ML20137W537
Jump to navigation
Jump to search
| ML20137W537 | |
| Person / Time | |
|---|---|
| Site: | Three Mile Island  |
| Issue date: | 02/12/1986 |
| From: | Conte R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20137W515 | List: |
| References | |
| 50-289-85-30, NUDOCS 8602200204 | |
| Download: ML20137W537 (89) | |
See also: IR 05000289/1985030
Text
{{#Wiki_filter:U.S. NUCLEAR REGULATORY COMMISSION
REGION I
Report No. 50-289/S5-30
Docket No. 50-289
License No. DPR-50 Priority -- Category C
Licensee: GPU Nuclear Corporation
Post Office Box 480
Middletown, Pennsylvania 17057
Facility At: Three Mile Island Nuclear Station, Unit No. 1
Inspection At: Middletown, Pennsylvania
Inspection Conducted: December 13 u 1935 - January 10, 1986
Inspectors: W. Baunack, Project Engineer, Region I
N. Blumoerg, Lead Reactor Engineer, Region I
L. Briggs, Lead Reactor Engineer, Region I
J. Bryant, Senior Resident Inspector (0conee),
Region II
P. Eapen, Chief, Quality Assurance Section, Region I
D. Haverkamp, Technical Assistant for TMI-1 Restart,
Region I
W. Oliveira, Reactor Engineer, Region I
D. Tricole, Resicent Inspector (Calvert Clif R)
Region I
R. Urban, Reactor Engineer, Region I
A. Weacock, Radiation Specialist, Region I
P. Wen, Reactor Engineer, Region I
J. White, Senior Radiation Specialist, Region I
F. Young, Resident Inspector (TMI-1), Region I
Other NRC Personnel: J. Thoma, Project Manager, Office of Nuclear Reactor
Regulation
Contractor Personnel: B. Gore, Senior Research Scientist, Battelle Pacific
Northwest Laboratories
Approved By: h J
O R. Conte (A M) _-1 Restart Manager
2/fkd
Date
TMI-1 Restart Staff
Division of Reactor Projects
Inspection Summary:
The TMI-1 Restart Staff conducted routine and special (NRC shift coverage)
safety inspections (838 hours) of power operation focusing on operator and
!
s G
t
.- ._ - _ - . - . __
2
:
management performance. Specifically, items reviewed in more detail in the
facility operations area were: control of fire doors; power operated relief
valve test results; pressure switch malfunction for makeup pump lube oil
sub-system, diesel generator standby coolant system; letdown cooler leakage;
out-of-specification log entries; instrumentation monitoring; and annunciator
black board conditions. Special focus continued on the ccmpletion of the
power escalation testing program, the radiation protection program as applied
to power escalation, and the non-licensed training program. Other review
: items included: reactor trip event and post-trip review; radiological release
i
events; emergency feedwater pump operability; licensee readiness for continued
power operations; steam generator cleaning techniqua; and licensee action on
previous inspection findings.
Inspection Results:
,
Licensee management and the quality assurance department continued their
presence and involvement in daily activities. Problems were noted in the
thoroughness and completeness of management and technical support personnel
review of plant problems. The problem of workers in the spaces having the
potential to cause safety system challenges persisted. Apparently the
licensee's corrective actions are not yet understood by all workers. The
licensee continued to be aggressive in the maintenance of safety-related
equipment. Operators continued to conduct themselves in a competent manner and their actions reflected a strong performance-oriented training program. Accordingly,
, operators were responsive to radiological esents and plant trips that occurred
during this period. A poor operational practice was noted of "living with" out-of-specification readings; although operator log-keeping actions were
, appropriate, it appeared that mid-level managers were unresponsive in taking
corrective actions. No performance problems were identified in the review of
non-licensed operator training. Overall, procedures were properly implemented,
but individual procedure step inadequacies continued to be observed. The
labeling on fire doors was confusing. An improperly coordinated surveillance
procedure change with respect to daily functional checks of fire doors resulted
in an apparent violation of technical specifications (paragraph 3.2.2).
The startup test program was effective in identifying system interface
problems and in providing training for plant operators along with carfirming
proper plant performance.
The Radiation Protection Program remained strong in support of licensee work
activities during the power operation mode. Personnel inattentiveness was the
cause for radiation area signs being left down; and, as a result in one case,
an individual entered a high radiation area without knowing it. No overexposure
occurred.
The licensee was responsive and took appropriate corrective action for NRC staff
findings from previous inspections.
The return to commercial power operations was conducted in a deliberate and
conservative manner.
l
._. _-- _ - - - . - -
-, . - - - , _ . . _ _ , - - . , , ,- . _ _ _ . _ - . _ -,,,_, , _ _ -
_ _ . _ . _ _ . . _ _ . - .. ..
,
t
3
DETAILS
1. Introduction and Overview
1.1 General
>
At the beginning of this inspection period on December 13, 1985, the
TMI-1 Restart Staff was providing on-shift inspection coverage 12
hours a day to assess restart operating activities. At 3:00 a.m. on
December 27, 1935, continuous coverage was initiated for the power
escalation to the maximum achievable power testing plateau, and at
'
6:00 p.m. on that day this inspection coverage was returned to 12
hours a day consistent with the reduced level of testing activity and
steady-state facility operation.
I
On both Christmas and New Year's Day periodic inspection coverage was
provided by the resident inspectors. At 7:00 a.m. on January 2, 1986,
continuous coverage was again initiated for a planned reactor trip
$.
test from maximum achievable power and the subsequent plant restart.
Shift coverage was terminated at 7:00 a.m. on January 5, 1986, follow-
ing the return to steady-state plant operation at 90% of rated power,
marking the completion of the restart test program. As of 8:00 a.m.
on January 10, 1986, the TMI-1 Restart Staff organization was secured
and TMI-1 inspection responsibility was returned to the normal Region I
organization alignment.
! The staff's on-shift observation of plant activities was maintained
by NRC personnel from Regions I and II and by a reactor operator ex-
aminer from an NRC contractor. Also, Region I inspectors conducted a
review of non-licensed personnel training and continued periodic
coverage of testing and radiation protection activities. Additional
i Region I personnel were on site during portions of the period to
augment the resident inspection staff.
1.2 Facility Restart Operations
During the period of December 13, 1985 to January 10, 1986, the sig-
nificant TMI-1 restart operational milestones included: (1) power
! escalation and main turbine generator operation and turbine runback
test at the maximum achievable power testing plateau, (2) turbine
trip / reactor trip from maximum achievable power, and (3) return to
j power operation, marking the completion of the restart test program.
Subsequently, the plant was able to achieve operation at 100% of rat-
ed power and additional full power test data were taken. The chrono-
j logical summary of plant operations during the period is presented l below.
1
I
.
---_
,e --- _ _ _ - y-. , . . _ _ , _ _ - - . _ _ _ . - - . - . - - - - - ~ . . - - - .
-
. . - . . - - - - - - _ . - . --. _ _ - - -
- _ _ - . . . . . - . - - _ _
4
Date Time Operational Highlight or Milestone
12/13/85 7:00 a.m. Reactor at 75% of rated power
12/27/85 4:16 a.m. Region I Administrator authorized the
'
; licensee to escalate power to the maxi-
mum achievable power plateau
4:20 a.m. - Increased power to 87.5% at 3% per
8:55 a.m. hour; power escalation stopped when
approaching the once through steam gen-
erator operating level limits
'
1:37 p.m. -- Conducted loss of main feedwater pump
1:50 p.m. runback test from 87.5% of rated power
and returned power to 87.5%
1/2/86 12:04 p.m. Conducted turbine trip / reactor trip
test from 88% of rated power
1:10 p.m. Tripped reactor coolant pumps for
emergency feedwater system initiation
testing
1:17 p.m. - Reactor coolant pumps restarted,
1:37 p.m. providing forced circulation flow
1/3/86 10:28 p.m. Region I Administrator authorized the
licensee to return to power operations
10:29 p.m. Commenced control rod withdrawal for
reactor startup
10:55 p.m. Reactor critical
1/4/86 3:26 a.m. Main turbine generator connected to
regional electrical grid
3:43 a.m. Reactor trip due to turbine trip caused
by high level in moisture separator IC
11:09 a.m. Commenced control rod withdrawal for
reactor startup
11:56 a.m. Reactor critical
2:15 p.m. Main turbine generator connected to
regional electrical grid
10:00 p.m. Reactor at 90% of rated power
5
Date Time Operational Highlight or Milestone
1/5/86 7:00 a.m. Reactor at 91% of rated power, secured
on-shift inspection coverage
1/6/86 2:05 p.m. Reactor at 100% of rated power
1/10/86 4:00 a.m. Reactor power reduced to 75% in order
to take a main feedwater pump out of
service to assess high pump vibrations
1.3 Operational Events
Several events or other matters occurred during this inspection peri-
od that were considered either operationally significant or of spe-
cial interest to the TMI-1 Restart Staff. These items are discussed
below.
On December 17, 1935, a contamination event occurred when noble gas
was released into the auxiliary building atmosphere during replace-
ment of a waste gas compressor filter. All workers leaving the
building were frisked by health physics technicians. Seven workers
and one region-based NRC inspector were slightly contaminated. In
the case of the NRC inspector, only his hard hat was contaminated.
The licensee estimated personnel exposures of about 4E-4 millirem
from the contamination.
On December 30, 1955 at 1:50 a.m., the licensee declared an unusual
event condition due to a release of noble gas monitored by the plant
effluent monitor. The unplanned release was due to failure of a
makeup pump (high pressure injection pump) shaft seal. Although the
plant effluent monitor went below the alert level at approximately
2:30 a.m., the licensee decided to stay in the unusual event condition
until the affected pump was drained and isolated. The licensee
secured from the unusual event at 4:20 a.m.
During power ascension on January 5, 1986, following the reactor
trips that occurred on January 2 and 4,1986, steam generator level
versus reactor power tracked lower than during the power escalation
program in which maximum achievable power was 88%, due to high opera-
;ing water level in the steam generators. The licensee attributed
the change to a redeposition of fouling material in the steam gener-
ators. Consequently, the plant was able to achieve full power opera-
' tion on January 6,1986, with steam generator water levels well
within operational limits. Subsequently, the licensee conducted full
power tests that were deferred because of the previous power
limitations. Startup testing was completed on January 9, 1986.
Reactor physics testing was completed by January 16, 1986.
,
6
1.4 Summary
This inspection included restart testing activities for portions of
the 75'; and maximum achievable power plateaus, and subsequent plant
operations at 100*o of rated power. During this period there were no
interruptions of the restart testing program due to equipment prob-
lems or for other reasons. However, an unplanned reactor trip de-
layed the plant startup, following a planned reactor trip from
maximum achievable power, for several hours. The shift inspectors
referred only implementation matters or status questions to shift
supervisory personnel and referred programmatic matters (event
followup, design or procedure adequacy problems) to resident and
region-based NRC personnel. Resident and region-based personnel in-
terfaced with licensee support groups in followup to shift inspector
referrals / concerns. The staff's observations and findings regarding
plant operations and testing and licensee response to operational
events is discussed in the report sections that follow.
2. Shift Insoection Activities
2.1 Scope of Review and Observations
During the period of December 13, 1935 - January 5, 1986, the TMI-1
Restart Staff continued its augmented shift inspection coverage. The
NRC shift inspectors assessed the adequacy and effectiveness of
operating personnel performance based on the inspectors' observations
of operating and startup activities to determine that:
--
operators are attentive and responsive to plant parameters
and conditions;
--
plant evolutions and testing are planned and properly
authorized;
--
procedures are used and folicwed as required by plant
policy;
--
equipment status changes are appropriately documented and
communicated to appropriate shift personnel;
--
the operating conditions of plant equipment are effectively
monitored and appropriate corrective action is initiated
when required;
--
backup instrumentation, measurements, and readings are used
as appropriate when normal instrumentation is found to be
defective or out of tolerance;
_ _ . _ _ _ -
l I
7
--
logkeeping is timely, accurate, and adequately reflects
plant activities and status;
--
operators follow good operating practices in conducting
plant operations; and
l
--
operator actions are consistent with performance-oriented {
training.
l
The shift inspectors' observations included, but were not limited to,
those reactor plant operation and startup testing activities, period-
ic surveillance and maintenance testing activities, and preventive
and corrective maintenance activities listed below.
Reactor Plant Operation and Startup Testing Activities
--
routine control room operations including annunciator alarm
response and control room logkeeping
--
operating and emergency procedures discussions with shift
supervisors, shift foremen, control room operators and
shift technical advisors
--
periodic inspection observation tours of areas outside the
control room, including diesel generator rooms, emergency
feedwater rooms, control building, turbine building, auxil-
iary building, intermediate building, electrical switchgear
rooms, and outside buildings and yard areas
--
routine operations and maintenance planning briefings be-
tween plant operations manager and shift supervisors
--
preparations for underwater inspection of a river water
pump to verify proper lubrication connections
( ( --
establishment and maintenance of cold weather protection
! controls
--
operating crew equipment tagging and lineup preparations
for turbine-driven emergency feedwater pump 1A controller
replacement
--
special shift crew briefings by the Vice President and
Director of TMI-l regarding past performance and expected
operating practices
--
shift crew actions in response to purge valve AH-V-113
found to be leaking greater than prescribed limits during
performance of surveillance procedure 1303-11.180
- _ _ - _ _ _ _ _
_ - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___ __ ______ __ _
8
--
operating crew response to spurious automatic start of
makeup pump 1A auxiliary oil pump
--
shift crew assessment and critique of contamination inci-
dent that occurred during waste gas system maintenance on l
December 18, 1985
--
operating personnel actions performed in makeup pump rooms
during startup of makeup pump 1B and reciring of makeup
pump 1A
-
operating crew changeover of letdown. coolers to investigate
possible source of primary water leakage into intermediate
closed cooling water system
--
operating crew preparations for post maintenance leak test-
ing of neutralizer pump 1B seal replacement, performed in
May 1985
--
operator response to control room annunciator alarms for
" rods symmetrical" and " inhibit rod out motion" received
twice in close succession
--
operating crew preparations for mechanical modification of
letdown radiation monitor RM-L-1 flow controller
--
operating shift response to failed river water inlet tem-
perature detector, assumed caused by river ice buildup
tearing detector away
,
--
operating crew performance of power escalation from 75% to
'
87.5% of rated power at 3% per minute
--
operating crew and test personnel briefings for and perfor-
mance of integrated control system tuning tests at 87.5% of
rated power
--
operating crew and test personnel performance of turbine
runback test on loss of main feedwater pump
--
power escalation to 90% of rated power when average coolant
temperature was raised 5 degrees F for physics testing
--
shif t crew and test personnel briefings for and conduct of
planned turbine trip / reactor trip on January 2, 1986
--
control room operator withdrawal of shutdown group rods
--
operating crew response to overheating pump for the
hydrogen / oxygen analyzer
- _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ __
-
9
--
operating crew preparations for plant startup following the
'
planned reactor trip
--
reference leg filling for OTSG 1A level transmitter for
LT-3 and LT-5 l
--
control room operator withdrawal of safety group rods
,
--
turbine generator startup
r
--
operating crew response to unplanned turbine trip / reactor
.
-
trip on January 4, 1986
r
--
operating and engineering staff post-trip review of un-
planned reactor trip
--
operating crew performance of reactor and turbine generator
startup following unplanned reactor trip on January 4,1986
I --
power escalation to 91% of rated power
Periodic Surveillance and Maintenance Testing Activities
--
main feedwater pump testing per operations surveillance
procedure OPS-5101
--
emergency feedwater pump 1A controller replacement
post-maintenance testing
--
portions of reactor protection system surveillance
activities
--
portions of control rod movement tests per surveillance
procedure 1303-3.1
--
pressurizer power operated relief valve monthly surveil-
lance testing per surveillance procedure 1303-11.45
--
portions of high and low pressure injection analog channels
quarterly testing per surveillance procedure 1303-4.11
--
portions of area radiation monitors monthly calibration
tests per surveillance procedure 1303-4.15C
--
fire system testing and flushing of deluge / sprinkler system
per surveillance procedure 1303-12.13
--
main feedwater pump flow instrument calibration to trouble-
shoot different transmitter indications at same feedsater
flow location, per job ticket CI403, dated December 20,
1985
_ _ _ _ _ _ _ _ _ _ _ _
.
10
--
post maintenance testing of new controller for
turbine-driven emergency feedwater pump 1A, using surveil-
lance procedure 1300-3G
--
portions of power range nuclear instrumentation calibration
per surveillance procedure 1302-1.1
--
portions of emergency feedwater pump flow testing per sur-
veillance procedure 1300-3G A/B and adjustments of valve
controller steam inlet pressure to various settings to as-
sure minimum pressure and flow design requirements are met
--
decay heat river water system weekly surveillance testing
--
special operability testing of letdown isolation valve
MU-V-3
--
control rod patch panel verification per surveillance pro-
cedure 1301-9.2
--
letdown isolation valve time delay relay testing
--
emergency feedwater pump 2A testing per surveillance proce-
dure 1300-3F
Preventive and Corrective Maintenance Activities
--
troubleshooting and repair of once through steam generator
IA level transmitters LT-3 and LT-5
--
turbine-driven emergency feedwater pump 1A controller
replacement
--
troubleshooting and investigation of makeup pump 1B lube
oil pressure switches
--
temporary repairs to leaks in secondary system valves and
an emergency feedwater riser flange using Furmanite sealant
--
repair of soldered hot spots in control room drive cabinet
per job ticket CI-231
In addition, shift inspectors conducted or contributed to the follow-
ing special reviews of facility design and operational matters or of
the licensee's administrative controls programs.
--
closure controls for fire doors
--
hydrolazing (water spraying) of bu.lding spray / decay heat
vaults to reduce contamination and radiation levels
--~- - -- ---
sMg m , . - - , ,-+r- w
_ _ .
11
--
partial valve lineup verification for emergency feedwater
pumps
--
control room computer alarm priority system
--
operating crew response to out of tolerance secondary plant
equipment readings, while operating at 75% of rated power
--
use of scaffolding for planned maintenance activities, such
as emergency feedwater pump 1A controller replacement
--
worker use of piping and valves as footholds / support in
, lieu of ladders or scaffolding
--
shift crew trending of reactor coolant system leakage
--
barrier control practices for entrances to high radiation
areas
--
effects of vent header pressure changes on reactor coolant
system leakage measurements
--
trending of pressurizer power operated relief valve
setpoint check monthly surveillance procedure test results
for tests conducted since June 1985
--
operating staff trending and followup to out-of-specification
readings for emergency diesel generators jacket cooling water
temperatures
--
storage controls of portable equipment in safety-related
spaces
--
makeup pump 1A inboard seal failure event that occurred on
December 29, 1985
--
nitrogen blowdown of OTSG 1A level transmitter root valve
FW-V-1093 to free blocked line caused by Furmanite repair
2.2 Assessments of Shift Inspectors
2.2.1 General
The shift inspectors assured that any potentially adverse
i
safety concern or regulatory finding was identified prompt-
ly to both appropriate licensee supervisors and the NRC's
TMI-1 Restart Manager. Those items requiring additional
staff review or followup are described in paragraph 3 of
this report. Also, at the end of their assigned period of
shift inspection activities, the inspectors provided their
general assessment of facility operational readiness and
._ _
___ . . , . _ _ _
. . _ _ _ _ _ _ _- _ _ .
12
l
personnel performance. As noted below, these general find-
ings included, as applicable, the collective inspector
views related to operating staff performance, maintenance,
surveillance, radiological controls, training, emergency
planning, and physical security. The overall conclusions
of the TMI-1 Restart Staff are provided in the sections
that follow.
2.2.2 Operator Performance
Licensed operators continued to demonstrate their detailed
knowledge of facility design and daily plant status. This
was evident in their routine actions on shift and in re-
sponse to alarms and changing plant conditions. It was
-
further evident in their overall good performance in re-
sponse to unexpected events such as the unusual event of
December 30, 1985, and for the unplanned reactor trip on
January 4, 1986. The shift organization in the control
room continued to demonstrate discipline and overall good
noise control in the control room. These same positive
attributes were evident during the off-normal situations
such as during the unplanned reactor trip.
Shift inspectors continued to note that, at the end of
shift, shif t foreman (SF) routinely cocied the control
room operators log entries for routine activities. Although
the shift foremen were diligent in independently recording
entry in to technical specification action statements and
other off-normal events required by administrative procedures,
for routine entries the SF log lost its independence. This
could impact post-event reviews, since routine entries could
be used in reconstructing events. The Manager of Plant Opera-
tions acknowledged the inspector's observations and he
indicated that he would review the matter. This will be
routinely followed by the resident inspectors in future
inspections.
Of particular focus during this inspection period was li-
censed operator performance during the turbine trip test of
, January 2,1986. Overall, crew performance was very good.
All members were disciplined and attentive. They main-
tained their stations, observed appropriate instrumenta-
tion, reported important readings and alarms, and made
appropriate control manipulations. They were directed for-
mally by the Manager of Plant Operations, who kept them
apprised of overall plant status, concerns, and impending
actions. Despite the presence of many operators and ob-
servers in the control room, order was maintained and the
noise level was acceptable.
.
w w- ---v es- --ec - = - - - -, --- --r----m-- +-y-
i
13 ,
!
The test was conducted according to procedure. Prior to
- the test, a clear, concise briefing was conducted by the
Manager of Plant Operations and a tast engineer. Prerequi-
! sites, limits and precautious and procedural steps were
read and discussed appropriately. Individual operator
responsibilities were clearly assigned. It should be not-
ed, however, that TP 800/9, STR-1 had been completely re-
written during the week before the test. STR-2 was not
approved formally until December 31; and, three hours be-
fore test time, the control room had not yet been provided
,
'
with a controlled copy. Fifteen minutes before initiation
of the test, operators reccmmended a change to an individual
step, which was incorporated as a test exception by the
engineer in charge of the test procedure. During the test,
the Quality Assurance (QA) observer independently cautioned
operators to ensure that this change was not overlooked.
Time pressures clearly affected this process but no adterse
safety conditions resulted.
The potential for operator error could have been reduced by
more timely provision of the controlled copy of the proce-
I
dure and by formal incorporation of changes identified
prior to test initiation.
Following the reactor trip, immediate actions of ATP 1210-1
were clearly read by the shif t foreman. He waited for re-
sponses to action or verification requirements before pro-
ceeding to succeeding items. Formal communications during
l the rapidly evolving post-trip situation were good. Immedi-
j ate actions were promptly carried out as evidenced by
starting of the second makeup pump four seconds after
reactor trip. The failure of MU-V3 to isolate letdown was
promptly detected. Since the pressurizer outsurge due to
post-trip cooling was stopped at an acceptable level, the
decision not to marmally intervene immediately was demon-
- strated to be acceptable.
i Annunciators were observed before they were silenced. Not
all of them were formally announced, but there was no indi-
cation that other than nuisance alarms were dismissed with-
out announcements. Overriding of the nuisance alarm for
reactor coolant system " Loop Flow High", which repeatedly
, entered, then clear,d, helped minimize noise levels and
distractions.
Data from the Bailey computer sequence monitor (acquired
for post-trip review purposes) were lost. There was no
,
indication of operator error in computer use during the
'
test.
- _ _
. _ _ . . . _ . _ _ - - _ .
1
14
Following the reactor coolant pump trip, equipment problems
prevented satisfactory demonstration of emergency feedwater
system performance and resulted in EF-P1 being declared out
of service. Operator actions, however, were appropriate
during conduct of the test. The decision to terminate
testing and restart the RC pumps to prevent further steam
pressure decrease and to control cooldown was appropriate.
Non-licensed personnel appropriately conducted themselves
to support licensed operators during routine and special
evolutions or testing. No adverse trends in performance
were detected.
Overall, personnel conducted themselves in a competent
manner.
2.2.3 Training
Shift inspectors made no specific observations on the
licensee's training program. However, based on past TMI-1
Restart Staff reviews into this area, previous conclusions
on the training program remain unchanged. Operator perfor-
mance continued to reflect a strong training program being
in place for providing performance-oriented training. Spe-
cifics on the non-licensed training program are in para-
graph 8 based on Region I inspector review.
2.2.4 Maintenance
For the maintenance activities selected for review, person-
nel were knowledgeable in the technical aspects of the work.
In general, their duties and responsibilities were understood.
However, there continued to be evident less than full apprecia-
tion of how poor work habits could challenge a safety system.
Inspectors continued to note potentially adverse safety
actions such as stepping on instrument air lines for climbing
in the vicinity of MS-V13A/B, the initiation / isolation
valves f,r steam admission to the turbine driven emergency
feedwater pump. Further, the operator for MS-V10A, a bypass
valve around MS-V13A, was stepped on for climbing. This
occurred despite the fact that applicable job tickets were
planned well in advance to provide temporary support or ,
ladders for these activities. Licensee management recognized
such poor practices as a problem for which they initiated
corrective actions. Apparently they were ineffective, to
date, in reaching all personnel to explain the shortcomings
of such poor work habits. This matter will continue to be
routinely reviewed in future inspections by the resident
inspectors.
15
The troubleshooting for the abnormal performance of the
turbine driven emergency feedwater was conducted between
January 2 and 4, 1986. A job ticket and surveillance pro-
cedure was used to operate the pump on recirculation while
it was declared out of service. However, no formal data
were taken since the pump startups were termed by licensee
representatives as a part of the troubleshooting effort for
the pressure regulating valve. Subsequent to proper pres-
sure regulator adjustment, a satisfactory test was con-
ducted and recorded. However, the inspector noted to
licensee management that informal aspects of the trouble-
shooting efforts was a poor practice and could lead to a
lack of quality in the overall repair effort if errors were
made. No conditions adverse to regulatory requirements
were noted but the conduct of troubleshooting maintenance
will continue to be routinely reviewed by the resident
inspectors.
The Furmanite repair to a leak in a steam generator level
sensing line, to a certain extent, lacked good engineering
support of the reinjection process. Apparently during their
planning effort, licensee representatives did not consider
the potential for plugging the instrument line due to the
injection of too much Furmanite (a sealing compound). As a
result the line became plugged and the Furmanite plug had
to be blown into the steam generator. The licensee wanted
to restart the plant without a complete understanding of
the effects of the Furmanite material in the steam generator.
After probing by NRC staff and subsequent verbal discussions
among licensee engineers, licensee management concluded l
that there were no unreviewed safety questions to plant
restart. Subsequent to the startup, licensee personnel
documented their basis for no unreviewed safety question in
a 10 CFR 50.59 evaluation. This problem was indicative of
poor technical support for this work item. This item is
unresolved pending NRC staff review of the licensee's eval-
uation (289/85-30-01).
Although approximately 50 work items were completed by the
maintenance department between January 2 and 4, 1986, the
above examples were symptomatic of a sense of urgency that
could have the potential to result in error. Although no
adverse safety condition resulted, this is an area that
will be routinely reviewed by the resident inspectors.
2.2.5 Surveillance
Surveillance activities were performed by both operations
and maintenance personnel. The activities observed were
handled in a professional manner. All parties involved
were knowledgeable of the various systems. Procedural ad-
herence was strictly observed.
h,me,. _ _ , __ _ _ _ _ _ , , _ _ . _ _ , . _ _ _ _ _ _ _ . _ _ _ _ _ ,, ___ , . _ _ _ _ _ _ _ _ _ _ _ _ _ _
16
2.2.6 Fire Protection
During inspection of the safety related areas, shif t in-
spectors continued to observe labeled fire doors " ajar"
(not fully closed, leaning on the latch mechanism). It
appeared that the doors had a problem with their closing
mechanisms and that personnel were not attentive enough to
assure proper closing of the doors. See paragraph 3.2.2 for
details on the TMI-1 Restart Staff followup review.
No adverse conditions to fire safety were noted.
2.2.7 Radiological Controls
During this period the licensee culminated efforts to de-
contaminate the decay heat and building spray vaults to
permit access sufficient for observations without protec-
the clothing. This was conducive to enhancing management
presence in all areas of the plant.
On December 22, 1935, a radiological controls area sign was
noted to be down and not observable. The group radiologi-
cal controls supervisor (GRCS) took corrective action but
this was a precursor of a more significant event. On Janu-
ary 9,1986, an individual walked into a high radiation
area without proper dose rate equipment. See paragraph 6
for details.
Overall, proper radiological control practices were
implemented.
2.2.8 Physical Security
No conditions adverse to plant physical security were
observed.
2.3 _Cqncl u si on
Operators continued to conduct themselves in a competent manner and
their actions reflected a strong performance-oriented training pro-
gram. Preparation for the turbine trip test was conducted at a fast
pace but the pretest briefing had an organizing effect such that no
adverse regulatory or s:fety conditions resulted. In general, opera-
tors were responsive to daily plant problems as they arose,
i
The maintenance department conducted an ambitious work list in a
short period for the outage after the turbine trip test. That pace
coupled with poor technical support on certain work items caused some
F
17
errors. Workers continue to exhibit poor practices in certain in-
stances in which their actions could have an affect on the primary
plant. Apparently licensee corrective actions to date where
incomplete.
Overall, procedures were properly implemented. Radiological control
practices were good, however, there was a lapse in personnel atten-
tiveness to reposting a high radiation area sign.
3. Plant Operations
3.1 Scope of Review
TMI-1 Restart Staff inspectors periodically inspected the facility to
determine the licensee's compliance with the general operating re-
quirements of Section 6 of the Technical Specifications (TS) in the
following areas:
--
review of selected plant parameters for abnormal trends
--
plant status from a maintenance / modification viewpoint in-
ciuding plant housekeeping and fire protection measures
--
control of ongoing and special evolutions, including con-
trol room personnel awareness of these evolutions
--
control of doc nents including logkeeping practices
--
implementation of radiological controls
--
implementation of the security plan including access con-
trol, boundary integrity, and badging practices
The inspectors also focused their attention on the areas listed
below.
--
control room operations during regular and backshift hours,
including frequent observation of activities in progfess,
and periodic reviews of selected sections of the shift
foreman's log and control room operator's log and other
control room daily logs
--
followup items identified by shif t inspector activities
(see paragraph 2)
--
areas outside the control r:en
--
selected licensee planning meetings
As a result of this review, the inspectors reviewed specific events
in more detail as described in the sections that follow.
- _ __. . _ _ _ _ . . - . ._ .- .- __ . _ _ _ _ _ . .--
i
18
i
3.2 Findings
I 3.2.1 General
Licensee management continued their presence and involve-
'
ment in daily activities. The quality assurance department
sustained their presence and detailed involvement in licensee
activities. However, the pace of activities between Janu-
ary 2 and 5 was hurried and it apparently caused a lapse in
complete review of technical problems in preparation for
- startup after the turbine trip. Specific problem areas
were addressed in paragraph 2.2 for the short outage and
performance problems noted below on a long term basis were
I
indicative of this problem, that is, the thoroughness of
management and technical support personnel review of plant
problems.
"
3.2.2 Fire Doors ,
,
The inspector reviewed the status of fire door maintenance
and surveillance activities. This review was conducted
because of numerous shif t inspector observations of fire
doors being not fully closed.
As part of this review, the inspector toured Unit 1, held
various discussions with licensee personnel, and reviewed
, the following documents.
--
Preventive maintenance (PM) procedure U-27, "Func-
tional Testing of Fire Doors," Revision 0, dated June
19, 1985
--
Surveillance procedure (SP) 1301-1, " Shift and Daily
Checks," Revision 58, dated September 5, 1985
--
SP 1303-12.20, " Fire Door Inspection - Control Build-
, ing," Revisions 3 and 4, dated December 4, 1984 and
December 20, 1985, respectively
1
'
--
SP 1303-12.21, " Fire Door Inspection - Primary Side,"
Revision 4, dated May 17, 1985
--
SP 1303-12.22, " Fire Door Inspection - Screen House,"
1 Revision 4, dated December 11, 1985
--
TMI-1 Fire Hazards Analysis, Fire Protection Program a
i
and associated documentation
. ! During a tour of the intermediate building on December 17,
1985, at approximately 7:45 a.m. , the inspector observed
fire door I-107 tied open with a piece of rope. No one was
.- . - . _ , . - - - - - -.
-. - - - . . _ _ . - -
19
observed in the immediate area. At approximately 11:30
a.m., the fire door was still open. The inspector brought
his concern to the attention of the shift supervisor; and,
shortly thereafter, the fire door was closed.
During discussions with cognizant licensee personnel, the
inspector was told that fire door I-107 does not separate
fire areas; it only separates fire zones. Also, it is not
part of a three hour rated fire wall and was not evaluated
in the licensee's Fire Hazard Analysis submitted to NRR as
part of their Fire Protection Program. After reviewing
! '
related documention, the inspector agreed with the licensee
representatives. The safety significance of leaving this
fire door tied open was minor and was not a violation of
technical specifications.
The inspector observed that most licensee personnel are not
l aware of which fire doors must be kept closed. Some fire
doors that do not have to be kept closed are labeled " Keep
Closed." Other fire doors that do not have to be kept closed
are not labeled at all (fire door I-107 was this type of
fire door). The inspector concluded that the labeling of
fire doors was confusing and brought this to the attention
of the licensee. Licensee management agreed that the
present labeling of fire doors was confusing and personnel
may not be sure which fire doors must be kept closed.
On January 6, 1986, the licensee removed " Keep Closed" signs
from all fire doors that do not have to be kept closed. Only
l '
those fire doors that are required by technical specification
to be kept closed are now labeled " Keep Closed." The licensee
stated that they would like to see the fire doors that are
not labeled remain closed as much as possible and will rely
on employee judgement to achieve this. The inspector found
l the licensee's action in this area to be satisfactory and
had no further questions.
The inspector also reviewed maintenance activities associ-
ated with these doors. PM procedure U-27 provides an ac-
,
ceptance criterion that normally-closed fire doors will
'
close and latch when started from a 3/4 opened position and
l released. However, during discussions with licensee per- l sonnel, the inspector was told that PM procedure V-27 had
not yet been implemented. The licensee committed to place
'
this procedure into their preventive maintenance schedule
beginning in January 1986, and will initially perform it
twice per year. Based upon an evaluation of results from
this procedure, the licensee may reduce the frequency to
once per year. The inspector found this action to be
satisfactory and had no further questions in this area.
- - _ - . - - - _ - - - - - . . - - - . - - - _ . - , _ _ - - - - - . - . .
20
During the inspector's review of surveillance procedures
associated with the fire doors, the inspector noted a
discrepancy. Fire door D-108, which is in a three hour
fire rated wall, did not appear in any surveillance
procedures. This condition existed from December 16, 1985
through December 19, 1985, when the fire door was not
checked as part of SP 1301-1.
Fire door D-108 was originally contained in SP 1303-12.22,
Revision 3, along with screen house fire doors. The
licensee decided that fire door D-108 would be more appro-
priate in SP 1303-12.20, along with control building fire
doors. Procedure change requests were initiated to remove
fire door D-108 from SP 1303-12.22, Revision 3, and to add
it to SP 1303-12.20 as Revision 4. These procedures were
attached and went into the change process as one package;
both should have been issued together as new Revision 4
procedures. However, SP 1303-12.22, Revision 4, was issued
on December 11, 1985 and SP 1303-12.20, Revision 4, was not
issued until December 20, 1985, when the inspector brought
this issue to the licensee's attention. Therefore, between
December 16-19, 1985, the licensee did not survey fire door
D-108. The inspector concluded that the above noted proce-
dure change issuances were not properly coordinated and
this resulted in an apparent violation of TS 4.18.7.2.c.
(289/85-30-02)
The licensee stated that security personnel routinely tour
Unit 1 and shut any fire doors that may be open. The in-
spector reviewed the security logs fcr the period of Decem-
ber 16-19, 1985. Security personnel did tour the diesel
room area; however, they did not sign off on fire doors,
only security doors. A senior licensee representative from
security stated that checking closed fire doors was not
part of the security guards' routine, but they would
probably close a fire door if it was open. The inspector
had no further questions in this area.
3.2.3 PORV Exception and Deficiency J &D) Data Review
The inspector reviewed data from surveillance procedure
1303-11.45, "PORV Setpoint Check," Revision 5, dated Sep-
tember 3, 1985, that was performed on December 20, 1985.
This review was conducted because the PORV was declared
inoperable due to three identified deficiencies.
21 The inspector found the surveillance package to be complete
with three deficiency sheets attached. The first deficien-
cy found the PORV high pressure reset to be out of toler- ance high. The second deficiency founc the PORV temperature interlock trip to be out c. tolerance high.
The third deficiency found the PORV low pressure reset to be out of tolerance low. The inspector questioned cognizant licensee personnel re- garding the cause of these three deficiencies. The inspec-
tor was told that instrument drift, dirty potentiometers,
or potentiometer adjustment relaxation were the most proba- ble causes. Corrective action for these three deficiencies consisted of recalibrating the bistables in accordance with procedures. A review of SP 1303-11.45 data from the past twelve months was conducted to determite if any obvious adverse trends existed. The inspector noted that the PORV temperature interlock trip was out of tolerance during the March and November 1985 surveillance tests. The inspector brought his concern to tha attention of the licensee. A representative stated that three recurrent de- ficiencies on the temperature interlock function would be picked up during trending reviews and action would probably be taken to correct the problem. In this case, the module would probably be replaced with a new module. As a result of the inspector's concern, the licensee initiated a cor- rective maintenance job ticket. The inspector concluded that corrective action taken by the licensee in response to the these deficiencies was appro- priate for the circumstances. The resident inspectors had no further questions in this area but will continue to rou- tinely follow deficiencies associated wfth the PORV. During the previously-mentioned review of past data from $P 1303-11.45, the inspector noted a confusing statement asso- ciated with the reset point for the PORV temperature inter- lock. The desired trip point is 275 F (-2.500 V) and the desired reset point is 278 F (-2.400 V). A footnote to the reset point states, " Required reset point is 0.100 1 0.050 y above trip point." However, the inspector noted several instances in which several I&C technicians used the wrong reset band. The reset band should be based upon the as-left trip point, not necessarily the desired trip point.
22
The inspector brought this concern to the attention of cog-
ni:ar.i licensee representatives. They agreed with the in-
spector that confusion exists among some I&C technicians as
to what the resat band should be for the PORV temperature
interlock. The licensee committed to revise SP 1303-11.45
so that it is clearly understood that the reset point is
based upon the as-left trip point.
The inspector concluded that the probability of setting the
reset point out of tolerance exists. Although the proce-
dure was somewhat confusing, the inspector found it to be
adequate. Several instances were noted in which the wrong
tolerance band was used; but, by chance, the as-left reset
point was in tolerance to the proper acceptance criteria.
3.2.4 Makeup pump Lube Oil Pressure Switches
On several occasions, the license has noted problems with a
oil pressure switch associated with makeup pump 1A in that
the switch was not stopping the motor driven lube oil pumps
after proper oil pressure had been achieved using the at-
tached lube oil pump. The motor driven oil pumps are used
to supply lube oil when starting the makeup pump. Under
normal running operations the motor driven oil pumps are
secured and an attached lube oil pump supplies the neces-
sary oil. To secure to the motor driven oil pump, when
proper oil pressure is generated by the attached oil pump,
the pump has oil pressure switches installed as part of the
control circuitry. The original lube oil pressure switches
were not environmentally qualified. The switches were re-
placed with a qualified switches. Evaluation by the
licensee, determined that the pressure setting of switch
was too high. The switch was eventually readjusted and is
working correctly.
The inspector reviewed and discussed the work that was per-
formed on all three makeup pumps with plant engineering and
maintenance personnel. Review of technical specifications
indicated that the licensee remained in compliance of their
technical specifications. Work performed was documented in
accordance with applicable station procedures. The inspec-
tor concluded that the inability of the lube oil switch to
function as designed did not degrade the ability of the make-
up pump to perform its safety-related function.
3.2.5 Diesel Generator Standby Coolant System
As part of the routine plant tour, the inspector reviewed
the auxiliary operators' logs. The Jacket coolant and lobe
oil temperatures kept warm by the standby coolant system
were consistently low out of specification. The normal
values for jacket coolant temperature were expected to be
_ __ - _ _ _ - - - _ - -- ._ - . - - _ _ _ = _ . - - _ - - _ _ - - .
_ --
l
23
l
120 F to 140 F. The readings obtained by the auxiliary
operators on a shif t basis were between 112 F and 115 F.
The temperature of the lube oil is to remain above the
alarm setpoint of 90 F. The cause of the low temperatures
was attributed to check valve leakage in the standby coolant
system. (See NRC Inspection Report 50-289/85-26.)
The inspector queried the licensee whether the lower tem-
peratures had an adverse affect on the operability of the
diesel. From discussions with plant engineering and opera- *
tions personnel, the licensee demonstrated that the low
Jacket coolant temperature (temperature greater than 100 F)
would keep the lube oil temperature above 90 F. The actual
lobe oil temperature critical limit was approximately 65 F.
Review of diesel support systems indicated that lube oil
temperatures less than 65 F was the limiting condition as-
sociated with "keepwarm" systems. Review of licensee pro-
cedures demonstrated that sufficient operator action was
required to ensure that the lube oil temperature would not ;
drop below 90 F.
Further, licensee representatives indicated that the check -
valves in question would be repaired during the next outage
af ter the new internal components were received f rom the ;
manufacturer. The inspector questioned the licensee's phi- '
losophy of daily logs having out-of-specification readings
(see also paragraph 3.2.8). It appears that the licensee
was willing to accept out-of-specification readings until :
repair of the check valves could be accomplished. The !
licensee acknowledged this concern and they stated that '
they would review the applicability of the current speciff- ;
cation values.
3.2.6 Letdown _. Cooler Isolation
.
The inspector reviewed the sequence of events and the chem-
istry data associated with determining which letdown cooler
was leaking. The inspector initially reviewed the sequence j
of events during NRC Inspection 50-289/85-27. At the com-
pletion of that report the licensee was evaluating the cur-
rent information to determine which cooler was leaking.
From the chemistry plots, the licensee was able to deter-
mine that the 'B' letdown cooler had a leak rate of approxi- :
mately 1 liter / day. The 'B' letdown cooler was isolated t
both on the primary and secondary side. The cooler has
been restricted to limited use under the direction of the ,
plant operation manager. Plant maintenance has started the '
necessary engineering and maintenance preparation to sup-
port replacement of the cooler during the next refueling
outage.
. _ - _ - - _ - . - - - . . - - _ . _ - _ . . . .
---
l l
1
i
;
24
I
!
'
I
The inspector reviewed the chemistry data and licensee rea- l
soning for their conclusion of the "B" cooler leakage. The
inspector discussed with plant maintenance and operational l
;
personnel the limited use and the licensee's schedule for ;
l cooler replacement. It was concluded that limited use of the !
- cooler would not adversely affect plant safety. The un- t
availability of a second cooler would only affect plant !
operations during startups, and affect the length of time j
to restart the plant from a cold shutdown condition. The
'
inspector noted that because the leakage was small, the r
licensee initially had difficulty in determining which . l
cooler was leaking. After the 'B' cooler had been isolated i
on both the primary and secondary side and the chemistry in !
l the intermediate close cooler was trended for a sufficient l period of time, the licensee was then able to determine I
which cooler was leaking. !
3.2.7 Temporary Digital Readout for RM-A5
I
In order to easily read the indicated counts for the main '
condenser offgas radiation monitor RM-AS, I&C technicians j
connected a portable test equipment to the permanent read- l
out in the control room. The portable test equipment was !
placed (unsecured in any manner) on top of the cabinet con-
taining RM-A5 readouts. The digital readout of the test
equipment was thus used by the operators as an accurate
method of quickly obtaining the count per minute reading
for RM-A5.
l
The inspector questioned an I&C foreman on whether the test l
equipment was considered permanent and should be considered I
a missile hazard. !
!
E
I The foreman stated that the equipment would remain until ! l corporate engineering had designed a permanent resolution. !
The permanent resolution may be a new readout or just using i
'
existing arrangements. The foreman, however, acknowledged
the inspector's concerns about the test equipment being a
missile hazard. He immediately directed that securing I
straps be placed on the test equipment to ensure that it l
could not become a missile hazard. :
l
The inspector reviewed the Interim solution and considered !
it acceptable until the final engineering resolution was r
,
developed. !
3.2.8 Auxil_iary_ Operator Logs I
The inspector reviewed selected primary system and secon- I
dary system auxiliary operator (AO) logs completed during :
the latter part of 1985. The A0 logs are used by auxiliary l
!
!
!
- _ . , _ . . _ - , _ _ .---- - - . , . ~ -
- -_ . - _ _ _ . . _ . - _ - _ _ - . _ - - - .__ _ _ _ _ _ _ - - - . . -
,
;
25
l
i
<
operators to record numerous plant data and observations,
j If any data or observations are out of specification as
i.
'
indicated by the reference values in the A0 logs, the aux-
'
iliary operator circles it to alert the reviewers (shif t
foreman and operations engineers). Appropriate action, if
l any, is then taken.
} The inspector cbserved that there were frequent red-circled
items throughout the A0 logs covering a period of several
i months. A licensee representative stated that the A0 logs
; are used for trending purposes and that the reference values
! may not be accurate because the plant had not been operated
! for a long period of time. The licensee representative
j also stated that the vast majority of the red-circled items
! are not indicative of problems, and if they are, they are
j .rpropriately dispositioned.
l
l The inspector concluded that the manner in which A0 log out '
} of specification entries are being resolved was a poor
) operating practice. A red-circled item should alert opera-
tions personnel that impending problems are evident. If
] red-circled items are passed by week after week, the effec-
tiveness of the A0 logs are greatly reduced. It could con-
dition operators to treat " abnormal" readings as " normal"
readings and corrective actions may not be taken. If the
4 reference values are not appropriate for the mode of opera-
'
tion, new reference values should be determined. The i
i resident inspectors will continue to routinely follow this ,
; area. r
i
!
3.2.9 Annunciator Black Goarcj '
j On several occasions, the shift inspectors noted certain { annunciator alarms that remained lit for extended periods i 1 of time and other alarms that remain on permanently. For each j '
alarm the inspector discussed the reason or cause for the
alarm with plant operations personnel. In certain instanc-
] es the alarms were due to testing in the plant, however
j some alarms were due to actual plant conditions and the way
the circuitry was designed.
1 i
Discussions with plant management indicated that the licensee !
was working toward having all annunciators out when the ;
; plant was operating at full power. The inspector discussed
j the major annunciators that were continuously lit during l
plant operations. In each instance, the licensee satts- '
i factorily resolved the inspector's concerns about alarms
that are continuously lit. However, the inspector questioned
1
{ why ground indications on both station batteries were period- -
i ically received. Plant operations personnel stated that i l Job tickets were being written or have been written to ' 4
correct the problem. The inspector verlffed that the i
i
a v
6
,
._ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _
. _ _ ... __ _ _ _ _ _ .-_ _.-__.._ _ _ _ _ _ . _ _ _ - . _ - - . _ _ _ _ . -
l
l
'
26
i
licensee was pursuing these problems; bat he also noted
that the battery ground problem was not being resolved ag- (
gressively by licensee management. The inspector had no (
!
,
further questions in this area.
3.2.10 Procedure Implementation
l
As a result of a shift inspector concern identified during
the plant startup conducted on January 4, 1986, which re-
lated to the performance of a heat balance at approximately
30*4 power, the completed plant startup procedure 1102-2 was
reviewed and discussed with the licensee. The plant ;
startup procedure requires the performance of a heat bal-
ance at "approximately 30*4 power." This step in the com-
pleted procedure was initialled as having been performed.
A search of records showed a heat balance was performed at
41% power. ;
Based on discussions with licensee personnel, the computer I
, used for performing the heat balance was unavailable at 30*4 !
power. Licensee management was present in the control room
l ; i when 30*. power was achieved. Since the heat balance was ! l
used to calibrate the power range instrumentation and the l
procedure calls for additional heat balances at progres- !
sively higher power levels prior to achieving 100*. power, i
l the decision was made to perform the heat balance specified , l l
at approximately 30*. power, later at 41*4 power when the l
computer was available. The actual power at which the heat !
balance is performed is immaterial for what it is intended ;
to accomplish, that is, for verifying the proper response
l of the power range instrumentation. Accordingly, licensee !
management determined the performance of the heat balance ;
at 41% satisfied the approximate 30*4 specified in the
procedure. Based on this determination, the step requiring i
the heat balance at approximately 30*. power was signed of f. l
At the 41*4 indicated value, the power range instrumentation l
was recalibrated to 43*4 based on the heat balance calcula- L
tion. The inspector had no further questions relative to
this matter .
1
Procedure 1102-4, " Power Operation," in discussing the heat ;
balance requirements specifics heat balances at 15 and 30*4 i
power. This was identified as a typographical error and a !
Procedure Change Request (PCR) has been prepared to have I
i
the procedure read 15 to 30*4 power to agree with the same
I requirement in Enclosure 5 of Procedure 1102-2. ,
I
i
l r -
l
i
- , . - , . -
-n -~,nn-- . . . - - . .-,--.,.,,,n , , , , , , .- - _ - - - , .
27
3.2.11 Reactor Plant 0peration and _ Testing Activities Between
January 5 and 10, 1986
_
As previously noted in prior inspection reports, steam gen-
erator water levels had been higher than expected due to
apparent deposits on the secondary side of the steam gener-
ator tubes. Following the most recent reactor startup on
January 4,1986, the steam generator levels were at a lower
level than had been expected. Due to the importance of
steam generator levels to plant operations, a review was
conducted to develop some assurance that the level readings
were accurate. To accomplish this, five steam generator
level instrument differential pressures on each steam
generator were compared and found to agree with the
instrument tolerances.
The inspector verified that technical specification sur-
veillance requirements for the instruments was verified to
have been perforrned as required. The required weekly sur-
veillance checks were verified to have been performed as
required in accordance with Surveillance Procedure
1301-4.1, Weekly Surveillance Checks, Revision 26, November
24, 1983. Completed surveillance procedures for December
28, 1985 December 20, 1985, December 13, 1985, and Decem-
ber 7, 1985 were reviewed. Data showed all instruments to
be accurate to within the specified acceptance criteria.
The inspector had no further questions in this area.
.
A sampling check of the licensee's control of locked valves
was conducted. Two valves BS-V-54A and BS-V-49A, normally
locked-open valves, were noted as being unlocked and
closed. A review of the control room locked valve log
showed the valve positions were under the control of a sur-
veillance test procedure and that their status was properly
documented.
The licensee's activities associated with a drop in instru-
ment air pressure were observed. Considerable personnel
activity was noted in the intermediate building in the area
of the instrument air compressors by individuals looking
for possible instrument air system leaks. Also, announce-
ments were made over the pubite address system for people
using instrument air to discontinue doing so. A further
evaluation by the licensee determined the loss of instru-
ment air pressure resulted from the simultaneous venting of
a number of air operated isolation valves. The venting of
the v41ves was in accordance with a surveillance procedure
1303-5.1, Reactor Building and " Isolation System logic
Channel and Component Test." This procedure had not been
perfortred previously with the plant at power and the loss
of instrument air pressure was not considered. Further
.. _. _ ._- _ _ . _ _ _ _ _ -_.. __ _ _ _ _ _ _ .- _ _ . . _ _ . _ _ _ - . .
!
28
,
I I r '
evaluation showed the use of instrument air was well within ;
the capability of two air compressors and that following
~
this experience with the procedure, no problems should be -
encountered during future perform 1nce of the surveillance. <
Also, the licensee is evaluating a change to the procedure
or the addition of a caution to help eliminate any future
problems.
!
During plant operation at 100% power on January 10, 1986, -
some increase in the "B" main feedwater pump vibration was
noted. Reactor power was reduced to approximately 78% and
the feedwater pump was removed from service. The source of
the vibration was traced to the pump coupling. The cou- ,
pling was disassembled, cleaned, lubricated, new "0"
rings installed, and reassembled. Portions of the pump i
maintenance and subsequent pump restart were witnessed by
the inspectors. The pump maintenance and startup were ob- l
served to have been performed in accordance with appropri- '
ate procedures and no deficiencies were noted. Quality
[
control involvement in the maintenance activity and subse- t
quent vibration data gathering wa> also noted.
The repair of the control room position indication of the
No. 4 main steam stop valve was observed. This repair
required the closing of the stop valve and the replacement
of a broken rod associated with the position indication
system by ISC personnel. The entire task was well planned
between operations and I&C personnel, good communications I
were established and maintained. The good planning and
cooperation between operations and I&C personnel resulted
in a swift and successful repair.
3.3 Conclusion
Licensee management and the quality assurance department continued
their presence and involvement in daily activities. Problems were
noted in the thoroughness and completeness of management review of
, plant problems or at least in assurino thorough and complete techni- , L cal support personnel review of plant problems. ! i
h
The poor operational practice of living with out-of-specification !
readings in daily logs was indirectly tolerated on a long term basis (
by intermediate managers, thereby conditioning operators to normal
" abnormal" readings. Individual procedure step inadequacies contin-
ued to be identified by inspectors indicating a problem with the at-
tentiveness of procedure reviewers and implementers to correct these :'
problems. The labelling on fire doors was confusing and an improper-
ly coordinated procedure review resulted in a violation of a regulatory
requirement. i
!
e
i
;
r i
_ _ _ - - . _ _ . _ _ _ _ _ _ - _ _ _ _ _ _ _ _. __ ._ _ _ . . . . _ .
.-
1
J
:
29 -
,
1
Records of maintenance and surveillance activities were properly re-
- tained. These records indicated that appropriate corrective action
i
was issued for routine items needing repair along with proper post !
maintenance testing. !
i ] The startups after reactor trips were conducted in a deliberate and
- conservative manner and operators continued to demonstrate their
- highly developed skills in operating the integrated control system in
i manual. ) 4. Star _ tup Testing 1 4
4.1 Test _ program Meeting Summary
i ! On December 17, 1985, a meeting was held at the TMI training center > in Middletown, Pernsylvania, to discuss the status of the power as- I cension program for TMI-1. Meeting attendees as well as the meeting j agenda and slides are identified in Attachment 1. The results of , l this meeting are summarized below.
Because of apparent fouling problems on the secondary side of the f
i
steam generators, TMI-1 was not expected to achieve 100'e power. The i
- licensee plan of action was to proceed past 75*. power around December
i 27, 1985, and to reach whatever maximum power would be achievable '
< with their present limitations. They would complete power ascension l test and around January 2, 1986, conduct the planned reactor trip l from maximum achievable power. Then if all prcceeded as scheduled, j the licensee would start up on January 3, 1986. The NRC staff agreed
l that completing the power ascension program at the maximum achievable
! power vice 100*4 power would be acceptable. However, the licensee was ! requested to document its plan of action via formal letter. (This : i was completed by letter, dated December 19,1985). 1 l The Itcensee also discussed in detail two situations which arose dur- 1 ing the startup. The first involved the EFW turbine-driven pump
steam supply safety valves lifting wnen the pump actuates. The licensee ,
,I
has proposed an alternative interim solution which was acceptable l
- while further investigations were to be conducted. The staff empha- l
l sized that a final solution must be achieved by the completion of ' i cycle 6 refueling outage. The licensee was requested to document i l their plan of action via a letter. (This was completed by letter, ;
dated December 20,1985), j
The second situation involved main steam safety valves controlling
j steam generator secondary pressure upon a reactor trip instead of the i turbine bypass valves. Based upon investigative methods in process,
- the staff concliund that the Ifeensee was appropriately addressing
i its concerns. (By letter dated December 20, 1985, the licensee con- s
cluded no safety concern exists and they committed to notify the NRC
'
! ! staff of their review of test results by the restart after the eddy l current outage planned for March 1986.) i i r
I
l
i J
. - . - _ - _ _ . .-. -----._ -- .- - _. - . . _ . -
1 i
,
i
'
*
h
30
- [
j 1 l l 4.2 PowerEscalationTestinL(MaximunAchievablePowerPlateauTesting} l i !
i 4.2.1 Scope of Review l
} !
j 4.2.1.1 Test Witnessing i
i At various times during the inspection period, the inspec- l 3 tors witnessed testing in progress on a sampling basis. ! However, test procedure (TP) 800/2, " Reactor Trip on Tur- j bine Trip," and TP 800/9, "EFW Pump Auto-Start Test," were
- witnessed in their entirety by the TMI-1 Restart Staff. !
l The tests were observed to verify that: s
l
i
1 --
tests were conducted in accordance with appropriate i
'
test procedures;
) --
prior to performing tests, an adequate briefing was
4
conducted for operations personnel; l
i !
--
4 test trerequisites and initial conditions'ere
w met; l
- L
j --
applicable tecnnical specifications were complied
4
l
with; -
f --
f
; operator actions were correct; ,
--
test engineers were knowledgeable in their duties;
and,
i
! --
test results were acceptable. "
;
- In addition to TP 800/2 and TP 800/9 witnessing, the fol- i
i Iowing tests were observed and/or their test results inde- i j pendently reviewed by the TM1-1 Restart Staff during this i l
inspection period. !
!
,
--
TP 800/5, " Unit Load and Steady-State Test" j
i ,
--
TP 836/1, "Feedsater System Operation and Tuning"
--
TP 849/1, "lCS Tuning at Power"
I i --
TP 846/1, "Incore Thermocouple Operations Test" !
i --
TP 8E5/2, " Turbine Bypass Valve Tes."
1
t
l --
TP 800/2, " Reactor Trip on Turbine Trip"
. l --
TP 800/9 "EFW Pump Auto-$ tart Test"
i i '
- !
: , _
31
--
CP 1105-14. " Loose Parts Monitoring System"
--
RP 1550-08, " Core Power Distribution Verification"
--
RP 1550-05, " Reactivity Coefficients at Power"
--
SP 1301-9.5, " Reactivity Anomaly"
--
SP 1303-1.2, "RCS Flow Surveillance"
--
OP 1103-16, " Heat Balance"
4.2.1.2 Test Results Review
Test results from the testing program for the 83*. and 100*.
power plateaus were reviewed by the inspector to verify
that:
--
test changes were approved and implemented in accor-
dance with acministrative procedures;
--
changes did not impact the basic objectives of the
test;
--
test deficiencies and exceptions were identified and
resolved and resolutions were acceptable;
--
the cognizant engineering group has evaluated the test
results and signified that testing demonstrated that
design conditions were met; and,
--
test results were within established acceptance crite-
ria or properly resolved.
4.2.2 Licansea T e s t_J e s u_1_t s_a nd_]J RC F i n d i_n g .s
Licensee performance of key tests is described in this sec-
tion. The discussion includes a summary of key test objet-
tives and test results; test performance including
operators, test engineers and equipment; and pertinent
findings and outstanding problem areas identified and/or
NRC findings as a result of testing. Cecause of the sig-
nificanca of IP 800/2 and TP 800/9, the details on the per-
formance of these tests and their results are discussed
separately in paragraph 4.3.
NOTE: Paragraphs 4.2.2.1 through 4.2.2.6 discuss special
tests performed at 88*. and 100*. Power levels except for IP
l '
800/2 and 800/9 which are discussed in paragraph 4.3. All
tests were completed by Janaary 10, 1986, prior to the end
of this reporting period. In somo instances, NRC review of
i
- - - - - - - - _ - - - _ _ - - _ - _ - _ _ -
- -. --
l
'
l , ! * l 32 i l
these data was conducted subsequent to January 10, 1986,
however the results of this review are included in this
inspection report for completeness.
4.2.2.1 Unit Load Steady-State Test (TP 800/5]
,
The steady-state plant parameters were measured per TP
800/5 at the SS*. and 100*. power levels. As noted in NRC
l Inspection Reports 50-289/85-27 and 85-28, the OTSG high l water level problem temporarily limited the achievable pow- l er to 88*i. Upon recovery from the TP 800/2 reactor trip l test, the steam generator water levels began to track with-
in the original predicted range up to 100* power level and
I no longer impose operation limits on the reactor power.
The rest of the steady-state plant parameters during I
power escalation continuously showed good agreement with
the predicted values. l
{
4.2.2.2 Feedwater System _ Operation and Tuning _(TP 836/1) t
. The feedwater system operation was tested per procedure TP j l 836/1 at the 85*. and 100*. power levels. The measured l l feedwater flows and temperatures (af ter second stage high- '
pressure heaters) agreed well with feedwater cycle perfor-
mance predicted values. No unacceptable feedwater system t
l oscillations were observed during these tests. l l l l 4.2.2.3 ICS Tunino at Power (TP_849/1) (
ICS tuning was completed satisfactorily at 88*. and 100*. l
l power. The significant portion of this test was a reactor
runback on loss of main feedwater pump. On December 27,
1985 at approximately 1:40 p.m., the licensee secured a
main feed pump as per the test procedure. In a period of
approximately five minutes the reactor rods inserted reduc- l
ing reactor power f rom 88*. to 70*.. The transient was wit-
nessed by the inspector. There were no plant equipment
problems and the operators handled the transient properly. ,
Test results were acceptable. l
4.2.2.4 Incore Thermocouple _(T/C) Operations Test (TP 846/1)
The incore I/Cs were checked at 88*. and 100*. power levels
per procedure TP 646/1; four out of fifty-two T/Cs, identi- j
fled as out of service during earlier testing, remained out l
of service during these tests. The remaining 48 operable l
T/Cs exceeded the minimum requirement of 16 operable T/Cs !
located in a specified pattern. ,
!
l
I
l
t
_ _ _ _ - _ .
33
A problem identified during previous inspections at lower
power plateaus was that the location of the five highest
T/C readings as read from Mod Comp computer Area 3 groups
8, 9, and 10 did not agree completely with the selection
as determined by the computer program "TCDSPL." The data
used for this comparison were taken at two different times.
Due to the quasi-steady nature of the system response, and
the small temperature difference among the higher T/C
readings, this discrepancy was explainable. At the 88's
power level test, as the temperature difference among T/Cs
increased, the discrepancy only occurred at the fifth
highest reading, whereas the first to the fourth highest
T/C readings were correctly selected by the computer
program "TCDSPL." At the 100*. power level test as
temperature difference increased even higher, all five
highest T/C readings were correctly selected by the comput-
er program "TCDSPL."
Other than as noted above, the functional check of all op-
erable T/Cs and the associated computer system continuously
showed acceptable results.
4.2.2.5 Core Power Distribution verification (RP 1,550-08)
The detailed core power distribution at the 88*. power pla-
teau was measured by the licensee per procedure RF 1550-08,
" Core Power Distribution Verification," using the incore
detector system. A procedure temporary change notice (TCN)
No. 1-85-0206, dated December 26, 1985), was written to
correct the predicted values for performing this procedure
at a higher burnup and lower power level than originally
planned. The inspector noted the following results:
--
The measured radial peaking factor for each assembly
was consistent with the analytically predicted value.
The comparison of the highest measured radial peaking
factor (1.272) at core location K-11 agreed closely
with the predicted value of 1.224.
--
The measured total peaking factor in each fuel assem-
bly also agreed consistently with the predicted value.
The highest measured total peaking factor of 1.463
agreed well with the predicted value of 1.416.
--
The measured linear heat rates accounting for various
uncertainty factors were within TS 3.5.2.7 limits, as
indicated in the following table.
- _ . .
_ . _ - .- - -- .. . _.
..~
,
! v 34 l
J i
'
-
h
; Measured Maximum
d
Axial Location Maximum * Allowable
; from Bottom Linear Heat Linear Heat
! , of Ccfe Rate Rate
, s (ft) .(KW/ft) [KW/ft)
* including:p wer spike factor
~
11.14 6.98 15.20
9.43 -11.82' 16.26
- 7.71 12.26- 17.10
6.00 11.88 17.50
I
,
4.29 11;75 16.31
l
.
2.57 12.00 14.98
! , 0.86 9.20 13.50
I ,
I
'
The licensee also measured the core power distribution at
"
'
100*. power on January 9, 1986. The inspector reviewed the
'-
test results and noted that the measured values were all
within TS limits and consistent with the data obtained from
i
88*. power results.
4
All results were acceptable.
4.2.2.6 Reactivity Coefficietts Measurement (RP 1550-05)
.
'
The temperature and power coefficients of reactivity at
! the 88*e power level were measured per procedure RF 1550-05. l The measured temperature coefficient was -10.17 pcm/F.
~
This value was in the test acceptance range of -14.0 pcm/F 1
3 4 pcm/F. The temperature coefficient is the sum of the
'i moderator temperature coefficient (MTC) and fuel doppler
- coefficient. The corresponding MTC at 88*. power as
derived from this measurement and the extrapolateJ value
-
for 100?; power at an all rods out cendition are as follows:
1 Power Level MTC TS Limits
(*41_ _ , IP.cm/F) apcm/F)_
; 88
-8.7 <5
!
100 -8.2 <0
! . The measur.,[[ pone coef ficient (power doppler coef ficient, , I since RCS temperature was maintained constant during this
test) was -8.59 pcm/*; power. This value satisfied the
safety analysis requirement that the power doppler. coef f t-
cient be more negative than -5.5 pcm/'. power. However, the
measured value was slightly out of predicted range of
-13.08 pcm/?; power + 4 pcm/?; power. The licensee nuclear
- engineer performed an engineering evaluation and concluded
j that the measured power doppler coefficient has no impact
on nuclear safety.
!
'
!
%
'
l , j
.-_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ . _
_ _ _ _ _ _ _ -
!
.
35
i
In addition, the licensee concluded that this value of the
power doppler coefficient would not have special signifi-
cance on operations. This is due to the relatively small
contribution of the power doppler coefficient on the overall
reactivity calculation. The predicted doppler coefficient
is used in two plant procedures: SP 1301-9.5, " Reactivity
Anomaly," and SP 1303-15A, " Reactivity Balance." The in-
spector reviewed the reactivity anomalies plot maintained
by the nuclear engineering group since the beginning of
this fuel cycle and noted that the measured values consis-
tently agreed with the predicted values.
Further, the inspector witnessed the plant startup on Janu-
ary 4, 1986, and observed that the estimated critical post-
tion (14% withdrawal on Group 7) was very close to the actual
critical configuration (14.2*. withdrawal on Group 7). Based
on these observations and independent verification of safety
analysis assumed parameters, the inspector concurred with
the licensee's engineering evaluation that there is no
unreviewed safety question involved.
4.2.2.7 Core Thermal Power (OP 1103-16, " Heat Balance
;
Calculations")
The core thermal power is usually determined by the plant
computer (Group 55) which monitors the required input pa-
rameters and performs a heat balance calculation. Hand
calculation of heat balance is allowed as a backup when the
plant computer is not operating. The inspector performed
an independent calculation using the plant parameters taken
at 10:38 a.m. on January 8,1986, for comparison.
Plant Inspector's
Test Date Group 55 Result Calculation
(MWt) (% Power) (MWt) ($ Power)
January 8, 1986 2532.91 99.92 2533.3 99.93
(10:38 a.m.)
This comparison indicated that the licensee's core thermal
power calculation is adequate. Through control room obser-
vation, the inspector noted that at about the same time,
the power range indications NI-5 through NI-8 were consis-
tent with the calculated core thermal power.
__ - _ - _ _ _ - _ - _ _ .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
.
.. . ..
36
4.3 Reactor Trip on Turbine Trip (TP 800/2) and Automatic Start of EFW
Pumps on Loss of Reactor Coolant Pumps (TP 800/9)
1
4.3.1 Reactor Trip on Turbine Trip (800/2)
4.3.1.1 Test Performance
Restart condition 2.C.(1)(d) requires that prior to comple-
tion of the power escalation test program, the licensee
demonstrate automatic reactor trip on turbine trip. This
test was performed on January 2,1986, in accordance with
TP 800/2, Section 9.2, " Reactor Trip Due to Main Turbine
Trip," during which the main turbine was manually tripped
from the control room console. Following the trip of the
main turbine, the following events were expected to occur.
--
reactor trip on the anticipatory turbine / reactor trip
--
turbine bypass value setpoint transfers to 1010 psig
--
containment isolation on reactor trip and reset features
function as designed
,
I
i
--
OTSG levels control at 30 inches by main feedsater
The results of this test are discussed below.
4.3.1.2 Test Results
The test results indicated that the reactor had automati-
cally tripped upon turbine trip. The maximum RCS pressure
attained following the turbine trip was 2150 psig. The
result was acceptable since it was below the reactor high
pressure trip set point of 2300 psig.
In addition to the automatic reactor trip on the anticipa-
tory turbine / reactor trip; OTSG levels were controlled at
approximately 30 inches using main feedwater; and partial
containment isolation on reactor trip functioned vith one
exception-letdown line isolation valve MU-V3 did not
close. See paragraph 4.3.1.2.1 for further discussion.
During the TP 800/2, reactor trip test, the inspector ob-
served in the control room that the readings from source
range instruments NI-1 and NI-2 indicated about 1 to 2
decades difference. This subject was discussed with a
cognizant licensee representative. Troubleshooting per-
formed by I&C technicians immediately after the test
indicated that NI-1 count rate amplification gain factor
was out of tolerance. The gain factor was subsequently
corrected per surveillance procedure 1303-7.2, " Source
_ _ _ _ _ .
. . . __ .__ _ _
l
1
!
37
Range Channel." The inspector witnessed the licensee I&C [
technicians perform a second confirmatory surveillance test .
on NI-1 and NI-2 on the following e y (January 3, 1986). '
Surveillance results indicated that both NI-1 and NI-2 had
been properly adjusted. (Both instruments properly responded
during the unplanned trip on January 4,1986.)
Another test objective, that turbine bypass valves control
turbine header pressure (1010 psig + 10 psi), was not satis-
fied due to lifting of the main steam safety valves. This
problem was first identified in the 40% power reactor trip
'
test (NRC Inspection Report 50-289/85-25, paragraph 4.2.3),
and a similar plant response was noted in the unplanned
reactor trip that occurred on December 1, 1985, at 75%
power. This problem is further detailed in paragraph
4.3.1.2.2 below.
There was also a concern as to the effect of the reacter trip
j on RCS and OTSG leak rates. This is discussed in paragraph i 4.4 where inspection efforts concerning RCS and 0TSG 1eak
rates are detailed.
, 4.3.1.2.1 Containment Isolation Valve MU-V3 Failure to Clost
As part of the turbine trip / reactor trip at 88% power per
TP 800/2, the licensee was to verify that containment isola-
tion valves shut on the reactor trip signal. During this
trip letdown isolation valve, MU-V3, failed to shut, al-
though it had previously shut during the trip at 40% power ,
per TP 800/2 during and a later inadvertent trip at 75%
power. The failure of MU-V3 to shut was identified as a
test deficiency.
Following the test, as part of troubleshooting mainte-
nance, the valve was successfully tested three times, in
that, it shut three consecutive times on a simulated reac-
-
tor trip signal. Further troubleshooting indicated that a
10 second time delay relay (62 X 1/MU-V3), which allows the
'
operator to open the valve after 10 seconds, was defective.
This relay was replaced. The retest on the job ticket
stated only that the valve was satisfactorily tested by the
- reactor trip signal. TP 800/2, paragraph 2.1, stated, in
part, that one of the test objectives was to verify con-
tainment isolation on the reactor trip and that reset fea-
tures function as designed. The inspector noted that TP
'
800/2 failed to document the testing of the override / bypass
features for certain containment isolation valves.
A licensee representative stated that this was an over-
sight; however, the reset feature actually applied to
letdown isolation valve MU-V3, to allow restoration of
, letdown based on operator action. The licensee provided
documentation that the reset feature for MU-V3 was previously
tested August 10, 1981, per pre-operational test, TP 334/2.
- - - -- . - - . - _ . . . . - - _ __.
38
The inspector reviewed applicable portions of TP 334/2 and
determined that reset function of MU-V3 had been adequately
tested at that time. A complete NRC review of TP 334/2
test data has been documented in an earlier NRC inspection.
The inspector questioned whether the reset function for
MU-V3 had been adequately t(sted during the power escalation
test program. The retest of job ticket CI 489, which
replaced the time delay relay for the MU-V3 reset function,
did not clearly document whether the timer and its reset
function were fully tested. The inspector determined that
there is no surveillance test procedure for the containment
isolation function on reactor trip since this is not identified
in the technical specification. The ability of MU-V3 to
operate is tested quarterly per SP 1303-11.26, " Reactor
Building Isolation Valve Cycle Test"; however, this procedure
was not used as a retest for MU-V3 repair since it does
test the automatic features of reactor building isolation
- valves.
On January 3, 1986, while the reactor was still shut down,
the licensee performed a retest of MU-V3 by inserting a re-
actor trip signal using the manual trip button. Safety
control rods, which had been withdrawn, were reinserted
prior to inserting the reactor scram signal. MU-V3 shut
after insertion of the scram signal and would not respond
to an open signal until after 10 seconds. The inspector
witnessed the retest of MU-V3 and observed that the isola-
tion and reset function did perform satisfactorily. There
were no further questions concerning the operation of
MU-V3.
4.3.1.2.2 Main Steam Safety Valve Actuation Fo11cwing Reactor Trips
(Unresolved Item 289/85-25-05)
During the reactor trip at 40% power performed on October 21,
1985, the main steam safety valves and turbine bypass valves
did not respond as expected. After the initial expected
lift of the safety valves, some of the safety valves con-
tinued to lift and reseat several times essentially providing
pressure control in the steam header. The turbine bypass
valves failed to completely open and control pressure.
This event was detailed in NRC Inspection Report 50-289/85-25.
On November 21, 1985, during normal reactor power operation
at 48% power, the licensee, tested and adjusted six select-
ed safety valves in order to correct the above problem.
The setpoints of five of the six valves tested were adjust-
ed higher in order to preclude the potential relifts of the
valves after a trip. This was detailed in NRC Inspection
Report 50-289/85-27.
,
~n - w - r
39
On December 1, 1935, the plant experienced an unplanned
trip while at 75% power. During this transient, the main
steam safety valves (safeties) on both steam generators
initially lifted. The main steam safeties on the B steam
generator appeared to reseat and the turbine bypass valves
associated with B steam generator automatically took over
and regulated the pressure in the generator. On the A
steam generator the main steam safeties continued to lift
and reseat several times thus controlling the pressure in
the A steam generator. Approximately 12 minutes after the
trip, the shift supervisor directed manual control of the A
turbine bypass valves to lower steam pressure to below
safety valve setpoints.
Results from this trip, indicated the problem had not been
resolved by the previous adjustment of safety valve setpoints.
As a result, the licensee changed its procedures to have
the operator take manual control of the turbine bypass
system following a reactor trip. In addition, special
monitoring equipment was installed on each valve and tele-
vision cameras and thermographic instruments were set in
- place to observe the main steam safeties lift during the
scheduled trip at 88% power per TP 800/2.
The trip from 88% power was performed on January 2,1986.
It was agreed by plant personnel that the operators would *
wait until after the third lift of the safety valves before
taking manual control of the turbine header pressure
setpoint. On the first lift after the trip, the lifting of
the valves were observed on three television monitors and
recorded on video tape. On the second and third lifts,
several plant personnel were sent on to the roof to person-
ally observe and record which valves relifted. It was ob-
served on the trip that most, but not all, safety valves
lifted. On the two subsequent relifts, only four to six
valves relifted. As was observed during the earlier trip
at 75% power, B steam generator pressure stabilized quicker
than A steam generator pressure.
The licensee is currently evaluating data and videotapes
to determine the need for further corrective action.
Licensee representatives stated that they will have a plan
of action and schedule by the eddy current test outage
scheduled for March 1986. This item remains unresolved
pending further licensee action.
__ _ _ _ . . . _ _ _
__
40
4.3.2 EFW Pump Auto-Start Test (TP 800/9)
4.3.2.1 Test Performance and Results
Restart condition 2.C.(1)(f) requires that prior to comple-
tion of the power escalation test program, the licensee
demonstrate automatic EFW initiation on loss of all four
reactor coolant pumps. This test was conducted following
the reactor trip test per TP 800/2.
Similar EFW pump auto-start tests have been performed twice
during the power escalation test program; once during the
40% power plateau transient test (TP 800/8, "RCS
Overcooling Test") and once during the 3% power plateau
natural circulation test (TP 700/2, " Low Power Natural Cir-
: culation Test"). These two previous tests demonstrated
that EFW pumps automatically started upon tripping both
main feedwater pumps. Actuation times and pump discharge
pressures were verified in these tests. The objective of
TP 800/9 test was to demonstrate that the EFW pumps automatical-
ly start on loss of all reactor coolant pumps. Test re-
suits indicated that the test objective was met. After
verifying the test objective had been accomplished, steam
driven pump EF-P1 and motor driven pump EF-P2B were secured
and motor driven pump EF-P2A was kept running to support
plant operations. Subsequently, all four reactor coolant
pumps were restarted in sequence with no unusual plant re-
sponse observed.
Paragraph 4.3.2.2 below details further licensee action
concerning the eperation of EF-P1 which were identified
during performance of TP 800/9.
4.3.2.2 EFW Steam Driven Pump Relief Valve Actuation (Unresolved
Item 289/85-22-02
During inspection 50-289/85-22 it was identified that both
EF-P1 steam supply relief valves lifted. Although EF-P1
operated properly, there was a concern that relief valves
should not be challenged during routine operation and that
a stuck-open relief valve could affect EF-P1 operability.
This concern was identified as NRC unresolved item
50-289/85-22-02. Licensee actions concerning this
, problem are detailed in inspection reports 50-289/85-24 and
50-289/85-25, and paragraph 7 of this report.
j During previous troubleshooting it was noted that the re-
lief valves did not lift when only one steam supply valve
was opened. As an interim corrective action for the
remainder of Cycle 5, the licensee replaced pressure con-
troller PC-5 for steam pressure control valve MS-V6. MS-V6
_ -_
,
41
.
reduces steam pressure from a main steam header pressure of
greater than 900 psi to EFW steam supply pressure of less
than 200 psi.
In addition, EF-P1 steam supply valve MS-V13B was jumpered
out so that it would not open automatically with the open-
ing MS-V13A during an EFW actuation concurrently. If need-
ed, MS-V13B can be opened manually from the control room.
Keeping MS-V13B closed reduces the amount of steam to EF-P1
precluding the lifting of the relief valves. The steam
supply to EF-P1 and the operability cf EF-P1 were satisfac-
torily tested several times during the startup test
prc ram.
During performance of TP 800/9 pumps EF-P1, 2A and 28 auto-
matically started as required when all four reactor coolant
pumps were secured. Although EF-P1 performed satisfactori-
ly in the recirculation mode, licensee personnel were con-
cerned that there may have been insufficient steam supplied
to reach sufficient flow and pressure for EF-P1 to supply
proper EFW flow to operating steam generators. Based on
this concern, EF-P1 was declared inoperable and a 72 hour
limiting condition for operation was declared.
Licensee calculations indicated that MS-V6 must open to at
least 45% open to rupply a sufficient volume of steam for
EF-P1 to properly supply operating steam generators. The
MS-V6 controller was re-adjusted and the pump tested sever-
al times until an opening of 45% and steam supply pressure
of greater than 180 psig (but less than 200 psig) was
achieved. EF-P1 was satisfactorily tested and declared
operable prior to restart af ter the 88% power trip.
Item 50-289/85-22-02 remains unresolved until the licensee
establishes permanent corrective action concerning the op-
eration of MS-V13A and MS-V13B. The licensee has committed
to a permanent resolution prior to the startup after the
next refueling outage (Cycle 6).
4.4 RCS and OTSG Leak Rate
4.4.1 Effect of Reactor Trips from 88% and 22% Power on RCS and
OTSG Leak Rates
The effect of reactor trips from 88% and 22% power on RCS
and 0TSG 1eak rates was followed up by the TMI-l Restart
Staff. Based on data review of surveillance procedure (SP)
1301-1, " Shift and Daily Checks," aod SP 1303-1.1, " Reactor
Coolant System Leak Rate," the inspector noted that both
RCS and OTSG 1eak rates on the days following these trips
;
,
!
_-- -- - - - _ _ - . - -
.
42
(January 3-5, 1936) remained well within the the technical
specification's limits and were consistent with the previ-
ous results. No abnormal conditions wer observed.
4.4.2 RCS Leak Rate Change Evaluation
During a routine NRC shift inspector review of plant opera-
tion data, the inspector noted that the caiculated uniden-
tified leak rate varied from -0.12 gpm (2 a.m.) to -0.26
gpm (10 a.m.) on December 22, 1985. The shift inspector
noted that while the plant was operating at steady-state *
75% power level, the only plant condition that could affect
the leak rate result appeared to be a change in reactor
coolant drain tank (RCDT) pressure.
To resolve the abeve discrepancy, this matter was further
reviewed by the TMI-1 Restart Staff. The RCP No. 3 seal
purge flow is directed to the RCDT. Since this flow is
supplied from the reclaimed water system and is not classi-
fied as RCS leakage, its contribution to the RCDT mass in-
vec. tory change has been properly factored into the RCS leak
rate calculation. The seal purge ficw is assumed to be 100
cc/ min per pump or approximately 0.10 gpm for four RCPs.
This assumption is consistent with the actual seal purge
flow guidance provided in the RCP operation procedure
(CP) 1103-6. The seal purge flow local instrument is lo-
cated inside the reactor building. Due to difficulty in
determining the actual flow rate on a routine basis, and
because this term is small in nature, the value of 0.10 gpm
is therefore taken as a constant in the calculation. Since
the RCDT pressure affects the seal purge flow, increasing
the RCDT pressure may cause the actual flow rate to deviate
from the assumed value of 0.1 gpm which in turn may have
a small impact on the leak rate calculation.
The current RCS leak rate calculation used by the licensee
was actually field tested during hot functional testing
conducted in 1984. Under controlled environmental condi-
tions, the current method was proven to be able to detect
RCS leakage in the order of about 1 gpm with a 0.1 gpm
variation. The relatively large variation, as observed by
the shift inspector on December 22, 1985, was regarded as
due to the increase in RCDT pressure and the inherent
calcciation allowances associated with the current surveil-
lance technique, and therefore was determined to be a
normal plant response.
4.4.3 RCS Leak Rate Eva c orative Loss Term
Through various inspections conducted since the TMI-1 hot
functional test, the evaporative loss term was identified
as having contributed to the negative unidentified leak
43
rate result. In the routine tracking of RCS leakage during
the extended period of startup testing program further
confirmed this finding; that the correct evaporative loss
term for the current plant condition is essentially zero.
In the meeting between TMI-1 Restart Staff and the licensee
management held on January 2,1986, the licensee agreed
that this evaporative loss term will be set to zero in the
RCS leak rate calculation and will be implemented before
January 31, 1986, pending the completion of computer soft-
ware changes.
4.5 Verification of Licensee / Restart Conditions
4.5.1 Verification of License Conditions
'
Licensee condition 2.C.(8)3 states that "GPU Nuclear Corpo-
ration shall complete its post-critical test program at
each power range (0-5%, 5%-50%, 50%-100%) in conformance
with the program described in Topical Report 008, Revision
3, and shall have available the results of that test pro-
gram and a summ y of its management review, prior to as-
cension from each power range and prior to normal power
operation."
Management reviews for the power ascension test program
were documented in three reviews called flag reviews. The
first flag review (Flag 28) was conducted on May 9,1985,
and authorized TMI-1 to proceed with hot functional testing
through 48% power. Flag 2B verified plant readiness for
testing and was based on an anticipated restart during June
1985. Because of court decisions, restart of TMI-1 was
delayed twice. Based on these delays, the licensee updated
its Flag 28 approval on August 27, 1985, and on October 2,
1985.
The Flag 3 review was conducted on November 15, 1985 and
authorized the plant to proceed from 48% to 100% power.
Because of high steam generator levels the plant was only
able to achieve 88% power per TP 800/2 on January 2, 1986,
the licensee considered the power escalation test program
complete. Flag 4 was conducted on January 3,1986, and the
plant was authorized to restart and achieve the maximum
power for commercial operation.
On January 6, 1986, the plant achieved 100% power. Further
test data were taken after reaching 100% power; however no
further flag meetings were required.
The inspector reviewed the summaries of flag meetings and
noted that there was a high level of management attention
given to these meetings. Each flag review was signed by
_.
I
44
the President of GPU Nuclear, five vice presidents, and the
chairman of the General Office Review Board. All test data
results, including test exceptions and deficien ies, were
made available to the NRC Restart Staff shortly after data'
reduction and review had been completed. All test data
were subsequently reviewed by the Test Approval Group (TAG)
at formal meetings. The T/G consisted of representatives
from Startup and Test, Plant Staff, Nuclear Steam Supply
Systems, Technical Functions and Quality Assurance. In
addition, most TAG members were supervisory or managerial
personnel.
Although a flag meeting was not conducted at 5*; power be-
fore proceeding to 48*;, the staff concludes that the Flag
2B meeting prior to restart and the TAG review of test
results at 5*4,15'J, 25*; and 40*; test plateaus meet the
license condition. Based on the above review, the inspector
determined the licensee review of restart test programs was
in conformance with license condition 2.C.(8)3.
4.5.2 Veri'ication of Restart Conditions 4.5.2.1 Res: art Condition (2) (d)
Restart condition (2) (d) states that " Prior to completion
of the Power Escalation Test program, GPU Nuclear Corpora-
tion shall demonstrate safety grade automatic anticipatory
reactor trip on loss of feedwater and upon turbine trip."
The reactor trip on loss of feedsater was demonstrated dur-
ing the trip test per TP 800/2 at 40*. power. Thi s was ver-
ified by the NRC and documented in NRC Inspection Report
50-289/85-25.
Performance of the reactor trip at 88*; power per TP 800/2
on January 2,1986, demonstrated reactor trip upon turbine
trip. Based on the above, the licensee has met the re-
quirements of restart andition (2) (d).
4.5.2.2 Restart Condition (2) (e)
Restart condition (2) (e) states that " prior to completion
of the Power Escalation Test program, GPU Nuclear Corpora-
tion shall demonstrate performance of the saturation meter,
the incore thermocouples, and the wide range hot leg tem-
perature instrumentation systems installed to recognize
inadequate core cooling."
The saturation meter was tested initially in 1981 per TP
645-1, " Test Functional Test," and then retested in 1984
following modification to Class 1E qualified temperature
inputs, per TP 700/1, " Controlling Procedure for Low Power
45
Physics Testing," enclosure 16, "Tsat Margin Comparison."
Wide range hot leg temperatures were tested initially
in 1981 per TP 657/1, "Non-Nuclear Instrumentation
Hot Operation and Calibration," and were retested in 1983
following installation of Class 1E RTDs, per TP 700/1, en-
closure 12, " Data Collection Sheets for Modifications
1-38, LM-43C and RM-13J." Verification of satisfactory
erformance and data review for the above tests were accom-
plished during several previous NRC in:cections.
The incore thermocouples were tested during the power esca-
"
lation test program per TP 846/1, "Incore Thermocouple
Functional Test." Verification of final satisfactory per-
formance of this test is detailed in paragraph 4.2.2.4 of
this report.
Based on the above, the licensee has met the requirements
of restart condition (2) (e).
4.5.2.3 Restart Condition (2)(f)
Restart condition (2)(f) states that " Prior to completion
of the Power Escalation Test program, GPU Nuclear Corpora-
,
tion shall demonstrate EP4 initiation on loss of all four
reactor coolant pumps." Performance of TP 800/9 on January
2,1986, verified the automatic start of all three EFW
pumps upon the loss of all four reactor coolant pumps.
This test is discussed in paragraph 4.3.2. Based on this,
the licensee has met the requirements of restart condition
(2)(f).
4.6 Startup Testing Program Completion
4.6.1 Test Program Summary
The licensee's startup testing program from zero power
. physics testing to the low power natural circulation test
at 3?e power was controlled by the test sequence procedure
TP 700/1, and power escalation testing was controlled by TP
800/1. The testing program included performance of various
tests and was, on a sampling basis, found to be consistent
with the FSAR, technical specifications and specific licensee
restart condition requirements. The startup testing program
was successfully completed on January 2, 1986, following
the completion of the 88*4 power reactor trip test, and with
the licensee's subsequent management review.
As described in previous inspection reports (50-289/85-27
and 85-28), the steam generator level versus reactor power
tracked higher than expected during the power escalation
program, in that it limited the maximum achievable power
to about 88*. Unexpectedly, as TMI-1 returned to power
-.- - - .. -
- -.
46
on January 4,1986, the steam generator water levels no
longer imposed an operational limit on the reactor power
level. The reactor trip test and the subsequent unplanned
trip at 22% power, which occurred during the plant restart
on January 4,1986, apparently resulted in redistribution
of deposits in the steam generators. This apparent redis-
tribution allowed the plant to achieve a higher power level.
The plant achieved full power operation on January 6,1986.
The licensee conducted additional full power tests that
were deferred because only 88% power could be achieved prior
to the reactor trip test. Subsequent to this inspection
period, the TMI-1 Restart Staff independently reviewed the
additional test data and found the test results acceptable.
4.6.2 Overall Test Exceptions and Deficiency (E&D) Review
The licensee startup test and reactor engineering groups
have successfully completed the startup testing program.
Test results were properly reviewed by the cognizant engi-
neers and/or testing personnel. Final test results were
reviewed and approved by Test Approval Group (TAG). All
test exceptions and deficiencies have been identified and
resolved, and resolutions were acceptable.
. 4.6.3 Restart Test Program Evaluation
On April 5, 1933, the licensee, by letter, submitted to the
NRC a description of its test program following NRC author-
ization to restart the TMI Unit I reactor. The test plan-
ning specification (SP-1101-06-008) was attached to this
letter and identified the items to be tested; the plant
status for testing; the formal commitment for the test; the
test scope and acceptance criteria; and the test procedure
which would perform the test.
During this inspection, the inspector reviewed the
licensee's test planning specification to determine if all
commitments identified for the power escalation test (PET)
program had been met.
In general, the licensee had met its commitments as defined
in the planning specification. However, some discrepancies
were noted and are identified below.
4.6.3.1 The test for modification RM-5C, " Containment Isolation on
Reactor Trip," requires the verification of proper contain-
ment isolation following the 40% and 100% reactor trips and
that the override and bypass switches allow valves to re-
open. The inspector noted that TP 800/2 failed to document
the reset feature of containment isolation valves. The
reset feature currently caly applies to one valve, MU-V3,
.- ._. . _- . _ _
47
" Letdown Isolation Valve." Resolution of this issue is
discussed in paragraph 4.3.1.2.1 of this report.
4.6.3.2 The licensee stated that turbine generator (TG) operation
would be tested per TP 8S5-1. TP 885-1 was to verify ac-
ceptability of various TG acceleration, speed and loading
parameters and characteristics and to verify proper
performance of the TG versus load at (15, 40, 76 and 100%
power). A review of TP 885-1 indicates that only steady-
state parameters such as temperature, thermocouple perfor-
mance, and vibration measurements were performed at various
power levels.
A licensee representative stated that turbine load testing
parameters were tested per a special GE load test performed
by GE during hot functional testing. The licensee provided
to the inspector documentation of turbine generator load
testing, " Control Systems Checkout Prior to Restart," which
was performed during the period of October 27, 1981, through
December 18, 1981. This testing performed a comprehensive
checkout of turbine controls under load conditions.
Based on the above documentation, the inspector considers
this test commitment to have been satisfied.
4.6.3.3 Turbine bypass valve testing acceptance criteria in the
test specification state that turbine bypass valves shall
open in < 3 seconds at < 1100 psia peak pressure. The pro-
cedure allows a peak pressure of < 1155 psig. A licensee
representative states that pressures were based on startup
test, TP 800/6 performed in 1974. TP 800/6 acceptance cri-
teria stated peak OTSG pressure should be < 1100 psia (1085
psig) and shall be < 1180 (1155 psig). These acceptance
criteria were not clearly translated into the planning
specification.
For the test (TP 885/2) the actual peak pressure recorded
by the transient monitor during the 40% and 88% trips did
not exceed 1060 psig. Because the main steam safety valves
'
controlled pressure (see paragraph 4.3.1.2.2), the opening
times of the turbine bypass valves could not be measured.
This is a test deficiency which, when closed, will be
resolved by the licensee. Based on the above, the inspector
considers that the intent of this test commitment has been
met.
6.1 4 The licensee stated that they would perform a natural
circulation level control test to determine the lowest lev-
el in the OTSG that sustains natural circulation flow while
on main feedwater. This was to be performed by TP 800/9
following the reactor trip from (an anticipated) trip at
..
- . _ . __ - _. _ - _ _ .
- ..
-
. . .-
. _ _ .-
48
100% power. This test was to be performed at the request of
the licensee's engineering organization.
During the startup test program several weeks prior to the
performance of the final trip test per TP 800/2, the
licensee notified the NRC Restart Staff verbally that the
natural recirculation level control portions of TP 800/9
would be deleted. This was acceptable to the restart staff
since this test was to be performed at the licensee's dis-
cretion and was not based on any formal regulatory commit-
ment or requirement.
A natural circulation level control test was performed
during the natural circulation test at 3% power during
performance of TP 700/2. During this test OTSG level was
controlled by emergency feedwater and natural circulation
was lost at approximately 35% OTSG 1evel on the A steam
generator.
Licensee representatives stated that sufficient data were
obtained during TP 700/2 because further natural circula-
tion level control testing was unnecessary. Based on the
above, the inspector considers that the licensee has met
the intent of this commitment.
4.6.3.5 As part of its restart commitment, the licensee stated
that after performance of TP 800/8, "RCS Overcooling Con-
trol Test," operating procedure (0P) 1102-16, "RCS Natural
Circulation Cooling," would be reviewed to ensure that it
provides adequate guidance to prevent overcooling during
.
!
any natural circulation cooling. The inspector questioned
whether or not this had been done as this commitment was
,
not part of the test program itself.
Discussions with licensee representatives indicated that
i the Technical Functions Division located at the corporate
;
office had the lead responsibility for changing this proce-
dure. The inspector reviewed several memoranda by Tech-
nical Functions indicating that extensive reviews were being
done as result of data obtained from the test. Simulator
programs and operator training will be upgraded as a result
- of data obtained from tests TP 700/2 and TP 800/8. Howev-
er, these memoranda do not address plant procedures.
The cognizant engineer stated that he was aware of the com-
mitment and was planning to revise OP 1102-16 in the near
future. To ensure that this commitment would be tracked,
'
the Manager of TMI-1 Startup and Test, Technical Functions,
submitted a technical functions work / task request (TR No.
BT 4410), which requests that the results of TP 700/2 and
.
- - . . - . , - . , - ,. -
, , -,-
- _ -
49
TP 800/8 be reviewed and that procedural recommendations be
incorporated into OP 1102-16. The inspector had no further
questions concerning this item.
4.7 Ouality Assurance / Quality control (0A/0C) Interface With the Startup
Test Program
Throughout the startup test program, the inspectors observed QA per-
sonnel witnessing test performance and reviewing test procedures
while in progress. The inspector discussed the test inspection pro-
gram with QA management, interviewed QA inspectors on shift and re-
viewed documentation of completed shift monitoring. QA overview of
test programs was under the cognizance of the operations QA manager
and performed by shif t monitoring engineers (SMEs).
SMEs monitor selected activities in the plant and there was at least
l one SME on shif t at all times. During testing, when needed, two SMEs
would be assigned to a shift. SMEs are engineering graduates who
have qualified for the position. During th test program several QC
i inspectors were qualified as SMEs in order to augment test program
coverage. There was at least one SME assigned exclusively to cover
1
i
the test program. Rather than oerform audits or per' arm specific
inspection points, SMEs performed a monitoring overview. SMEs on a
sampling basis reviewed data, verified test procedure compliance,
'
assured that the latest revision of procedures were in use and veri-
fied test equipment calibration.
Prior to the start of the test program, the operations QA section
developed a detailed monitoring plan. This plan detailed areas to be
monitored in each phase of testing, assigned responsibilities, and
documented subsequent monitoring by recording QA monitoring report
(QAMR) numbers. The inspector noted by documentation of QAMRs that
extensive monitoring was done in each area. In addition, the inspec-
tor reviewed seven randomly selected QAMRs which were accomplished
during different phases of the test program. The QAMRs were compre-
hensive, in-depth, and indicative of detailed QA inspection coverage.
In addition to tests in progress, QA reviewed all completed test data
and exceptions and deficiencies. The operations QA manager is a mem-
ber of the test acceptance group (TAG). QA has an input to the
fi acceptability of test results. As a result of QA participating
in IAG, 100% of all completed data is reviewed. In addition to test
procedures, QA provided documentation that license conditions, startup
conditions, and other regulatory commitments made by GPU Nuclear were
being met. Periodic memoranda are forwarded to the Vice President /
Director of TMI-1 concerning the status of GPUN's ef forts to satisfy
licensing commitments to the NRC.
Based on inspector observations of QA monitoring and extensive docu-
mentation of QA monitoring, the inspector considered the monitoring
program to be thorough and extensive.
,
7--- - - , - , _ , - - - - . , ,, - _ _ - - - - , - _ - - . , - - - - . - - - - - - - - - - - , , - , , , - - - -
50
4.8 Conclusions
The nuclear engineering group consistently performed well throughout
the entire startup testing period. The physics testing program was
well prepared and executed. The test data were properly evaluated
and information was disseminated to the appropriate groups. The
startup testing group also performed well in coordinating and identify-
ing plant equipment problems. Equipment or procedural problems
identified in the early stages of testing were quickly corrected and
implemented during the remaining testing period.
Overall, the testing program was effective and provided an adequate
test of overall plant equipment performance, plant operator perfor-
mance, and the ability of safety systems to perform adequately.
5. Event Review
5.1 Reactor Trip / post-Trip Review
The resident and region-based inspectors attended the licensee's post-
, trip review of the trips on January 2 and January 4,1986. This re-
view was required by administrative procedure 1063, " Reactor Trip
Review Process." The January 2, 1986, reactor trip review identified
several problems that were corrected or appropriate corrective action
was initiated by the licensee. Additionally, the review noted that
portions of the computer program (transient monitor) used to compute
sequence of events did not function. The sequence of events computer
program was subsequently retested successfully and the licensee was
unable to reproduce the problem. The sequence of events monitor
functioned properly during the January 4,1986, trip, but the licensee
decided to closely monitor its function on a daily basis;
The cause of the trip on January 2, 1986, was a manual turbine
trip / reactor trip as part of the power escalation test sequence which
was described in paragraph 4. The trip on January 4,1986, was due
to failure of the level controller associated with a feedwater heater
drain tank which resulted in high level in a moisture separator tank.
The moisture separator 1-of-1 instrument logic provided a turbine
trip which caused the reactor trip.
For both trips, the inspector independently reviewed the completed
enclosure 1, " Post Trip Review of AP 1063," and the recorder strip
charts and verified that the data required by enclosure 1 were re-
trieved with no significant deficiencies noted. The inspector con-
cluded that the procedure adequately evaluated plant performance to
the extent necessary to reach a decision related to startup. In ad-
dition, the inspector confirmed the l'censee's conclusion,
i
51
5.2 Radioactive Gas Release from the Waste Gas System
5.2.1 Event Chronology
On December 17, 1985, between 7:40 and 8:00 p.m., an uncon-
trciled release of radioactive gases occurred in the waste
gas compressor cubicle. The plant was at steady-state 75%
power and the reactor coolant system was at normal operat-
ing temperature and pressure.
Repair personnel were logged in on RWP 31377 to perform
preventive maintenance on the 'A' waste gas compressor
(WDG-P-1A). The waste gas compressor system was tagged out
in accordance with the switching and tagging order. Howev-
er, a decision was made on December 16, 1985 by the shift
supervisor, shift maintenance foreman and radiological
controls supervisor to use check valve WDG-V-20 for isola-
tion from the pressurized vent header. Isolation valve
WDG-V-100, which is downstream of the check valve, was not
closed because preventive maintenance was also being
performed on that valve.
A radiological controls technician was assigned to monitor
the job. A carboy was used to drain the separator tank so
that any releases would be limited to noble gases, as re-
quired by the RWP. While drain'ng, an increase was noted
on the RM-14 monitor. A local gas sample was obtained, and
results indicated the presence of Xe-123 at 3.73 E-7
uCi/cc. No further problems were foreseen, so the job was
continued on the basis of these dos,e rates.
At 7:40 p.m., a strainer was removed from the system and
prompted an immediate response on radiation monitors. At
7:45 p.m., monitors RM-A6 (particulate and gas) and RM-A8
(gas) showed increases. The vent header was checked and a
small pressure drop was indicated. Steps were taken to
stop the job and return the system to a safe condition. At
7:55 p.m., RM-A6 (particulate) was changed out, but the
monitor continued to rise. Local gas sampling was taken.
The shift foreman and shift maintenance foreman entered the
cubicle to determine how to stop the release as the radio-
logical controls technician was removing the other people
from the area. The check valve was determined to be leak-
ing, so the strainer was reinstalled at 8:00 p.m.; this
stopped the release.
5.2.2 Sc_ ope of Review
The inspectors reviewed the details of the leak and the
licensee's review of this event to determine:
52
--
details regarding the cause of the event and event
chronology;
--
consistency of licensee actions with NRC license and
procedural requirements; and,
--
licensee actions to correct the cause of the event.
The inspector's review of this incident included discus-
sions with cognizant licensee personnel and review of the
following documents:
--
system drawings;
--
CR0 and SF logs;
--
switching order 85-1882, dated December 15, 1985;
--
administrative procedure (AP) 1044, " Event Review and
Reporting Requirements," Revision 14, dated June 4,
1985;
--
various radiological data and calculations;
--
RWP #31377; and,
--
critique meeting minutes.
The inspectors also attended the radiological investigative
critique that was held on December 18, 1985.
5.2.3 Licensee Findings / Actions
The cause of the event was leakage past a check valve that
was being used for isolation purposes. Immediate correc-
tive actions were to close the system by reinstalling the
strainer into the heat exchanger, and closing valve
WDG-V-100, for further isolation.
The RM-A6 (particulate and gas) and RMA-8 (gas) monitors
showed maximum increases from 300 cpm to 2000 cpm, 80 cpm
to 200 cpm, and 90 cpm to 220 cpm, respectively. These
values were below alert setpoints.
A total of eight people were determined to have some amount
of contamination, either to skin, clothing, or both mostly
due to a short lived noble gas decay. The licensee esti-
mated personnel exposures of aoout 4E-5 millirems from the
incident, which is significantly lower thari NRC require-
ments. The amount released to the environment was estimat-
ed to be less than 0.001 percent of the quarterly limit for
noble gases.
53
As part of the licensee's followup actions, this event will
be presented to other shifts through the use of the required
reading book. A procedure change request will allow
WDG-V-100 to be worked on at a more opportune time, when
the probability of a release will be reduced. Also, pre-
ventive maintenance procedures for gas compressor work will
be split up to ensure adequate isolation is provided in the
future.
5.2.4 NRC Operational Assessment
Licensee immediate corrective actions during the event were
appropriate. Followup corrective actions from the
licensee's self-review were also appropriate and reasonably
thorough. During and after the event, the licensee fol-
lowed procedures adequately. However, the inspector con-
cluded that licensee representatives used poor judgement in
relying on a check valve for isolation purposes. They
should have completed preventive maintenance on WDG-V-100
before or after performing preventive maintenance on the
waste gas compressor; therefore, this valve could have been
used to isolate the system to preclude the radiological
release.
5.2.5 NRC Radiolocical Assessment
The NRC confirmed the licensee's estimate of activity re-
leased to be about 1.4 curies over a 54 minute duration.
Offsite exposure due to this event was calculated to be
about 1.35 E-3, total dose, based on very conservative as-
sumptions, compared to the licensee's values of 5.2 E-5
mrad gamma and 1.2 E-4 mrad beta, derived from the
licensee's offsite dose calculation manual. Licensee ac-
tion relative to recovery, radiological controls and evalu-
ation of personnel exposure were found to be adequate and
in accordance with regulatory requirements. This release
constituted less than 0.001*. of the quarterly release limit
specified in technical specifications.
5.3 Makeup Pump Leak
5.3.1 Event Chronology
At 1:26 a.m., makeup pump 1A was started to supply normal
makeup to the reactor coolant system while realigning elec-
trical power to makeup pump 18. At 1:31 a.m., an auxiliary
operator reported slight leakage from mechanical seals on
makeup pump 1A and increased radiation activity levels in
the makeup pump cubicle. At 1:35 a.m., auxiliary building
radiation monitor RM-A6, gas and particulate channels,
reached the alert level. At approximately 1:38 a.m., the
inboard mechanical seal on makeup pump 1A failed and water
-._-- .- - . _. - ._. . _ _ __ . , . . _ _ _ . - .
,
54
a
was sprayed onto the walls of the cubicle. The plant oper-
ators restarted makeup pump 18 and secured makeup pump 1A.
Approximately 300 gallons of water was spilled onto the
floor. At 1:42 a.m. the combined fuel handling building
and auxiliary building vent radiation monitor RM-A8, gas
channel, reached the alert level. Based on the fact that
both RM-A6 and RM-A8 were in an alert status, the shift
supervisor declared an unusual event at 1:50 a.m.
The shift supervisor directed that makeup pump 1A be iso-
lated and drained. Station health physics personnel began
to clean up the room.
Both radiation monitors immediately began to decrease; the
RM-A8 alarm cleared at 2:00 a.m. and the RM-A6 alarm
cleared at 2:27 a.m. The licensee decided not to secure
i '
from the unusual event condition since a planned pump
draining evolution had the potential for additional radio-
active releases. At 4:00 a.m., makeup pump 1A had been
crained and the unusual event was terminated at 4:20 a.m.
As a result of the release, nine people were found to be
contaminated. Of the nine, three were found to have actual
skin contamination (two on the skin and one on his hair).
The two with actual contamination on their kin were decon-
<
taminated using soap and water.
5.3.2 Scope of Resiew
The inspector reviewed the details of the leak and the
licensee's review of this event to determine:
--
details regarding the cause of the event and event
chronology;
--
consistency of licensee actions with NRC license and
procedural requirements; and
--
proposed licensee actions to correct the cause of the
event.
The inspector's review of this incident included discus-
sions with cognizant licensee personnel and review of the
following documents:
--
systems drawings
, --
CR0 and SF logs
--
applicable RWP and radiological assessment report
--
critique meeting minutes
i
,
-
. _ - - . ._. _ . _ _ _ . _ . _ . _ _ . _ _ _ . _ . _ _ _ __
_ , _ . . _ _ _ _ _ _ , , . _ . _ . _ . _ _ .
- . - . - . _ - - _ . . - . .- __ -_ __
55
5.3.3 Licensee Findines/ Actions
The licensee's investigation determined that makeup pump 1A
(MU-P-1A) experienced an inboard mechanical seal failure.
The seal failure resulted from a loose thrust bearing lock-
l nut. This resulted in the loss of the thrust bearings
function, allowing the pump shaft to move axially and to
'
cause the seal failure. The locknut had been secured with
a set screw, which had loosened allowing the locknut to
subsequently loosen. In order to prevent a recurrence of
this failure, the licensee dimpled the shaft at the loca-
tion of the set screw and applied "Locktite" (permatex
sealant) to the shaft and set screw. Radiological releases
were within regulatory requirements.
5.3.4 NRC Operational Assessment
Review of licensee actions during the event were found to
be appropriate and proper use of procedures was noted. The
root cause was appropriately determined to be a failed me- I
chanical seal caused by a makeup pump thrust bearing shift.
The inspector reviewed the engineering evaluation for the
dimpling of the shaft and the application of the sealant.
Also, the actual work that implemented the modification was
witnessed. The work was found to be adequately controlled,
, supervised, and performed by experienced and knowledgeable ,
workers. Quality control personnel also observed the work. '
One minor problem arose with regard to the use of Permatex
sealant on the thrust bearing housing cap and end plate
joints. All three pumps had previously been sealed with
Permatex in these areas. During reassembly of the 1A pump,
the use of Permatex sealant was not approved by QC; and,
therefore, it subsequently leaked minutely during testing.
The 18 and IC pumps were made up using an approved sealant.
The licensee will review the tasks that used Permatex to
determine if it is acceptable. Following the completion of
the modification, all three pumps were tested prior to
being declared operable. The inspector had no further
questions with regard to this matter.
During observation of the work being performed on MU-P-1A,
it was noted that the installed thrust bearing locknut was
made of brass, while the vendor's manual for the pump spec-
ified the thrust bearing locknut to be steel (SAE 1020).
The licensee contacted the vendor and determined that four
_- -- -. . - . _ - - - . , _ - _ _ - - - - . - - - _ - - - . . - . - - . --- -
_ __ _-- -- __ _ __ _ . _ _ _ _ _ _ . _ - _ __ ..__
4
56
replacement locknuts made of brass, instead of steel as or- l
dered in 1974, had been mistakenly shipped. This was at-
tributed to human error in reading the equipment material
list; brass is the specified material for the item directly
belcw the locknut on the material list. The locknuts in-
stalled on MU-P-1B and 1C were verified to be made of
steel.
An engineering evaluation was performed to justify the use
of a brass locknut on the 1A pump. This evaluation used
4 the pump manufacturer's calculations using worst case axial j thrust loads; the evaluation showed that the brass locknut
would not be subjected to shear failure of its threads.
1
The brass locknut will be replaced with a steel locknut at
- the earliest convenient opportunity.
From the licensee's operational and engineering review, the
, root cause was identified and corrective actions were ,
appropriate.
5.3.5 NRC Radiological Assessment
i
Analyses of the event indicated that the predominant radio-
logical contributor was noble gas activity; i.e., radioac-
tive krypton and xenon gas. As a result, about nine people
in the auxiliary building were contaminated with particu-
late noble gas decay daughters; i.e., Rb-88 and Cs-138,
which caused negligible exposure. Workers involved in re-
covery and clean-up operations received less than 50 mrem
exposure for the event.
The spray caused a spill of about 300 gallons of reactor
! coolant which was collected via the floor drain system in
the auxiliary building sump. Off gasing of the noble gas
constituents during the event resulted in release of noble
gas activity through the station vent for about 74 minutes.
Evaluation of the vent monitors indicated that about 46
curies of noble gas activity, principally Xe-133, were
released. Off-site dose consequences of this release were
calculated to be 3.69 E-3 mrad gamma and 5.97 E-3 mrad beta
at the site boundary. Such exposure constitutes 0.07% of
the quarterly limit permitted by the license technical
specifications; i.e., 5 mrad gamma air dose and 10 mrad
beta air dose. These values were independently confirmed
by the NRC inspector.
1
-- - - - -- ._.. __ _- ._ __. -_ -. -. . .- _ --
57
The licensee's followup and review of the occurrence was
thorough and timely. Radiological control coverage was
sufficient to rapidly identify the hazard and minimize
'
personnel exposure, The licensee's critique of the event
identified areas in which response should be improved and
action was initiated to implement lessons learned.
No conditions adverse to regulatory requirements were iden- i
tified in this area.
1
l 5.4 Conclusion !
Operators were responsive to these events with respect to
isolating the radiological release and/or following appro-
priate plant and emergency procedures. The licensee's i
event review or critique was thorough enough to identify
a tne root cause of the events. Appropriate corrective at-
tion was taken or planned. Licensee dose calculations were
I conservative and independently verified by the TMI-1 Re- l start Staff. '
6. Radiation Protection
6.1 Radiological Program Area Review
The licensee's program was reviewed against, (1) criteria contained
in 10 CFR 20 relative to personnel exposure control and radiological i
!
surveillance, (2) applicable license conditions relative to internal
,
audits and assessmen+,s and procedure establishment and implementa-
1
tion, and (3) personnel qualification and training. Additionally,
j management controls were reviewed.
Adequate procedures and protocols are established and implemented to
control personnel exposure both in plant and offsite. The licensee
is consistently conservative with regard to dose assessments, and
,
generally exercises good judgement in analyzing and assessing events
I which have the potential to cause increased exposure.
l
!
Internal audits for the 1935 period were reviewed relative to the
Radiation Protection Plan. The audits were conducted by qualified
personnel and exhibited sufficient technical depth to evaluate pro-
gram performance. Though most audit findings identified in this pe-
riod were administrative in nature, sufficient technical insight was
demonstrated.
However, while this particular audit process is formalized and car-
ried out in accordance with schedules, there is no assurance that all
4 program areas will receive sufficient review, since the audit scope
is not tracked. Rather, a general audit assignment is made and the
! i
<
1
l
i
i
.. . . __ -. .. - _ = . - - - - _ - - - - . - - - . .
L
53
i
!
l individual auditors decide on the scope and depth of the assessment,
. with management approval. There is no system currently in place that
!
tracks audits of individual program elements that can be utilized to
; verify that the entire program as defined by the Radiation Protection
i
Plan is subject to periodic critical review. This item will be
reviewed in a subsequent inspection (289/85-30-03).
! Radiological surveys are performed thoroughly with sufficient detail
j to support radiological work. Current surveys are conspicuously
'
posted at the control point to support worker briefings and at indi-
1 vidual radiologically controlled areas as a reference to actual
I conditions. '
Personnel exposure control practices involving RWP usage appeared
sufficient to prohibit inadvertent, unanticipated or unplanned per-
sonnel exposure, provided personnel conform to established policies
and procedures. During this inspection, the licensee identified that
an individual inadvertently entered an area requiring an RWP without
j such authori:ation, due to misplaced area posting. Immediate and
'
adequate action was taken to assure that the area was adequately con- s
'
trolled and exposure to the individual assessed. In this event, the
individual received less than 10 millirem due to the occurrence. The
i licensee initiated action to prevent recurrence. Though technically , 1
a violation of technical specification 6.13, "High Radiation A eas," '
this event meets all of the criteria for licensee identified events
specified in 10 CFR 2, Appendix C, and will not be cited.
- The licensee's training and qualification program was reviewed and
- appears sufficient to assure competency and technical proficiency
among radiation protection personnel. While actual operating experi-
ence is limited in the department, sufficient management oversight
and control is exercised in an effort to compensate.
i
Management personnel demonstrate a high degree of involvement and
- understanding of plant conditions, communicate effectively internally
and inter-departmentally, and appear to plan, organize and direct
effectively.
! No conditions adverse to regulatory requirements were found in any of
the areas reviewed.
! 6.2 Control of Work Activities
- The licensee's program for controlling work activities and worker
i
exposure was evaluated by interviews with selected members of the
radiation protection staff, and revies of the following documents and
procedures.
,
i i
_ _ _ - - - . - .-. . - - . . . -_ _ . . , - , . - - . _ - . , . . . - - . .
__ _ .__ - - _ _ _ - . - - _ - _ . - _ _ _ _ - . . -_.--.-__.. - -- .---
i
59
4
--
procedure 9100-ADM-4110.04, " Radiation Work Permit"
: --
RWP # 31259, " Remove FS Detector 8-2 from Service," and associ-
ated ALARA review
,
4
--
RWP # 031201, " Change MUF 2A&B and MUF 1A&B," and associated
ALARA review
1 --
RWP # 031202, "Decon Letdown Prefilter Cubicle and Hallway"
P
j
--
RWP # 031204, " Remove Broken Filter from Letdown Filter Room"
--
selected radiological surveys associated with the above RWP
packages
.
l
--
selected RWP signature verification and entry-exit times
,
--
the health physics air sampling log book
, The licensee's radiation work permit (RWP) system specifies rcdiolog-
*
ical controls for specific and routine work activities in radiologi-
, cal areas. The inspector reviewed the RWPs mentioned above, along
*
with selected associated documentation and determined that:
--
radiological surveys used in the development of RWPs were gener-
ally of sufficient depth and scope;
l
--
radiological controls outlined in the RWPs were appropriate for
the work being performed;
--
administrative requirements of the controlling RWP procedure (HP
] supervisory review, pre work briefings, etc.) were being com-
plied with;
--
air sampling was performed as required by various RWPs; and,
--
workers wearing respiratory protection on selected RWPs had re-
! ceived required respiratory protection training and fit-testing. , i .
Overall licensee control of work activities and worker exposure was , found to be adequate. The inspector noted the licensee's method of
l choice for job specific air sampling is to provide lapel air samplers ;
to the work parties. This ensures a representative breathing zone '
air sample is collected and used to calculate MPC hours.
6.3 Posting and Area _ Control
a ! The licensee's posting and control of airborne radioactivity areas, 4
contaminated areas, and radiation and high radiation areas was re-
- viewed with respect to the following criteria.
} i
Technical Specification 6.12, "High Radiation Area"
'
--
>
- - - - - - - - - _ - - . - - - - - - - . - - - . . , - - , . ., , - - -
, , . , . , - -- . _ - , . . - - , - _ , - - - . . . . . . , - - -
- - - - - - . - - - - _-. - - - . . - - - _ - _ - - t !
l
60 t
j . l .
--
10 CFR 20, " Standards for Protection Against Radiation"
The licensee's performance in this area was reviewed by the following
,
methods:
1 --
inspector tours of the various radiological work and storage
-
areas
! {
--
performance of independent surveys by the inspector
2
--
interviews of radiation protection supervisory and
technician-level personnel
Housekeeping in the licensee's radiological areas was generally good.
l The inspector noted the licensee has implemented the guidance con-
tained in NRC IE Notice No. 84-82, " Guidance for Posting Radiation
Areas," in that posting has in general been limited to small,
discrete areas around the radioactive source, rather than simply
posting at cubicle doorways. This effort by the licensee to minimize
over posting helps prevent worker desensitization to radiological
'
posting.
i
- This inspection effort included tours of the auxiliary building and inde-
pencent survey verification of licensee radiological boundaries. The
j inspector noted during a survey of the mini-valve alley on the 286
foot elevation that a high radiation area posting inside the cubicle
was not appropriate. A "high radiation area" sign and posting is
required to indicate the presence of a high radiation area, i.e. an
area where radiation dose-rates exceed 100 mrem /hr. Consequently,
the boundary and posting for the high radiation area should be locat-
ed outside the area, where dose-rates are less than or at 100
,
mrem /hr, to alert personnel.
>
1
'
The inspector determined that the radiation field inside the cubicle
i
was non-uniform and that while the licensee's boundary was
adequately placed when dose-rate measurements were made at waist lev-
] el, dose-rates greater than 100 mrem /hr could be measured outside the
'
high radiation area boundary at the head level.
4 Immediately the licensee satisfactorily reposted the cubicle when
informed of the inspector's findings. Subsequent surveys in the cu-
i
bicle and comparison with previous routine surveys indicated that
dose-rate conditions had changed in the cubicle in the interval be-
,
tween the last licensee survey and the inspector survey in the
cubicle.
The licensee indicated that HP technicians are instructed to make
" general area" dose rate measurements by recording the highest
waist-level dose reading in a 360 inch circle. The licensee indicat-
'
!
ed some technicians may have extended this technique to the posting
of radiation and high radiation areas, using waist-level dose rate
I
i
j
.- ,- . - . - - - . - - - - . .- . - , - - - . . - - - - - , . - , - - - - - - - - - - , . - - - - - - - - - .
61
readings. The licensee indicated that technicians would be notified
as to the posting problem indicated above and the need to survey fcr
dose to the whole body and post accordingly. Licensee followup ac-
tion will be reviewed during a subsequent inspection (289/85-30-04).
Within the scope of the above review, no conditions adverse to regu-
latory requirements were identified.
6.4 Recurrence of Noble Gas Contamination of Personnel
During this review it was noted that occurrences involving the use of
the auxiliary building floor and equipment drain system to direct
reactor coolant runoff to the auxiliary building sump of ten result in
personnel contamination. Such contamination is due to the shortlived
particulate daughters of the noble gases Xe-138 and Kr-88; i.e.,
Cs-138 and Rb-88, respectively.
While individual exposure to this type of contamination is generally
not significant due to the short decay period of Cs-138 and Rb-88,
the recurrence of events in which large portions of the auxiliary and
fuel handling buildings are affected by noble gas activity is not nor-
mal to the operation of the facility.
In an effort to understand the conditions that cause extensive noble
gas migration in the auxiliary and fuel handling building, the
licensee initiated actions to test floor and equipment drain systems.
As a result, a drain header has been recently identified as being a
primary conduit of noble gas through the auxiliary building and into'
the TMI-2 portion of the fuel handling building.
In an effort to restrict and control this pathway, suspect drains
were either incorporated into the loop seal surveillance program,
sealed or covered. Additionally, action has been initiated to detect
and seal major leakage points through the Unit 1/ Unit 2 environmental
barrier. These efforts should prevent the spread of Unit 1 airborne
activity to Unit 2. However, noble gas involvement of large portions
of the auxiliary building due to localized activities such as flushing
operations, reactor coolant sampling, and component leakage, is still
not under control and results in the frequent contamination of several
personnel.
Though the licensee is investigating various mechanisms that may be
responsible for this problem, there is no concerted or organized ef-
fort yet developed to assure the timely completion of investigative
action items, the analyses of results, and the development of an ef-
fective resolution.
This item will be reviewed in a subsequent inspection and is current-
ly being followed by NRC staf f (289/84-16-01).
4
.
A
.
62
7 '. Emergency Feedwater Pump (EF-P1) Operability
.
7.1 Review
'
The ilispector assessed the operability of the turbine-driven emergen-
cy feedwater pump based on a revtew of licensee maintenance (preven-
tive and corrective) and surveillance activities to verify that:
--
procedures required by Technical Specifications (TS) 4.2.2 and
4.2.1.1 are'being properly implemented;
--
applicable procedures have the properiformat and technical con-
tent in accordance with the ASME BoileFarM Pressure Vessel
Code, Section XI, Subsection IWP, and applicable sections of
ANSI N18.7-1976;
--
surveillances and preventive maintenance (PM) were conducted at
the proper frequency; and,
--
machinery history records and related surveillance and PM
records were retrievable.
In addition to discussions with cognizant licensee personnel (mainte-
nance, operation, and en;ineering), the inspector reviewed selected
portions of the folicwing'iicensee documep p and records:
--
surveillance procedure (SP) 1300-3G A/B, " Turbine Driven mer-
gency Feedwater Pump Functional Test and Valve Operability
Test," Revision 20, dated November 13, 1985, including data ob-
tained on December 11, 1985, November 13, 1985, October 13,
1935, and September 15, 1985;
'
--
PM' procedure 1410-P-3, " Lube 011 Change,'! Revision 6, dated Feb-
ruary 2, 1935;
--
selected job tickets from machinery history files, numbers
C8969, CE756, CF763, and CH760 between the period of January 1,
1973 to November 5, 1985; and,
--
Worthington Pump International Technical Manual 2134-EIC, "WT
Pumps" and Technical Manual W-9C, " Turbine."
7.2 Findings
On December 16,1985. EF-P1 was out of service to test a new governor
valve controller. This controller was bsing tested to determine if
it would prevent the relief valves on the steam line to the EF-P1
turbine from lif ting durir,g the start sequence and if it could ade-
quately control EF-P1 during operation. The inspector reviewed tag-
ging order 85-1881 that was used to place EF-P1 out of service. The
inspector determined that appropriate valves were tagged locally and
in the control room and appropriate breakers were opened.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___
l
63
l
- While the licensee was testing the new controller, the inspector ob-
!
served good work practices, except for two instances when workers
used the limitorque operator for MS-V-10B as a stepping point. Dur-
,
ing testing, the controller did not perform satisfactorily and the
licensee replaced the new controller with the old controller. In
'
order to return EF-P1 back to service, SP 1300 3G A/B was used. The
inspector reviewed this package and no discrepancies were noted.
The review of machinery history on EF-P1 revealed no major problems
i
over the last twelve years, except for a bearing replacement during
j the spring of 1985. This history file provided a useful summary of
work activities on the turbine-driven EFW pump. The job ticket
records were retrievable using the licensee's microfiche and micro-
film systems.
Preventive maintenance procedures appropriately reflected vendor ret-
ommendations to change lubricating fluids at six-month intervals on
both the pump and turbine. The pump manual also recommended that the
pump should not be routinely disassembled; this should only be per-
formed when there is a problem with the pump. The licensee is plan-
ning to follow this recommendation.
Surveillance and preventive maintenance procedures were current and
being conducted at the proper frequency in accordance with technical
specifications. The records of these tests were retrievable. SP
1300 3G A/B was found to meet all applicable requirements of Subsec-
tion IWP, Section XI of the ASME Code.
7.3 Conclusions
The inspector concluded that the licensee properly tagged out of ser-
vice EF-P1 and appropriately returned it to service. Licensee
records were well kept and reflected applicable procedures that were
being properly implemented. The licensee is continuing to follow the
recurring problem with the EF-P1 steam relief valves. Pending
satisfactory resolution of this issue, in place surveillance and pre-
ventive maintenance procedures should provide adequate reliability
that the turbine-driven emergency feedwater pump will be operable
when needed.
8. Non-Licensed Staff Training
8.1 General
The purpose of this portion of the inspection was to assess the ef-
fectiveness of the licensee's non-licensed training. As a represen-
tative sample, the inspectors chose the training programs for
instrumentation and control (I&C) technicians, electrical and mechan-
ical maintenance personnel, and auxiliary operators (A0s). The in-
spectors noted that the licensee is in the process of obtaining INP0
_ _ _ _ _ _ _ _ _ _ _ _
64
accreditation for these training programs. The effectiveness of the
existing program was determined by reviewing the following
activities:
--
monthly reactor protection system surveillance by the I&C
technicians
--
shif t turnover for A0s (0700 twice,1500 twice, 2300 twice)
--
corrective maintenance (CM) for hydrogen analyzer by I&C techni-
cians (CM-1430-HA-2, paragraph 6.10)
--
A0 activities on all three shif t (changing over the hydrogen
supply to the makeup tank and taking secondary plant readings)
--
minor maintenance by mechanical maintenance personnel (mainte-
nance of leaking valves in accordance with CF 498)
--
vibration test on main steam drain pipe by electrical mainte-
nance staff
--
repair work on water screen
--
troubleshooting for the AH-E-12 fans circulating systems (pre-
liminary action to determine if a CM or a work order (WG) is
required)
--
welding activities (DF 2A&B piping in accordance with WO-C-7409)
These activities were selected from day, eyet*ig, and midnight shifts.
The problems associated with the power operated relief valve (PORV)
, surveillance on October 25, 1985, was used as an example of personnel
error.
8.2 Details of the Review
For each of the job functions identified in Section 8.1, the inspec-
tors reviewed the established training program, implementing proce-
dures, qualification, and experience of personnel, qualification and
training of instructors, quality of on-the-job training, and the ef-
fectiveness of training as evidenced in daily activities. The com-
ments and evaluations from the trainees, line supervisors, and
instructors were also reviewed to establish how this feedback was
factored into future training. Lessons learned from the events at
~
the licensee's facilities, as well as other facilities, were also
reviewed to detern.ine how such information was factored into the
training program. The management and QA involvement in the training
area was assessed to determine their effectiveness. No violations
were identified. The observations in each of the review areas are
given below.
-- _--
_ _ _ _ _ _ _ _ _ _ _ _
65
l , 8.2.1 Observation of Work Activities l
The inspectors witnessed the work activities identified in
Section 8.1 above. The activities were conducted in accor-
dance with the procedures. The personnel were knowledgeable
- in the requirements of the procedures and technical specifi-
l cations. Experienced personnel were providing adequate J guidance to inexperienced personnel. All of the activities l witnessed were conducted satisfactorily.
Two situations, the monthly reactor protection system sur-
veillance and the followup of the PORV surveillance error,
are discussed in detail. For the reactor protection system
surveillance, the inspector witnessed the interaction be-
tween experienced and inexperienced personnel. The experi-
enced technician read the procedure requirement and the
inexperienced technician performed the action. Adequate
time was provided to the inexperienced technician to con-
7 duct the action. The experienced technician was patient to
let the inexperienced technician learn from his mistakes.
For example, while performing step 8.5.14 of procedure
1303-4.1 (revision 47), the inexperienced technician con-
nected the digital voltmeter to the test jack of the auxil-
iary power supply instead of the power range detector power
supply. The inexperienced technician sensed a concern when
he had to over adjust to bring .Se output within tolerance.
At this juncture, the experier ;e technician stopped the
procedure and initiated the i v'stigation. The foreman
brought the lead foreman to th' work area to provide fur-
ther assistance. In step 8.8.8 of the same procedure, the
technicians noted that the pressure / temperature bistable
trips were not within tolerances. In this case, the I&C
supervisor came to the work area and assisted the personnel
in identifying and correcting the concern. The problem was
caused by an above-normal zero/ gain setting on the signal
converter. The problem was corrected and the supervisor
instructed the technician to monitor the output from the
signal converter carefully for potential problems in the
future.
The inspectors also reviewed the actions undertaken by the
I&C group to preclude the problems developed during ana
after the PORV surveillance test described in paragraph
8.1 above. The maintenance department issued a verbal or-
der to require that all techr.ical specification related
surveillances are to be conducted by senior personnel. The
inspectors reviewed the log for monthly surveillances of
RPS channels. Since October 25, 1985, these surveillances
were conducted by senior personnel. The inspectors found
the activities to be conducted in a competent manner by
qualified and properly trained oersonnel.
66
8.2.2 Training Policy and Programs
Procedure 1000-POL-2600.01, Revision 0-00, dated April 15,
1985, established the licensee's training policy. This
policy : tresses items such as the basis for training,
commitment to training, on-the-job performance, and effec-
tiveness. It also established the role of the Training
Advisory Council and the commitment to pursue and maintain
INPO accreditation. This policy is implemented through
system-wide procedures. The inspectors reviewed the imple-
menting procedures listed below as a representative sample.
Documents Reviewed
Number Title Revision Date
6211-PGD-2612.01 Auxiliary Operator Training 5-00 8/1/85
Program TMI-1
'
6210-ADM-2621.01 Maintenance Training Program 3-00 7/20/84
TMI-1/TMI-2
1
6210-ADM-2621.02 Maintenance Training Instructor 2-00 10/24/84
Certification / Qualification
Training Program
6210-ADM-2621.03 Maintenance Indoctrination Training 1-00 5/17/85
Program TMI-1 and IMI-2
6210-ADM-2604.03 Control of Examinations Units 1/2 0-00 3/1/85
6210-ADM-2620.01 Revision Review / Required Reading 3-00 11/1/85
Book for Technicians and Support
Training Instruction
6210-ADM-2610.02 Operator Training Instructor 01 8/23/82
Indoctrination / Qualification
Training Program
6200-ADM-2682-09 Standard Evaluation While In Training 0-00 4/15/85
6200-ADM-2682-10 Trainee Evaluation Once Back 1-00 9/13/85
On-The-Job
6200-ADM-2682-11 Program Evaluation 0-00 4/15/85
6200-ADM-2682-12 Course Evaluation 0-00 4/15/85
Maintenance Department Standing Orde s Nos. 17 and 18
l
67
The training for non-licensed personnel consists of both
formal and on-the-job training. The formal training is
provided at the training facility and the laboratory. It
includes discussions of theory, hands-on operation, and
maintenance of generic components such as pumps and valves.
At present, the formal training is conducted in classes
containing all levels of personnel. The instructors pre-
pare the lessons for the persons with average experience.
As a result, the formal training tends to be boring for
t
senior staff and difficult for the entry-level personnel as
determined through inspector interviews. This is identi-
fied as an area of concern by both the instructors and
trainees. The licensee's self evaluation report identified
similar problems.
Training and development educational procedures and job
descriptions for instructors of various grade levels are
developed by the Corporate Human Resources Department. The
inspector reviewed the training and qualification records
of the maintenance and A0 instructors. The instructors'
files were complete and current. The files contained eval-
uations by peers, supervisors, managers, and Corporate spe-
cialists. The inspector reviewed the evaluations by the
students and noted that the comments are reviewed and cor-
rective action is taken. In one case the instructor re-
ceived additional training in a specific area.
The inspector reviewed the A0 Crew Required Reading Book
index. The A0s signed off that they have read items such
as IE Notices, significant events at TMI-I and INPO
reports.
An Operator Training Work Request form is submitted by per-
sonnel requesting additional, modified or new training.
Request form 84-01 regarding additional limitorque valve
training was submitted by an A0; he is now being trained in
response to his submittal.
The training program for non-licensed personnel is expected
to change considerably in the future due to INP0 accredi-
, tation and an automatic mode of progression for technicians
and maintenance personnel. The automatic mode of progres-
sion is in effect for I&C personnel. However, the electri-
cal and mechanical maintenance personnel have not ratified
this option in the union contract. This option will be
considered for union ratification in the Spring of 1986.
,
'
The licensee has completed the I&C job and task analysis
required for the INPO accreditation. This analysis was
done by an outside contractor. An on-the-job training pro-
gram was being drafted at the time cf this inspection. The
licensee has completed a self evaluation in the I&C area
68
and the report was being drafted for submittal to INPO.
The inspector reviewed the licensee's activities and deter-
mined that the licensee is progressing towards INP0 accred-
itation. The licensee intends to resolve all concerns
identified in the self evaluation report by March 1, 1986.
8.2.3 Training Audits and Evaluations
In addition to direct observation of the effectiveness of
training through the conduct of work, the inspectors inter-
viewed at least one experienced and one inexperienced per-
son from each work group for each shift. The interviewees
were complementary to the training program and instructors.
The inspectors discussed student and industry experience
feedback with instructors. The instructors went over the
lesson plans with the inspectors and demonstrated the depth
of preparation. In new areas, the instructors spent up to
eighteen hours of preparation for one contact hour of
classroom instruction. The instructors also demonstrated
that the comments from trainees and line supervisors were
incorporated in the lesson plans in a timely manner. The
instructors provided specific lesson plans to show how they
incorporated industry experience gathered from NRC bulletins
and INPO documents in such lesson plans.
The quality assurance staff conducts routine audits in
training and review of the effectiveness in audits for
maintenance. The inspector reviewed QA audit reports
S-TMI-85-04 (Training) and S-TMI-84-01 (Maintenance) and
noted that these audits were indepth and technically ori-
ented. QA personnel demonstrated good knowledge in the
requirements of non-licensed training and industry practic-
es for maintenance and calibration. In addition to the
audits, OA monitors the training and the activities con-
ducted by trained personnel. The findings from the audits
and monitorings are forwarded to various levels of manage-
ment for actions and feedback. The findings are also
trended regularly to identify and correct potential
problems.
The line supervisors actively participated in training ef-
forts. Annual training requests were provided to the
training personnel in a timely manner. Managers and super-
visors monitored the quality of instruction through occa-
sional class attendance. Some line supervisors expressed
reservations when instructors signed-off practical factors
for technicians and mechanics.
The inspectors noted that several practical factors for
maintenance personnel were signed off by the on-the-job
training (0JT) coordinator on the same day. In discussions
. - _ - _ _ - _ - _ _ . .. . _ _ _ __ __ __ -
-
69
with the trainees and DJT coordinators, the inspector
learned that the date used does not necessarily imply that
the specific practical factor was conduct 9d on that day.
Certain practical factors go on for weeks under the DJT
coordinator's cognizance. The date signed merely indicates
that the DJT coordirator reviewed the effort on that day in
accordance with Maintenance Department Order No. 17. This
order allows for the coordinator's evaluation of those in-
dividuals who attained the required skill level through
previous experience. The coordinators and the trainees
indicated that the existing practical factors records do
not always adequately reflect the extent of training re-
ceived. The inspector reviewed several oractical factors.
In the mechanical area, it is possible to complete several
practical factors on the same day. In the electrical area,
when the practical factors were signed off on the basis of
. previous work experience, it was not always possible for
the trainee or the coordinator to recall the job ticket
number or the component designators used for such sign-
offs. This was discussed in detail with licensee represen-
tatives and they acknowledged the inspector's observation.
The adequacy of the records for on-the-job training will be
reviewed in a future NRC inspection (289/85-30-05).
8.3 Conclusions
Within the scope of this inspection, the licensee's training program
was effective and is geared to improve on-the-job performance with
management support and commitment. QA is actively involved in train-
ing. The personnel were knowledgeable in the work and procedural
requirements and conducted the activities with care. When faced with
problems, the personnel demonstrated the ability to take conservative
measures and seek help. In brief, with the exception of certain
records-related problems and licensee identified areas of concern,
overall, the non-licensed training program was implemented and main-
tained in accordance with license requirements and applicable indus-
try standards.
9. Startup to Power Operation Review
During the inspection period, the staff reviewed selected portions of the
licensee's maintenance and surveillance and startup testing programs to
assess the readiness of the plant for startup and release from the final
power escalation test hold point. The selected areas inspected included
outstanding licensee identified items in the surveillance, maintenance and
startup test areas. The objective was to identify equipment operability
problems that could adversely affect safe operations of the facility. The
startup testing area was reviewed as documented in paragraph 4.
;
- - - - - - , - - _ - - . . . - - - - - . - - . . - - - -
-
,. - - - - - . - , . -- ,
70
The inspector reviewed selected portions of the licensee's applicable cor-
rective action tracking systems to determine if any adverse condition for
safety-related equipment operability existed. The inspector's review in-
cluded tracking systems for open maintenance job tickets. The inspector
reviewed the open job tickets and discussed all work that had been classi-
fied as priority one, two, and three. Of the priority one, two, and three
- work, the inspector determined that the work required by these jobs would
not have an adverse affect on the plant safety if not performed prior to
returning the plant to operations.
Also, the licensee maintains a system for E&Ds associated with surveil-
lance test procedures in order to identify and track procedure and hard-
ware problems identified during performance of the tests. The
surveillance test coordinator (STC) has placed a summary of unresolved
E&Ds on a computer system and periodically reviews a computer printout to
ensure that E&Ds are ultimately resolved.
Prior to the restart of the reactor subsequent to the turbine trip test,
the inspector requested a computer printout of open surveillance test
E&Ds. The inspector along with the STC and an operations engineer re-
viewed each item listed on the printout. No open test exceptions or defi-
ciencies were identified on the computer printout which would have
precluded restart of the plant.
The inspector asked wnether outstanding E&Ds were routinely reviewed prior
to any reactor startup. The operations engineer stated that enclosure 1,
" Plant Critical Check List (Shutdown greater than 24 hours) " and enclosurc
2, " Plant Critical Check List (Shutdown less than 24 hours)," to operating
procedure 1102-2, " Plant Startup," require the verification of the adequate
performance of numerous surveillance tests. As 'ch test is verified,
each exception or deficiency which is unresolved must be reviewed to
assure that surveillance acceptability is not affected. The inspector
reviewed enclosures 1 and 2 to procedure 1102-2 and observed the verifi-
cation of surveillance tests and had no further questions concerning this
area. There were no E&Ds open that would preclude startup after the
turbine trip test.
10. Steam Generator Mechanical Cleaning
Between January 7-9, 1996, TMI-1 Restart Staff inspectors observed the
,
water slap process demonstration and test program at Lynchburg, Virginia,
with the objective of understanding the process and familiarizing them-
selves with the test program. The water slap process involves accelerating
a column of water toward a tube support plate at a velocity sufficient to
dislodge accumulated scale or corrosion products from the spaces between
tube and support plate surfaces. The test program was to establish that
the steam generator integrity would not be compromised by process damage
to components such as tubes, tie rods, tube support plates, and tube-to-tube
sheet attachments. The testing included multiple applications of the
71 process to a full scale mockup up to the fourth support plate. Process transients such as pressure, stress and tube or tube support acceleration were measured. From these measurements, stress values of components were established for comparison to the ASME Code, Section III, NB, 1983 edition. The testing conducted to date indicates that stress values are signifi- cantly below the ASME code allowable levels. The testing and evalua- tion of data are expected to be finalized by late February 1986 and documented in a safety evaluation. The NRC inspectors observed testing in progress, examined equipment in use and reviewed portions of previous safety evaluations that were applicable to use of a similar process. The test program was per- formed with calibrated sensors and equipment and was noted to be carefully planned and performed. Discussions with utility personnel indicated that the process was successful when previously used.
11. Licensee Action on Previous Inspection Findings
Tne inspector reviewed licensee action on previous inspection findings to
ensure that the licensee took appropriate action in response to the find-
ings or by self-initiative, and that the licensee's action was timely.
11.1 {0p5 n) 'nspector Follos Item (289/83-26-03): Temperature effects on
RM-L10. Since a question still remains relative to temperature ef-
r
ect on RM-L10 the licensee has taken or will take the following
actions:
--
The RM-L10 set point has been reduced by 10% to compensate for
the expected decreased performance of the detector with increas-
ing sump temperatures.
--
A work order has been initiated to modify RM-L10 with a thermom-
eter device that will cause increased voltage as a function of
temperature to automatically compensate for decreased efficien-
cy; such modification will be made pending the success of test-
ing by the vendor.
--
Procedural modifications are being considered relative to taking
RM-L10 out of service in favor of composite sampling when the
reactor is shutdown and the auxiliary boiler is expected to dis-
charge to the turbine building sump.
Operatior - 1 data so far indicates that turbine building sump tempera-
tures may not exceed the design specification for the detector. The
licensee will continue further monitoring of the system.
._ - _ _ _ . - _ - _ . ._ .
l 72
11.2 (Closed) Inspector Follow Item (289/84-18-01): Licensee to review
the applicability of using pressurizer uncompensated water level in
the reactor coolant system leakage calculation. The licensee com-
pleted that review and decided not to use the uncompensated pressur-
izer water level signal in the computer calculation for reactor
coolant system leakage. The licensee determined that the gain in
,
'
accuracy was only 0.05 gpm and, accordingly, this gain in accuracy
did not warrant major procedural and software changes.
Since the use of compensate water level introduces only a + 0.1 gpm
error in the calculation, the inspector considered the licensee's
leak rate calculation is acceptable.
'
11.3 (0 pen) Inspector Follow Item (289/84-18-02): NRC staff to review
applicability of the evaporator loss term in the reactor crolant
system leak rate calculation. By internal memorandum, the Office of
Nuclear Reactor Regulation advised Region I that the evaporative loss
term should either be recalculated to a new baseline number or delet-
! ed since the current value introduced a negative unidentified
leak rate which was unrealistic. Concurrent with that determination,
.
'
the licensee decided that a readjustment of the evaporative loss term
was appropriate. A licensee procedure change was reviewed and ap-
1 proved to incorporate the licensee's decision in this matter but the
computer program revision still need work. The licensee committed to
, revise the evaporative loss term to zero and to implement necessary
'
procedural and software changes by January 31, 1986.
I
11.4 (0 pen) Unresolved Item (289/85-19-02): Licensee to review steam
generator primary to secondary leak rate actions to factor in experi-
ente gained as a result of the_ power escalation program. Based on
TMI-1 staff review, the leakrate Calculation was relatively stable
despite transients on the plant. It appears that the limit for
shutdown 6 gph (gallons per hour) above baseline (0.5 gph) is
realistic. However, current procedural actions call for extended
evaluation of the leakrate for 6-12 gph above baseline. This extended
evaluation period appears to be no longer necessary. The licensee
committed to review this matter and issue appropriate procedure
revisions by April 1, 1986.
11.5 (0 pen) Unresolved Item (289/85-22-02): Licensee to resolve problem
with steam relief valve actuation upon initiation of the turbine
driven emergency feedwater_pu_mp. Details of licensee actions are
,
discussed in paragraph 4.3.2.2. The licensee is committed to provide
!
a fully automatic emergency feedwater initiation system by the
startup af ter the next refueling outage (approximately rpril 1987).
11.6 (0 pen) Unresolved Item (289/85-25-05): Licensee to resolve prob.l_em
'
with continuous actuation of steam _ generator safety valve on high
power reactor trios. Details of licensee actions are discussed in
paragraph 4.3.1.2.2. The licensee committed to having a plan for cor-
rective action and implementation schedule by the end of the eddy
current outage (approximately April 1986).
,
. , - _ , , . _ . _ . _ , . - ,__.._-.-. - __ _ . _ _ _
,
1
73
11.7 (0 pen) Unresolved Item (289/85-25-06): Licensee to review and
provide a response letter to NRC Inspection Report 50-289/85-25
with respect to the discarding of a test exception and deficiency
,
record. The licensee committed to provide the response letter by
'
Janu_ry 31, 1986.
! 11.8 (Closed) Unresolved Item (289/85-26-04): NRC staff to review
licensee assessment relative to dose effects of leakage from makeup
'
pump (MU-P) 1A. The licensee has completed investigation of anoma-
lous releases due to problems encountered in the maintenance and op-
eration of MU-PIA. The following occurrences were reviewed and found
to be acceptably evaluated.
,
Activity Fraction of
.
Date Event Released Quality Limit
10/28/85 Maintenance 0.7 Ci 0.0086%
i problem causing
150 gal. spill
12/30/85 Operation 46.3 Ci 0.07%
problem causing
300 gal. spill
Additional details in this report relative to recurrence of noble gas
contamination are in paragraph 6.
11.9 (Closed) Unresolved Item (289/85-27-02): Licensee to evaluate high
steam generator water level limit with respect to decreased heat
transfer efficiency. The TMI-1 Restart Staff review of licensee ac-
tion is documented in NRC Inspection Report 50-289/85-28 but this
item number was inadvertently missed from being closed. It is ad-
dressed in this report for administrative purposes although the high
limit is no longer currently needed (see paragraph 4.2.2.1).
11.10 (0 pen) Inspector Follow Item (289/84-16-01): Licensee to balance
auxiliary / fuel handling building ventilation. Based on a review of
radiological incidents (paragraph 4) it appears that the leaving of
the ventilation system in an unbalanced condition has contributed to
the spread of noble gas out of potentially contaminated or contami-
nated areas resulting in short-lived daughter product contamination
of personnel. Additional details are in paragraph 6.4.
12. Exit Interview
i The inspectors discussed the inspection scope and findings with licensee
management at a final exit interview conducted on January 10, 1986. In
addition to the testing program status meeting of December 17, 1985, in-
terim exit interviews occurred on: December 20 1985 (Non-licensed train-
ing); January 8, 1986 (Startup Test Program); and January 9, 1986
,-- - - . -, - .- .- . ._ . . _ _ _ - . _ _ . -
74 (Radiation Protection). The following licensee personnel attended the final exit meeting:
J. Colitz, Plant Engineering Director, TMI-1
T. Hawkins, Manager, Startup and Test, Technical Functions H. Hukill, Director and Vice President, TMI-1 C. Incorvati, TMI-I Audit Supervisor, Nuclear Assurance B. Leonard, Operations Training Manager R. Masoero, Engineer, Plant Analysis R. Neidig, TMI-1 Communications
M. Nelson, TMI-1 Review Program Supervisor S. Otto, TMI-1 Licensing Engineer
L. Ritter, Administrator II, Plant Operations, P11-1
M. Ross, Manager, Plant Operations, TMI-I D. Shovlin, Manager, Plant Maintenance, TMI-1 P. Sinegar, Administrator II, Manager, Maintenance M. Snyder, Preventive Maintenance Manager C. Smyth, TMI-1 Licensing Manager R. Toole, Operations and Maintenance Director, TMI-1 The exit meeting was also attended by S. Maingi, a nuclear engineer repre- senting the Commonwealth of Pennsylvania. The inspection results, as dis- cussed at the meeting, are summarized in the cover page of the inspection report. Licensee representatives indicated that none of the subjects dis- cussed contained proprietary information. Unresolved Items are matters about which information is required in order to ascertain whether they are acceptable items, violations, or deviations. Unresolved item (s), discussed during the exit meeting, are documented in paragraphs 2.2.4, 4.3.1.2.2, 4.3.2.2, 6.4 and 11. Inspector Follow Items are matters which were established to administra- tively follow open issues based on inspector judgement or on licensee / staff commitments prior to the TMI-l restart. Inspector follow item (s), discussed during the exit meeting, are documented in paragraphs 6.1, 6.3, 8.2.3 and 11.
-- . - _
_
4
ATTACHMENT 1
Startup and Testing Meeting
December 17, 1985 - 2:00 p.m.
Licensee:
W. Bloomfield, Nuclear Safety Engineer (IOSRG)
J. Carroll, Jr., TMI-1 Startup and Test Director, Technical Functions
J. Colitz, Plant Engineering Director, TMI-I
E. Hammond, Nuclear Safety Compliance Committee
R. Harding, TMI-1 Licensing
T. Hawkins, TMI-1 Startup and Test Manager, Technical Functions
H. Hukill, Director & Vice President, TMI-1
M. Knight, TMI-1 Licensing Engineer
R. McGoey, TMI-1 Licensing
R. Meyer, Nuclear Safety Compliance Committee
R. Neidig, TMI-1 Communications
J. Pearce, Senior Engineer I, TMI-1
M. Sanford, Mechanical Systems, Technical Functions
C. Smyth, TMI-1 Licensing Manager
R. Toole, Operation and Maintenance Director, TMI-1
NRC:
L. Bettenhausen, Chief, Operations Branch
A. Blough, Chief, Reactor Projects Section lA
N. Blumberg, Lead Reactor Engineer
R. Conte, TMI-1 Restart Manager
W. Kane, Director, TMI-I Restart Staff
J. Thoma, Project Manager, PWR Project Directorate #6, NRR
P. Wen, Reactor Engineer
State Representative:
W. Dornsife, Department of Environmental Resources, Commonwealth of Pa.
S. Maingi, Nuclear Engineer, Commonwealth of Pa. ,
_..
-. . _ _ _ _ _ , __ . _ _ __ _ _ _ . _
-. . -. . _ . - . . . _ --
-- - - - - . - -
"
,
ATTACHMENT 1
. $ d
NRC MEETING ON TMI-1
'
STARTUP AND TEST PROGRAM
i AGENDA - DECEMBER 17, 1985 AT TMI-l
1. Introduction R. J. McGoey
.
2. Overview and 0TSG Crud Phenomenon J. J. Colitz
l.
,
3. EFW Turbine Driven Pump Safeties M. Sanford
Discussion
.
r
4. Main Steam Safety Discussion M. Sanford
I
5. Conclusion R. J. McGoey
.
I
i
a
2153f
:
1
. _ . . . _ _ - - - . _ _ _ _ _ - _ _ , , . _ _ , . . . _ _ _ . - , _ _ _ _ . , _ . _ _ , . . , _ . . , _ . . . _ _ _ _ _ _ . _ _ _ , . . _ . _ _ . _ _ . . . . _ - - . . . -
.
POWER ESCALATION TEST SEQUENCE
Perf. TP 800/1 Perf. TP 800/1 Perf. TP 800/1 Perf. TP 800/1 Perf. TP 800/2 Perf. TP 800/9 Prerequisites
Reactor Trip EFW Pump Auto for
CPFPET Increase-> CPFPET Increase 4 CPFPET Increase
Power to 80% Power to 90% Power to 95%
>' Increase
Power to 100%
^? " $$ ot
_3 Start Test
4 Flag 4
Tu Tdo
l l l l
Perf. TP 849/1 Perf. TP 800/3 Perf. IP 800/5 Perf. TP 885/1 RP 1550-08 and Perf. TP 835/2
ICS Tuning at Thermal Exp. Unit Load Steady -
TG Ops. -
SP 1301-9 8 Con -> Turbine Bypass
Power Checks for Hgr. State Test Test Power Dist. Valve Test
I I I I
P:rf. TP 836/1 Perf. TP 849/1 Perf. TP 846/1 Loose Parts
FW Sys. Ops. Main Steam ICS Tunin9 ]Perf.TP800/3
Thermal Expan. -
Incore Thermo- 9 Vibration Moni-
rnd Tuning Safety Valves at Power lChk. for Hanner couple Ops Test toring Data
Perf SP 1302-1.1 Perf. TP 800/1
-
Power Range -
Plant 3
Calib. Performance
Perf. TP 800/5 LM-29 High
-
Unit Load Steady -
Rad Noise @
State Test Measurement
Perf OP 1103/16 Perf TP 200/1.3
-
Heat Balance -
8 Sta Feed Wtr 9
Calculations Sheli Side Test
Perf. TP 836/1 JI #C-7563
- FW Sys. Ops. - Acception to 3
and Tuning SP 1302-5.4
Perf. TP 849/1 RP 1550-05
-
ICS Tuning -
React. Coeff. 3
100% at Power
SP 1301-9.5 RP 1550-09 : RESTART LICENSING
- Reactivity -
Unit Acceo-
ance Test
9 CONDITION
Anomaly
SP 1303-1.2 Perform
-
Rx. Coolant -
Neutron
Surveys
)
Flow Surv.
R. C. Flux Loose Parts 23
- -
Vibration Mont- 3 *
Flo" toring Data $
C
i
$
-
a
_ _ _ _ _ _ . _ _ _ _ _ _ _ _ _
'
ATTACitMENT 1 ,
.
.
I Longitudinal Section of the OT5C
i l
!
l_
*
Primary inlet
V
Tubesheet
*
. Primary Head
,
-
. , . -
\ f D4 ,
s s
N ,
I
'
\
-
4 N
s
l j I- s
..
Tube support 7 secondary shell
Plate s %
, g g i ,
l I o s .
'
( "
4
M
"
'
4s,.a.,,
'
-
*
.
.
,
$ttam *#N 8
s e.&
I
f:d .
>
,
-
a
l
2-
}_: r. .a.. . ,
r..a.ai.,7
" "
T .
- -
<
[
(s '
-
;
-
l 4
he .
_:4 _ . ,
g ,,s
\/ \ /Uk
,, , :
,
.
., s
, {% sopport swiri
Primary Outlet -
_' f ,
. l
*
6
.
- . -..........q
ATTACHMEflT 1
... .
nA frn rm ( A n '
\ \ 1 1
O
rn n
avg
'
iT; _ p>< % h
>
,
...... l
- ,
$ l
5 ,
"
h
-g .
'
'
.
f% J
'
k-
-
C, u..
d k-.O
e
^C rC
r '
, /,' . <
/ / /
'
, j/ v
'
, /
'
/, ,
<
f e /
' f, /4 ,
(_ a da
4 $& &-.
4 WJ
SECT A-A
ATTACHMENT 1
i
9 '
-- E
%n_ ..
, -
B
-
,
! ! -
i .
,
8
. .
..._....?.'......... . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . .
.g
: Y
.
,
.. -
.:....... . ...._.. 13
i 4
. . w................. ..... s ... .. . . .
D
^
. .
.
.
: ' 3c. v.
g ,
,. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.g
< .
-
%o) k
. ... .... . . . . . . . . .
,
. . . . . . . . . <,.
er
. . .g
.
i
v !
$.
.... ....... . . . . . .
.g m
e
P II
; ,
d
. ..........._..._. . . . . . . . . . . .. . .g
-
. .
.....'......._.]....._........_..........._....... .a
,
..
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ._ . .... .g
; '
. .
.
',
;....
,
. . . . . . . . . . . . . . . . - . . . . . . . . . . . . . . - . . . . . . . . . . . _ i -e
.
. .
,
,
, . .
o .
. . . . . .
y l I I I I
I I F 1 'I - l l I
3888RR888898 RRgaeea o
x - un ZMW CNuMMO l
ATTACHMENT 1
AGENDA
THI-1
EfWTURBINEORIVENPUMPMAINSTEAMSAFETYVALVES
I. CONCERN
MSV22BLIFTSUPONINITIATIONOfEF-P-1
CONCERN:P0TENTIALSTARVINGOfEF-P-1 TURBINE.
II. EVALUATION
1. MSV228 LIfI ONLY OCCURS WITH SIMULTANE0US MS-Vl3AlB
OPENING.
2. PUMP WILL OPERATE WITH SAFETY VALVE OPEN.
3. PC-SlMSV-6 RESPONSE T00 SLOW TO PREVENT MSV228 LIFT.
-
ATTACHMENT 1
!
III.INTERIMACTIONTAKEN
GUIDANCEPROVIDEDTOOPERATORIfSAfiTYVALVELIFISDURING
EF-P-1 OPERATION:
o OPERATOR INSTRUCTED ON HOW TO RESEAT THE SAFETY VALVE
o OPERATOR INSTRUCTED ON WOW TO TAKE EF-P-1 0UT Of
SERVICE (IfSAFETYCAN'TBERESEATEDANDADEQUATE
MOTOR DRIVEN FLOW IS AVAILABLD
- o OPERATOR INSTRUCTED ON WOW TO OPERATE EF-P-1 WITH
l SAFETYVALVEOPENORGAGGEDCLOSED
l l
I
,
ATTACHMENT 1
I
IV. CURRENT ACTION PLAN
1. REVISE INTERIM (CORRECTIVE) ACTION:
l USESINGLEVALVEOPENLOGICF0RSHORTTERM(CYCLE 5) j
0 JUMPER ACROSS MS-V13B AUT0 START LOGIC T0 KEEP
HS-V138 CLOSED
o MANUALLY OPEN MS-V13B FROM CONTROL ROOM If MS-V13A
- FAILSTOPROVIDESuffICIENTSTEAM
l 2. IMPROVE RESPONSE TIME CHARACTERISTICS Of PRESSURE l
- CONTR0LERPC-5.
o INSTALL NEW PRESSURE CONTROLLER & VOLUME B00 STER
,
o TEST FOR ACCEPTABLE PERFORMANCE WITH SIMULTANE0US
VALVEOPENING.
ATTACHMENT 1
,
SUMMARY
1. MSV22B LIfI OCCURS ONLY WITH SIMULTANE0US MSV13AlB
OPENING.
2. O INTERIMACTIONTOBEREVISEDTOONLYOPENONEMSV13
-
o IMPROVED CONTROLLER TO BE TESTED AS CORRECTIVE
ACTION
3. AB0VE ASSURES EF-P-1 OPERABILITY.*.N0 NUCLEAR SAFETY
CONCERN
ATTACHMENT 1
MS-V-13A MS-V13B
C Q
2" l 2"
a a
[ggM
, 32" k 2 -u,- k ; '12",,[gRQfi
0TSG
J VT F7 3 OTSG
fiS-V9A f1S-V10A ____ . MS-V10B I!S-V9B
' r
PC-5 _
-
7, e,
( [ MS-V6
MS-V22A 8"
f1S-V22B
-C>
_
f1
Ir l 4
L J -
F10M
A!JX BOILER
-
4,, 73 .i
AS-V4
-
EF-U1 EF-P1
N
Tf;I-1
STEAM SUPPLY TO
EFW PUMP TUPSIf1E
._
ATTACHMENT 1
MAIN STEAM SAFETY VALVES (MSSV'S)
EVENT
DATE: 10/21/85
0 401POWERTRIP
o ALL MS$V'S LIFT
o NUMBER Of MS$V'S RELIfI
o PEAK RECORDED PRESSURE 1050 PSIG
o RELIfTS AT APPR0X. 1010 PSIG
-
.
ATTACHMENT 1
-
, I
.
EVALUATION
l 1. VALVE SETP0INTS WERE CONflRNED CORRECT.
2. VALVE PERFORMANCE
l
a) ALL VALVES INITIALLY LIFT FRON PRESSURE WAVE AS
EXPECTED.
b) BLOWOOWN IS WITHIN ACCEPTABLE B00N05 OIS$V'S 010 NO
EXCEED 101 BLOW 00WN)
i c) PROLONGE0 INITIAL LIfI COULD RESULT IN NSSV SEIP0lNT
ORIfI(ASMUCHAS31)
d) CONSISTENT WIIW OBSERVATIONS AI OTHER PLANTS.
1 i
ATTACHMENT 1 3. ANALYSIS SHOWS A SYSTEM PRESSURE SPIKE RELATED T0 IURBINE
( STOP VALVE CLOSURE IN EXCESS Of HIGHEST MS$V SET P0 INT. !
4. SAFETYVALVEPERFORMANCEMAINIAINEDPLANTWITHINDESIGN
ENVELOPEDURINGTRIP.
5. INADVERTANT 75% POWER TRIP SHOWE0 REPEATABILITY.
l < l L
-__
(,- .
,
ATTACHMENT 1
,
'
CONCLUSION
l ,
l. THI-l MSSV SEIP0lNIS ARE SET CORRECILY.
l l 2. IMI-l MSSVS RESPONDED WITHIN EXPECTED TOLERANCES DURING 1
THEPLANTTRIP.
l '
3. THI-l CURRENT SEIP0INTS ARE ADEQUATE FCR CONTINUED S
PLANTOPERATIONATALLPOWERLEVELS.
l
4. GPUN WILL CONTINUE TO INVESTIGATE MEANS TO IMPROVE TH
l ALREADYACCEPTABLESITUATION. l l l , < }}