upon discovery of the work performed, and its potential
effect on control rod operability, NMPC complied the Technical
Specification Action Statement regarding control rod drive systems. An
orderly reactor shutdown was accomplished within ten hours, an Unusual
Event was properly declared, and the NRC was notified in a timely manner
by NMPC management.
NMPC's immediate investigation and SORC review of the event failed to
uncover certain pertinent factual information and was considered to be
not sufficiently thorough or self-critical to identify potential root
causes beyond the obvious notification and approval of Operations
personnel. This suggests an overreliance on licensed operators to
control plant maintenance activities. Although the event was analyzed
in LER 86-026, no corrective actions have been proposed beyond the short
term revision to AP-5.0, such as in the areas of vendor technical
manuals, training and verbal / written direction to maintenance personnel.
.
t
. 46
'
F
.
8. PROGRAMMATIC ISSUES
In its August 15,1986 letter to the NRC, NMPC included an overview of
the factors that potentially contributed to the problems covered in the
allegations. NMPC concluded that the following four programmatic issues
were involved in the allegations and warranted further consideration.
1. Procedures - NMPC found a number of problems with procedures,
including some situations not covered by procedures, unclear QC
involvement, areas of ineffective procedural coverage, and a weak
procedure updating process.
2. Material Control - NMPC found that the control of loose LPRM and
IRM connectors was inadequate, and a general review of material
control would be undertaken.
3. Root Cause Analysis - NMPC found that the root cause analysis of
problems at Unit I needed to be improved based on the repetitive
maintenance on LPRMs and CRD pumps without correcting the problems. -
4. Management Effectiveness - NMPC concluded that supervisory
deficiencies, though not widespread, needed to be corrected.
In the August 18,1986 meeting between NMPC and NRC, the programmatic
issues received considerable discussion. Based on this discussion NMPC
expanded their review to the following programmatic areas.
1. Effectiveness of front line supervision
2. Informality of operations
3. Tendency for concerns to be kept at a low level
4. Quality Assurance involvement in programmatic issues
5. Ability to deal with hardware issues
6. Unit 2 effects on Unit 1
7. Repeat NRC inspection findings
8. Adequacy of root cause evaluations
9. Is management looking down for concerns
10. What level in organization are decisions made
11. Procedural issues, adequacy and informal practices
12. Radiation control
These topics were addressed by NMPC during a meeting with the inspection
team on August 28 and were covered in the NMPC letter to the NRC on
August 31, 1986.
. _ - _ _ _ _ _ - _ _ _ _ - - - _ _ - -_ - . _ _ _ _ - _ - _ . _ _ _ .
e
.
47
-
e
D
9. SUMMARY AND CONCLUSIONS
The special inspection was in response to allegations initially
presented to the NRC Unit I resident inspector on July 11, 1986 and
subsequently investigated and evaluated by NMPC as presented in the
August 18, 1986 meeting and in a summary report provided to Region I by
letter dated August 31, 1986.
The special inspection team independently assessed the scope of the
alleged technical issues and their broader programmatic implications for
continued operation of Unit I and the potential licensing and operation
of Unit 2. Although most of the allegations were found to be factually
correct, their safety implications were subsequently determined to be
minor and have been addressed by NMPC corrective actions.
The inspection reviewed the general effectiveness of the quality
assurance program and its ability to find and correct the problems
involved in the allegations. The review found that its capabilities
were very limited regarding the potential for discovery of these types of
problems. This review is documented in Section 5.
The inspection also incorporated the results of a Region I inves-
tigator's initial review of the NMPC internal investigation of the
allegations. The review concluded that the investigative staff had an
independent charter, free of undue management influence in their actions
and conclusions, and that the assigned staff were experienced and
credible. The NRC stated that NMPC should provide a final report of this
investigation for NRC review and notify the NRC of any safety significant
issues which the investigation might reveal. The NRC review is documented
in Section 6.
The inspection reviewed the CRD valve maintenance activities, the
resulting Unit 1 shutdown on August 22, and the NMPC internal
investigation of the events. In this review the inspection found a
fundamental breakdown in work controls, insufficient QC involvement, and
a flawed NMPC internal investigation. This review is documented in
Section 7.
Based on these reviews we concluded that there were certain programmatic
weaknesses evident in the NMPC management system that allowed these
issues to develop and in some instances spread, in that:
1. Methods within the organization to identify shortcomings and poten-
tial problems have not been effectively implemented. As a result,
problems identified by NMPC staff are not always brought to the
attention of management for resolution.
2. Once issues are identified, there are weaknesses in the NMPC review
methods and management oversight which in some cases effect the
ability to:
.-
t
. 48
e :
e
determine contributors to the problem or event;
identify the root causes; and
evaluate the impact on broad program effectiveness.
3. The NMPC Operational Quality Assurance (QA) program was not as
effective as it should be in helping the line organization to find
and correct problems.
The inspection team acknowledged the alleged harassment of the I&C
technician by his peers and supervisor for bringing these issues to NMPC
QA and to the NRC. As contained in an August 18, 1986 letter the NRC
recommended that these issues be presented to the U.S. Department of
Labor (DOL) by the alleger and therefore, further NRC action will be
dependent upon DOL action and NRC review of the final NMPC investigation
report.
In the course of this special inspection, the NRC became aware of
additional material that raised questions about the effectiveness of the
overall QA program at Unit 2. In order to review these issues a QA
audit team conducted an onsite followup inspection during the period of
September 8-12, 1986, and the results are documented in Attachment 5 to
this report. The team determined that there were no safety issues and
no unresolved hardware issues. However, similar to the above special
inspection the inspection concluded that some programmatic issues
existed, primarily due to organizational problems. A response to these
issues was requested in a November 19, 1986 letter.
The special inspection identified a number of apparent violations as
documented in Section 4 of this report. Additionally, Attachment 3,
Supplemental Report 50-220/86-16 contains proposed violations associated
with the radiological controls program. A related programmatic
inspection conducted September 10-12 and 15-19, 1986 identified a number
of indications of inadequate control of operations, surveillance,
maintenance and modification activities as discussed in Attachment 4 to
this report.
An enforcement conference will be scheduled to discuss these issues and
their programmatic implications.
r
. ATTACHMENT 1
'
- UNigDSTATES
cf 'o,,
NUCLEAR REGULAT!RY COMMISSION
. ! o
E REGION I
- . O, [ 631 PARK AVENUE
S,s , [ KING oF PRUSSIA, PENNSYLVANIA 194o6
.....
File No. RI-86-A-0080
Docket No. 50-410
jjgg
50-220
Niagara Mohawk Power Corporation
ATTN: Mr. C. V. Mangan
Senior Vice President
300 Erie Boulevard, West
Syracuse, New York 13202
Gentlemen:
>
Subject: Allegations by Nine Mile Point 1 Instrument and Control Technician
6
Enclosed is a summary of allegations made by a Nine Mile Point Unit 1 Instrument
and Control Technician about activities at Unit 1 expressed to our Resident
Inspector initially on July 11, 1986 and subsequently amplified in discussions
with our regional staff. We understand from the individual that he has informed
your staff of all but the last two concerns, items 13 and 14.
Based on discussions between our staff and you and your staff on August 6 and
7, 1986 at the Nine Mile Point site, we understand that your investigation of
these concerns is nearly complete. Please provide us with a written report of
the results of your investigation. This letter is being placed in the Unit 2
docket as well as the Unit I docket because these potentially significant
allegations could impact the schedule for Unit 2 11 censing.
Following your submittal of the report, we ask that you arrange to meet with
us in our Region I office as soon as possible to discuss the report. We
appreciate your cooperation.
Sincerely,
.
William F
N
ane, Director
-
Division of Reactor Projects
Enclosures: As stated
W
'
l
-
.
. een
.
a m
-
_
==_ _ __._. --- -
.
-
K :-
_ -
-.
-
__----
.
. _
.
_
_
.
6
k eO p
.
o: .
- 24 Niagara Mohawk 2
4~ Power Corporation 1i AUG 1986
cc w/o encl:
Connor & Wetterhahn
John W. Keib, Esquire
J. A. Perry, Vice President, Quality Assurance
W. Hansen, Manager of Quality Assurance
D. Quamme, NMP-2 Project Director
T. J. Perkins, General Superintendent ,
R. B. Abbott, Station Superintendent
T. E. Lempges, Vice President, Nuclear' Generation
T. Roman, Station Superintendent
J. Alrich, Supervisor, Operations
W. Drews, Technical Superintendent
Dirgetor, Power Division
Be fartment of Public Service, State of New York
Public Document Room (PDR)
Local Public Document Room (LPDR)
Nuclear Safety Information Center (NSIC) '
NRC Resident Inspector
- State of New York
-- bec w/o encl:--
Region I Docket Room (with concurrences)
Management Assistant, DRMA (w/o encl)
) DRP Section Chief
Region I SLO
Robert J. Bores, DRSS
.
W
- -
4
M=
..
e-=~ -e
- - - - -- - - -- -
,2 _ .
_. .
._ . _
_
_
.
n- .
_
. .
- .
,
__
- R ,
.
.
!
. - ,..
'
t
,
,
l
l
SUMMARY OF ALLEGATIONS
CRD Pump Vibration Testing
1. In March,1986, after weeks of daily vibration tests of the CRD pump,
testing was suspended when it was apparent that the increasing vibration
would exceed the action limit of the ASME requirements and a plant
shutdown would have been required prior to the scheduled March 8, 1986
.
shutdown.
1
Helium Leak Tests
N
2.A In March, 1986, the chemistry supervisor noted that errors existed in the
procedure for helium leak testing the stack gas system, in that portions
of the system would not be tested. The alleger found the supervisor's
conclusion to be correct. The I&C supervisor assigned the alleger to
review the leak testing procedure and propose changes to it. After
! completi,og this work, the I&C supervisor sat on the proposed changes and
! later told the alleger to do the testing with the old procedure. The
leak testing was done in April. _
3. The alleger was instructed to apply 100 psi air to seat the feedwater
< check valve af ter it had failed its initial, test. It failed the second
j test also. Then the eechanic installing the replacement valve told the
- alleger that the valve seat was hammered in place. The valve passed the
l 1eak test, but stuck shut during startup.
4. The shift supervisor diverted flow in the feedwater lines to free the
i
stuck feedwater check valve. There appeared to be no procedure for this
and no management review. Eventually, the valve opened.
l 5. During the outage non qualified technicians installed LPRM connectors in
that A techs were installing them without direct supervision from C techs.
6. During the outage and years prior LPRMs connectors were routinely
l
installed without proper Work Request (WR) paperwork, connectors
l
replacements were represented on WRs as troubleshooting, and the
t installation and test procedure, LPRM-1, was routinely not used or filled
I out afterward.
!
'
7. Since the cable replacement six years ago the LPRM cables have not fit
properly into the connectors. The cable 41 electrics have been melted
smaller (per LPRM-1) or the connector bores have been drilled larger to
fit them together.
) -
- -
..
l - -
_ = ..
-
.
_
- . _.
~
y- .
--
_
,
-
-
,
.
..
[ - --
.
-
~
_
_
.
n]
-.
_
_ _
49
L -
. *
.
.
,
i~
.
a,
)
8. QC involvement in the LPRM connector work was improper in that I&C techs
frequently did not inform QC that connectors were being replaced, and
even when aware of the connector replacements, QC inspected only paper
and never went under the vessel because they knew the work was
unacceptable to specifications.
9. On July 10 a different design connector was installed on some LPRMs
(prior to being discovered by the resident inspector), and no design
change had been submitted for it. In addition, no work requests or LFRM
maintenance procedures were prepared until after the resident inspector
came down to witness this activity at which time the workers involved
took a break to generate the paperwork and get it approved by the shift
supervisor. -
9
10. 5During the outage the alleger was harassed by fellow workers and
discriminated against by his supervision due to his raising concerns
.
about the LPRM connector work. The supervisors did little or nothing to
correct his harassnent.
11. The connector on IRM 18 was replaced on June 7,1986, and%as not
documented on the WR.
) * -
12. The plant Was started up on the morning of June 17,198Qased on
falsified surveillance test records for the replaced IRM connector. The
I&C tech's and assistant supervisor falsified the test record without
performing any of the required surveillance testing.
Other
13. An I&C technician working on LPRM connectors received a dose of 1.25 REM
which was in excess of his administrative limit.
I 14. A piece of an aluminum tool about 1 inch by 8 inches was lost in the
reactor vessel during the outage. The tool was used for installation
and removal of feedwater line plugs.
.
.
)
-- - .
W
me
H, 6
. ===-
'
.._ ~~
~~ ~- ~ ~ - * - .
- _.
- - - . ._ _. _
_ __
.
-
g