ML20211C929
ML20211C929 | |
Person / Time | |
---|---|
Site: | Three Mile Island |
Issue date: | 05/29/1986 |
From: | Blough A, Conte R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
To: | |
Shared Package | |
ML20211C898 | List: |
References | |
50-289-86-05, 50-289-86-5, NUDOCS 8606120433 | |
Download: ML20211C929 (36) | |
See also: IR 05000289/1986005
Text
. ____ _____ _
.
.
U.S. NUCLEAR REGULATORY COMMISSION
REGION I
Report No. 50-289/86-05
Docket No. 50-289
License No. DPR-50 Priority -- Category C
Licensee: GPU Nuclear Corporation
Post Office Box 480
Middletown, Pennsylvania 17057
i Facility: Three Mile Island Nuclear Station, Unit 1
Inspection At: Middletown, Pennsylvania
Inspection Conducted: March 7, 1986 - April 11, 1986
Inspectors: R. Conte, Senior Resident Inspector (TMI-1)
D. Johnson, Resident Inspector (TMI-1)
A. Krasopoulos, Reactor Engineer, Region I
D. LeQuia, Radiation Specialist, Region I
R. McBrearty, Reactor Engineer, Region I
J. Rogers, Resident Inspector (TMI-1)
A. Weadock, Radiation Specialist, Region I
F. Young, Resident Inspector (TMI-1), Region I
Reporting Inspector: 8 T 2P-JC
fr R. Conte,6enior Resident Inspector (TMI-1) Date
Approved By: M M 7-8(,
A. Blouffi, Chief Date
Reactor Projects Section No. lA
Division of Reactor Projects
Inspection Summary:
Resident and region-based NRC staff conducted routine safety inspections (440 1
hours) of power operations, focusing on plant and personnel performance. Spe-
cifically, items reviewed in detail in the operation and maintenance area were:
outage preparation and shutdown /cooldown; significant performance appraisal
team findings; and local leakrate testing. Special focus occurred on: the ;
eddy current testing of steam generator tubes; feedwater nozzle weld cracking;
and significant events of March 15, 22, and 24-25, 1986. Other review areas
included: fire protection program -- fire brigade training; outage radiation
protection; and licensee action on an emergency diesel generator 10 CFR 21 ;
report.
l
l
l
8606120433 860606~ !
PDR ADOCK 05000299 )
O PDR- i
l
. . . _
_ _ _ _ _ _ - - - - - --_
.
. +
2
Inspection Results:
Overall, licensee management took many initiatives to prepare for and control
outage work impact on control room personnel. Licensed. operators continued to
exhibit overall professionalism and demonstrated detailed knowledge of plant
design and outage work status. However, during the transition from operations
to outage conditions, the degasification event pointed out the need for enhanced
attentiveness by technicians and operators to special evolutions. Also, the
licensee apparently failed to properly survey or evaluate conditions in the
reactor building for the buildup of radioactive iodine; this failure was con-
trary to regulatory requirements (paragraph 3.4.5.2). Contributing to that
event were communication factors during the high pace of activities on March
24-25, 1986. Once the iodine problem was recognized, licensee corrective
action and dose evaluation methodologies were quite good.
! .There was also a failure to properly post a radiation area (paragraph 6.2.2).
l The implementation of the licensee's radiation protection program was generally
- proper, except as noted above.
1
Plant and personnel performance for the reactor trip event of March 15, 1986,
was essentially as expected. The licensee properly performed post-trip
evaluations and identified followup actions.
^
Steam generator tube eddy current testing was properly conducted and controlled
with respect to oversight of vendor activities. Proper corrective actions were
planned and implemented for the emergency feedwater nozzle weld crack problem.
,
Licensee planned followup actions on the emergency diesel generator 10 CFR 21
report are appropriate to the circumstances.
I
a
!
4
!
I
_
, , _ . . _ , . _ . . .. , , _ . - . ~ . - , , - - - ..-<-,e- - - - - - ,--w - ,
l
.
.
DETAILS
1. Introduction and Overview
1.1 NRC Staff Activities
The overall purpose of this inspection was to assess licensee
activities for the transition from power operations to cold shut-
down outage activities as they related to reactor safety, worker
radiation protection, and fire protection. Within each area, the
inspectors documented the specific purpose of the area under review,
scope of inspection activities and findings, along with appropriate
conclusions. The inspector made his assessments by reviewing infor-
mation on a sampling basis through actual observation of licensee
activities, interviews with licensee personnel, measurement of
radiation levels, independent calculation, and selective review
of applicable documents.
1.2 Licensee Activities
From the beginning of the inspection period to March 21, 1986, the
licensee operated TMI-1 at full power, except for the reactor trip of
March 15, 1986 (sae paragraph 3.2). The primary licensee focus during
that period was preparations for the eddy current outage testing which
started over the weekend of March 22-23, 1986. The reactor trip was
due to a secondary plant trip caused by a combination of procedural
weaknesses and operator actions during the tran:fer of inservice
coolers for the main lube oil system.
The remainder of the period involved cold shutdown and reactor vessel
drained down conditions to support eddy current testing (ECT) of a
sampling of steam generator tubes. Initial bubble and drip test of
all tubes showed very good results with no indication of leaking
tubes. At the end of the inspection period, ECT data collection was
essentially complete and licensee analysis was proceeding. )
l
Two events occurred during the initial stage of the outage. There
was a noble gas release during RCS degasification on March 22, 1986, l
which resulted in the declaration of an unusual event (see paragraph j
3.3). Also, there was a buildup of radioactive iodine and lesser '
significant contaminatica incidents on or about March 24, 1986, with
the RCS vented to the RB etmosphere (see paragraph 3.4). The iodine
buildup was due, in part, to inadequate planning. l
l
i
_- _ _ - _ _ - _ _ - _
?
Q
O
4
2. Plant Operations
2.1 Scope of Review
The NRC resident inspectors periodically inspected the facility
to determine the licensee's compliance with the general operating
requirements 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,
including plant housekeeping and fire protection measures;
--
control of ongoing and special evolutions, including control
room personnel awareness of these evolutions;
--
control of documents, including logkeeping practices;
--
implementation of radiological controls; and,
--
implementation of the security plan, including access control,
boundary integrity, and badging practices.
Enhanced resident office coverage at this facility enabled the
inspectors to more fully assess the adequacy and effectiveness of
performance of operating 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 followed as required by plant policy;
--
equipment status changes are appropriately documented and
communicated to appropriate shift personnel;
--
the operating conditions uf plant equipment are effectively
monitored and appropriate corrective action is initiated when !
required;
)
--
backup instrumentation, measurement, and readings are used as
appropriate when normal instrumentation is found to be defective
or out of tolerance;
--
logkeeping is timely, accurate, and adequately rsflects plant
activities and status;
l
l
_
4
..
.-
.
.
5
--
operators follow-good operating practices in conducting plant :
operations; and, '
--
operator actions are consistent with performance-oriented
training.
The inspectors focused their attention on the areas listed below.
General / Operations
--
Control room operations during regular and backshift hours,
including frequent observation of activities in progress, 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 on activities that could affect plant safety or
impact on plant operations
--
Areas outside the control room
--
Selected licensee planning meetings.
--
Plant shutdown and cooldown t
--
RCS degasification
--
Instrument air walkdown on Emergency Feedwater (EF) flow
control valves
--
EF flow indication alignment (indicator to transmitter
alignment)
--
EF standby status
I
Maintenance I
!
--
Testing of motor-operated valves
1
--
"A" battery bank cell replacement ;
Surveillance i
1
--
Local leakrate testing of the Reactor Building penetrations
--
EF surveillance testing
!
!
!
l
rm ,cl.
~
l
.. ..- s- ..
. , m ,
-
.
_
. <- ; -
.
6 -
-
Radiological Controls-
--
Locked high' radiation doors
--
Radiation Work Permitsposting
i
--
Weekly survey maps
2.2 Findings
2.2.1 General -
T
J Licensee manag'ement continued their presence and involve-
ment.,in da,ily activities.
t Good preparation and preplanning
was poted for operatiqnal aspects of the scheduled eddy
current outage. The work list and scope of work to be
pepformedwasaggressivebutachievablewithinthetime
limit established by the licensee. During the weekend
i
prioito :the start of thg outage, work appeared orcanized
[ Equipment was staged and made ready
'and"well contro1]ed.
fo'r entry ,into radiologically controlled areas. A separate
control point wi.th two nealth physics computer terminals
for reactor building (RB) entries was established in
readiness for.the substantial Monday morning work force
a rri val .
s s
y
Licensea operators remained professional in performing
-
.,their duties, especially in control room activities. The
control room operators were knowledgeable of plant status
>
from a work activities viewpoint in addition to being
knowledgeable of detailed reactor plant conditions. Shift
foremen were additionally stressed by interfaces with the
work force for equipment control measures, but they remained
in overall control and those pressures were somewhat alle-
viated by management: 'For instance, additional "off-shift"
- -
, shift supervisors _were. assigned to be coordinators of
activities.in various locations; such as, the reactor
building. Also, a standard set of cold shutdown " tag outs"
were made ready prior to the start of the outage to avoid
-
numerous case-by-case equipment control measures and hand-
- ling by the on-shift shift foreman.
There were two events that occurred at the start of the
j' 'utage. Details are addressed in' paragraphs 3.3 and 3.4.
o
,
It appears that the licensee could have better anticipated
(by calculation and/or survey) the quantities of iodine
"
released from the reactor coolant system during their
preparatory efforts. Further, personnel could have paid
>
closer attention to detail during the degassing operation
, at the chemical sampling room to avoid the actuation of
the relief valve and subsequent release of noble gas.
?- .
,
4
.
.
7
2.2.2 Performance Appraisal Team Significant Findings
During this inspection period, the resident inspectors were
also involved in and supported the Performance Appraisal
Team (PAT) Inspection (No. 50-289/86-03). The primary
functional areas followed by the resident inspectors were
plant operations and the design change, engineering and
modification areas. The below-listed PAT findings were
considered significant in that more information would be
needed in order to determine impact relative to safety
related equipment operability for the upcoming startup.
--
There was a lack of documentation on the environmental
equipment qualification for certain cables used in the
"
--
10 CFR 50.59 safety evaluations were not performed for
lead shielding on certain piping, and technical evalu-
ations did not consider the additional stress on piping
during dynamic conditions in distinction to static
conditions.
--
The remote shutdown panel (RSDP) was apparently not
electrically isolated from control room (CR) panels.
Accordingly, a fire in the CR could adversely affect
the RSDP.
--
With the battery room temperature kept below minimum
design temperature during certain time periods, the
effect on battery capacity was not fully evaluated by
i
the licensee.
For these items, the licensee will be conducting an
internal evaluation to determine the affect of these find-
ings on safety-relatad equipment operability. Licensee
action will be verified during the next NRC inspection (No.
50-289/86-06).
Prior to the outage startup, the NRC resident office staff
will follow up on the below-listed additional PAT findings !
for which the licensee was in the process of taking correc-
tive action during outage work:
--
proper adjustment of limit switch and torque switch
settings for those valves in which motor operated
,
!
valve testing will be complete;
--
quality of the compressed air purchased for emergency l
feedwater two-hour backup air system in meeting design
specification; and,
I
.- -
_
__.
._.__
.
.
8
--
completion of licensee established preventive mainte-
nance for MU-V16C, High Pressure Injection Isolation
Valve (verified complete by the inspector during this
period).
These and all other PAT findings will be documented in
detail in NRC Inspection Report No. 50-289/86-03.
2.2.3 Reactor Building Local Leak Rate Testing
The inspector conducted a review of the local leak rate
testing that was being accomplished as a result of a
scheduler exemption that was granted to the licensee in a
NRC letter, dated February 20, 1986. This letter allowed
the licensee to postpone 10 CFR Part 50, Appendix J local
leak rate testing that was to be accomplished by February
23, 1986, until the present o'utage. The leak rate testing
was to be accomplished on seventy-one (71) valves which
were required to be individually leak tested per the
requirements of Appendix J,Section III.D.3.
At the end of the inspection period, the licensee was in
the process of completing the remaining leak tests. The
licensee had chosen to test all ninety-four (94) valves and
penetrations required by their procedure. This testing
encompassed the seventy-one (71) valves that were required
to be tested. At the end of the inspection period, seven-
ty-nine (79) of the ninety-four (94) valves were complete.
The inspector discussed with licensee personnel involved
in conducting the testing and reviewing the data, the
requirements for this testing and general overall test
conduct. Licensee personnel were aware of the require-
ments, and personnel conducting the tests were knowledge-
i
2
able of test methodology. The inspector witnessed selected ,
test evolutions and observed that equipment was calibrated
i
and testing was accomplished according to procedures.
Corrective action for discrepancies found during testing
was' acceptable. The test procedure SP 1303-11.18 was re-
viewed and found to be adequate. Review of final data and
total leak rate will be reviewed in the next resident
office inspection (No. 50-289/ 86-06). The inspector had
no concerns on the licensee's action to date.
2.3 Conclusion
Operators continued to conduct themselves in a professional manner.
In general operators were responsive to daily plant problems as they
arose. Overall, procedures were properly implemented.
,
wu-- ,, c w , , - - --
+n-r
-.
.
O
9
i Management took initiative to control outage work and alleviate
undue pressures on control room personnel. Appropriate licensee
action is being planned to address the significant PAT findings noted
above.
3. Event Review
,
3.1 Introduction and General Scope of NRC Staff Review
1
During this inspection period, there were three events that the NRC
staff reviewed in detail. They were: the reactor trip of March 15,
1986; the unusual event of March 22, 1986; and, the airborne radio-
activity / contamination events of March 24-25, 1986. In general, the
i following aspects were considered for each of these events:
--
details regarding the cause of the event and event chronology;
--
functioning of safety systems as required by plant conditions;
--
consistency of licensee actions with license requirements,
approved procedures, and the nature of the event;
--
radiological consequences (on-site or off-site) and personnel
exposure, if any;
--
proposed licensee actions to correct the causes of the event;
--
verification that plant and system performance are within the
limits of analyses described in the Final Safety Analysis Report
(FSAR); and,
--
proper notification of the NRC was made in accordance with
For each of these events the inspector provided a chronological /
factual summary; specific scope of NRC staff review; licensee find-
ings, both operational and radiological; and, NRC staff findings,-
both operational and radiological. An overall conclusion on licensee
performance is provided in paragraph 3.5.
3.2 Reactor Trip
3.2.1 Event Chronology and Background Information !
,
Shortly before 10:00 a.m. on March 15, 1986, operators
started to shift main lube oil coolers. A low pressure
transient developed in the main lube oil system and pres-
sure sensing instrumentation tripped the main turbine
generator. A restart modification functioned properly
(turbine-to-reactor trip) and it tr.ipped the reactor at
10:02 a.m. This resulted in the actuation of the steam
,
, ,. . , , - - , _ _ -,. r. , , - , , . , ,-
. - - - - - _ _ _ _ _ _ _ _ _
. o
.
10
generator safety valves causing both visible steam and loud
noises for several minutes. This was noticeable from off
site locations.
The main lobe oil systems provides relatively cool lubri-
'
.
cating oil for the main turbine generator journal bearings
and thrust bearing. If this system were inoperable as
sensed by pressure switches, significant damage could
result to the main turbine generator shaft and bearing
assembly. Accordingly, a secondary plant trip (turbine
trip) it designed to occur when low pressure is sensed,
which is what occurred on March 15, 1986. The turbine-
to-reactor trip is an anticipatory function to minimize the
pressure transient on the reactor coolant system upon a
secondary plant load rejection (which occurs on a turbine
trip). Shifting lube oil coolers is a weekly operation to
rotate operating equipment.
Being on site at the time, the senior resident inspector
heard the steam generator safety valves actuate and he
immediately responded to follow licensee post-trip activi-
ties, including the post-trip review. Plant response was
as expected and operators established stable hot shutdown
conditions. The licensee post-trip review identified
proximate sequential causes of the trip back to the main
lube oil pressure transient; however, a root cause of the
lube oil transient could not be established. It was sus-
pected that either a valve malfunctioned or the transfer
1 valve was not properly operated. The licensee planned to
continue to review this area during the eddy current
outage.
The licensee restarted the plant Saturday night, March 15,
1986, and achieved full power by early Monday, March 17,
1986.
3.2.2 Specific Scope of NRC Staff Review for the Reactor
Trip
Specific to the reactor trip event noted above, the i
inspector verified the below listed items:
--
initial proper response of the plant to the post-trip
window on the pressure-temperature (P-T) plot;
--
personnel properly implemented ATOG procedures and
prudently acted on unusual conditions;
! --
identification of the^ sequential proximate causes for
the trip along with a reasonable determination of the
i root cause;
1
k
.
l
a
11
--
post-trip review was conducted in accordance with
Administrative Procedure AP 1063, Revision 6, " Reactor
Trip Review Process;" and,
--
No ur veriewed safety issues identified in post-trip
review data.
In addition to discussions with cognizant licensee person-
nel, the inspector:
--
made an independent assessment of post-trip. parameter
response based on visible strip chart and indicators
in the control room shortly after the event;
--
attended the licensee's post-trip review;
--
attended the independent safety review of abnormal
t
response of the feedwater system;
.
'
--
reviewed the' complete post-trip review package (No.
86-03); and,
--
reviewed Administrative Procedure (AP) 1063, Revision
7, dated March 24, 1986, " Reactor Trip Review Process"
for adequacy.
3.2.3 Licensee Findings 1
No unreviewed safety questions were identified. Prior to
the startup, the licensee imposed a requirement to further
investigate duplicating those conditions that caused the
low pressure situation in the main lube oil system. ,
Licensee personnel were able to duplicate a low pressure
transient in the lube oil system during a transfer to
standby cooler operation. However, the results of that
review were inconclusive in identifying a specific root
cause. The licensee became more convinced that either
personnel error or a procedure weakness contributed to
the pressure transient in the lube oil system.
Subsequent to that startup.and during shutdown for the eddy
current outage on March 21, 1986, the licensee conducted a
cooler transfer with the main turbine at full speed but off
the regional electrical grid. Licensee personnel learned
that the degree of opening the cooler fill valve affects
'
how long it takes the standby cooler to fill, vent, and be
pressurized. After five minutes, when personnel were
satisfied that the standby cooler was filled, vented, and
pressurized, personnel shifted the transfer valve. No
turbine trip occurred and the licensee continued the plant
shutdown.
.
, , --
e-- r-, -
.. .
'
l
12
The licensee's internal trip review also identified that
personnel physically forced the transfer valve during
repositioning. Applicable procedures were not clear in
-
assuming the standby cooler was filled, vented, and
pressurized. Licensee operations management concluded that'
.the major (root) cause of the event was weak procedures and
, a minor (root) cause was personnel error. Procedure changes
are planned.which are to include keeping the fill valve
, . opened. Operations engineers will continue to monitor lube
oil cooler transfers and to identify potential valve
,
malfunctioning.
- Two other post-startup items were identified for review by
the licensee's Technical Functions Division.
--
At the time of the trip, the saturation margin monitor
(SMM) instruments responded by indicating 25-30 F sub-
cooling margin. One computer channel alarmed (indica-
ting'less than 25 F). Redundant subcooling instrumen-
tation and the post-trip P-T plot indicated that
saturation margin was above 50 F. Licensee personnel
believe that the response of the SMMs was due to
resistance temperature detector (RTD) response.
--
The plant did not get to the post-trip P-T window
until about twelve minutes; whereas, the procedural
guideline is 10 minutes. The post-trip review deter-
mined that the cause of the problem was feedwater flow
going to zero or the regulating valves going shut.
This was unanticipated for proper Integrated Control
System (ICS) OTSG level control. I&C personnel were
to work on the applicable controllers during the eddy
current outage.
Total radioactivity (mostly Xe-133) released as a result of
the steam generator safety valve actuation was 37 micro-
curies which was well below technical specification limits.
3.2.4 NRC Findings
Operator response to the event was consistent with emer-
gency procedures. Operators were conscious of, and oriented
their action toward, decay heat removal. Plant response was
essentially as expected, except as discussed below.
The inspector independently confirmed the licensee's
findings as noted above. The sequential proximate causes
' of the trip were reasonably-determined prior to startup and
the licensee provided an extensive and thorough review to
identify a root cause of the event. Although a specific
root cause could not be identified, the trip could have
i
,. , , - . -m1
,
.
.
13
been avoided had personnel used better judgement by stand-
ing back and further reviewing the situation, instead of
forcing the transfer valve to reposition.
The post-trip review package was not distributed until
shortly before the post-review meeting. Consequentially,
it appeared that certain plant engineering personnel were
i reviewing and analyzing data during the conduct of the
meeting. However, a majority of the meeting members were
4
knowledgeable of the data; and it was evident that they had
( analyzed and made conclusions prior to the meeting. The
late distribution of the data package did not adversely
affect the adequacy of the post-trip review. However, the
inspector discussed with licensee management potential
problems in certifying the AP 1063 requirement (paragraph
3.1)..." group... convene to review and concur with the
. conclusion of the post-trip review" with the untimely
3
distribution of a post-trip review package.
Licensee management acknowledged the inspector's comments
and reiterated the engineering personnel were adequately
briefed and prepared for the meeting. The inspector had no
i further comments on this matter.
i At the conclusion of the inspection, Technical Functions
Division (TF) was still reviewing the post- startup items
(SMM response, and feedwater response).and the licensee
event report to the NRC was not received. Accordingly,
,
this area is unresolved pending licensee submittal and
Region I review of the applicable LER. Further, Region I
will review TF disposition of the post-startup items noted __"
above (289/86-LO-06).
3.3 Inadvertent Release During Degassing Operations
3.3.1 Event Chronology
At 1:50 a.m. on March 22, 1986, during degassing opera-
tions, a first alarm (alert) was received on RM-A-6,
Auxiliary Building gaseous effluent radiation monitor.
This was followed at 3:01 a.m. by a similar alarm on RM-A-8, Auxiliary Building ventilation stack vent radiation
monitor. The plant was in a cooldown mode at approximately
260 F in preparation for a five-week outage and the primary
system was being degassed via the pressurizer steam space
-
sample line to the make-up tank. 1
The licensee secured the degas flow path and at 3:05 a.m.
the radiation monitor alarms had cleared. The release path
was determined to be the inadvertent lifting of CA-RV-5, a
relief valve in the degas flow path.
>
. _ . - _ - . _ _. , ,. . , - , . _ . _ .
-
. . . , . .
___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
.
l
l
.
14
3.3.2. Specific Scope of NRC Staff
'
The inspector conducted discussions with cognizant licensee
personnel and reviewed the following documents:
--
system flow diagrams; ,
--
Operating Procedure (0P) 1102-12, " Hydrogen Addition
and Degassification;" and,
--
OP 1104-43, " Nuclear Plant Sampling."
3.3.3 Licensee Findings 1
The licensee determined that the lifting of CA-RV-5 was
,
due to improper throttling of CA-V-10 as the cause of the
release. When CA-RV-5 lifted, a flow path was set up to
the Auxiliary Building sump which vented to the environment
via the Auxiliary Building ventilation. system. Initial
calculations showed that approximately 30 curies of noble
gases, primarily Xenon-133, were released. This calcula-
tion was subsequently changed to approximately 15 curies
upon systematic evaluation of the strip chart recorders
from RM-A-6 and RM-A-8. Calculation of dose at site
boundary revealed a cumulative dose of 4.3 x 10 -4 mrad.
.
'
All releases and site boundary doses were far below the *
limits of 10 CFR Part 20 and the technical specifications.
Although applicable procedures contained precautions about
the potential lifting of CA-RV-5, the operating procedures
, are being revised to more clearly specify pressure limits
to be used when degassing to preclude lifting of the relief
4
valve.
The licensee is also reviewing the entire degassing opera-
tion; flow paths and methodology to determine if operating
methods can be changed to preclude future degassing
problec3. These recommendations will be reviewed by the
Plant Review Group (PRG) in future meetings.
3.3.4 NRC Staff Findings
The licensee's initial actions in response to this event
were timely and adequate to control the release. The
inspector agreed with the licensee that the cause was
improper throttling of CA-V-10, which caused CA-RV-5 to
lift and discharge to the Auxiliary Building. 1
i
1
l
.
.
15
The licensee acknowledged that, due to the process flow
path liquid / gas / steam mixture coupled with the difficulty
!
of pressure control, the potential exists to lift CA-RV-5
and release radioactivity to the Auxiliary Building spaces
instead of to designed processing systems. The licensee
has agreed to review this evolution oy referral to the
Plant Review Group for possible long-term corrective
actions. Present corrective actions to prevent recurrence
consist of operating procedure changes to ensure more
adequate control of the degassing evolution as presently
conducted.
This area will remain will remain unresolved until Region I
reviews the completion of licensee action as stated above
(289/86-05-01).
3.4 Airborne Radioactivity and Persc....el Contamination Event of
March 24-25, 1986
3.4.1 Sequence of Events
On March 24, 1986, starting at approximately 9:00 a.m., the
licensee removed the "A" and "B" once-through steam genera-
tor (OTSG) handhole covers and commenced ventilation of the
"A" and "B" OTSGs to the Reactor Building (RB). Ventilation
was effected by connecting the suction from a 500 cfm port-
able HEPA unit directly to the handholes after cover
removal. These filter units provided particulate cleanup
only and exhausted directly to the RB. By 9:49 a.m., the
licensee had collected air samples from the exhaust of the
HEPA units as required by their Radiation Work Permit (RWP)
, and ALARA review for the job, but failed to analyze the
! iodine samples until 8:47 p.m. that evening.
In addition, the "B" OTSG sample result, which showed
'
,
3.8 MPC iodine, was not evaluated by the counting room
technician or reviewed by the Group Radiological Control
1
Supervisor (GRCS) until the following day (March 25,1986).
- This delay in sample processing held true for the majority
- of samples collected on March 24, 1986, with the exception
'
i
of the worker breathing zone samples discussed below. Delay
was due to a large sample backup in the counting room; the
licensee was concurrently concerned with noble gas levels
t
inside the RB, and noble gas air samples were receiving
l counting priority.
i
The licensee continued to work on both OTSGs throughout the
day (removing manways, etc.). The licensee indicated that
although they discussed the potential for airborne radir-
iodine in the RB, they dismissed it partly due to the
4
o
.
16
[The inspector reviewed the applicable RWP and determined
these samples could not be considered representative of
general RB airborne iodine conditions since: they sampled
the worker's breathing zone air on OTSG platform; and, a
portable ventilation spot intake was positioned immediately
adjacent to the handhole during removal and would, there-
fore, be pulling any airborne iodine away from the worker.]
By 6:46 p.m. on the evening of March 24, 1986, a RB general
area air sample fr)m the 346-foot elevation indicated 0.4
MPC iodine. Additionally, a worker who had been contamin-
ated during the day received a whole body count (WBC) at
5:50 p.m., which indicated I-131. The licensee indicated
this data was recognized as indicators of an iodine problem
by a member of the Radiological Controls Field Operations
dayshift staff; however, this information was not effec-
tively communicated to the relieving shift and no actions
were taken. During this time the licensee also failed to
notice a 6000 cpm increase on RMA-2, the RB airborne monitor
iodine channel, which increased from 10,000 cpm at 3:00
p.m. to 16,000 cpm by 8:00 p.m.
At approximately 9:00 p.m., two OTSG jumpers exiting the RB
were found to be contaminated. They were showered and sent
for a whole body count (WBC). WBC results at approximately
11:00 p.m. indicated that 'aese individuals had internal /
external contamination levels of Iodine-131 up to 340
nanocuries (nc).
The WBC technician notified the GRCS of the iodine results.
About this same time (11:00 p.m.), another set of OTSG HEPA
exhaust samples were counted, showing 3.6 MPC and 1.5 MPC
from the "A" and "B" 0TSG, respectively. The licensee also
collected a " grab" sample from RMA-2 at 11:10 p.m., which
subsequently showed 0.48 MPC in the RB recirculating air.
At that point, approximately fourteen (14) hours after the
start of the iodine event, the licensee first became aware
of radioactive iodine concentrations in the RB. Shortly
after this identification was made (approximately 1:00 a.m.
on March 25,1986), the licensee restricted access to the
RB.
As a result of this failure to assess airborne radioactivity
concentrations in a timely manner, one hundred and twenty-
four (124) personnel were sent for whole body counting with
seventy-seven(77) showing positive indication of unplanned
iodine intakes up to 120 nCi.
.
_ . _ _ _ _ - _ _ - _ _
.
J
.
17
3.4.2 Scope of NRC Review
The inspectors assessed the licensee's identification and
followup actions to the airborne radiciodine event on March
24, 1986, by the following methods:
--
review of licensee air sample log sheets for March
24-25, 1986;
--
review of the HP log book for March 24-25, 1986;
--
review of RWP 31641, " Remove' Insulation and Manways
,
and Handholes from "A" and "B" OTSG"
--
review of associated ALARA Review No. 86-03-06;
--
discussion with cognizant licensee personnel.
--
review of OP 1102-12, " Plant Shutdown;" and,
- --
review of applicable maintenance and surveillance
procedures,
i
- 3.4.3 Licensee Findings - Radiological
Licensee identification of airborne radiciodine in the RB
i occurred at approximately 11:00 p.m. on March 24, 1986, and
stemmed from the following information input:
--
WBC results from two individuals contaminated while
working in the OTSGs (see paragraph 6.0 of this
report);
--
air samples taken at the exhaust of HEPA units venti-
lating OTSGs, showing 3.6 and 1.5 MPC, respectively;
and,
--
grap air sample result showing 0.48 MPC in RB general
- recirculating air.
Licensee response after this identification of airborne
radioiodine included the following:
'
--
collecting and evaluating additional samples;
j. --
scheduling individuals present in RB on March 24,
j
1986,(124) for whole body countings for dose
- assessment;
- --
limiting access to the RB on March 25, 1986, until
j airborne levels were reduced;
--__________._____.m__
- .
.
.
18
,
--
shortening iodine charcoal filter count times from
1000 to 500 seconds to expedite sample processing;
and,
<
--
purchasing charcoal filter units to attach directly
in line with the 500 cfm HEPA units to provide iodine
cleanup capability. These units were purchased and on
site by March 27, 1986.
The licensee had not performed a formal investigation or
critique into the causes of the airborne event as of April
-
2, 1986; the last day of NRC radiation specialist review of
this event. However, informal discussion between the
inspectors and cognizant radiological control personnel
indicated the following factors contributed to a failure to
identify airborne radioiodine levels in a timely manner:
--
reliance on two unrepresentative breathing zone air
samples as indicators that no iodine problem existed;
--
long delay in sample processing, compounded by a
prioritizing of noble gas samples over iodine samples;
--
failure to note the gradual increase in RMA-2 monitor
readings; and,
,
--
lack of communication among shift personnel and
licensee divisions.
Licensee followup included whole body counting of 124
personnel present in the RB Nrch 24, 1986. Seventy-
seven people were found *o external / internal contam-
ination, predominantly Iod :-131. The highest WBC
indicated 340 nCi of I-131; however, this individual had
been externally contaminated while working in the OTSG.
Selective shielding of body areas and recounting of this
individual indicated the majority of this activity was
t
located on the hands, approximately 80 nCi of I-131
! activity is being assigned as an intake. The highest
intake value, discounting skin contamination contributions, ;
is 120 nCi of I-131. With respect to radiological releases.
for the period March 25 to April 7, 1986, the licensee
reported the release of 194 Ci of noble gases from reactor
building purging. Also released were approximately 3.156 ;
E-5 Ci Particulate, Tritium, and Iodine, predominantly
'
Iodine 131. The value is based on ventilation filter effi-
ciency calculations, since release rates were below detect-
' able limits of effluent iodine monitors. The potential
release converts to an integrated dose at the site boundary i
i
of 4.94 E-3 mrad, gamma, and 2.16 E-2 mrad, beta, which is l
well within technical specification limits.
i
l
-
.
.
19
3.4.4 Licensee Findings - Operational
A review of this event from an operational perspective was
performed by the licensee to see if the methodology used in
the operation of the plant contributed significantly to the
event. A review of plant procedures and the order that
plant evolutions were performed was conducted. The review
did not identify any design problems that would directly
contribute to the root cause of the event.
3.4.5 NRC Findings - Radiological
3.4.5.1 Once the iodine buildup in the RB was identified, licensee
response actions were appropriate and oriented toward worker
radiation protection. Intake dose calculations were con-
servative and confirmed by state-of-the-art equipment.
Reasonable efforts were made to distinguish skin contamin-
ation versus actual radioactive material intake based on
whole body counting results. No overexposure were
identified. However, the unplanned exposures could have
been prevented as noted below.
The inspector verified the licensee's. list of individuals
identified as having intakes based on the preliminary WBC
and that these res/ ts are being followed and recounted by
1
the licensee. The licensee appears to have adequate tech-
nical expertise in this area.
3.4.5.2 Requirements of 10 CFR 20.201 are that each licensee make
such surveys as may be necessary to comply with all sections
of Part 20. Requireraents of 10 CFR 20.103(a)(3), in part,
are that "... the licensee shall use suitable measurements
of concentrations of radioactive materials in air for
detecting and evaluating airborne radioactivity ... as may
be necessary for timely detection and assessment of indi-
vidual intakes of radioactivity by exposed individuals ..."
The inspector considered that the licensee's failure to
provide a timely detection and assessment of airborne
radiciodine conditions in the RB on March 24, 1986, was i
an apparent violation of 10 CFR 20.201. As a result of I
an approximate 14-hour delay prior to detection of the l
airborne radiofodine, 124 people present in the RB were 1
exposed to then unknown concentrations of airborne
radiotodine. Seventy- seven individuals were later found
to have unplanned uptakes of radiciodine.
Licensee radiological control personnel indicated that
iodine had been surveyed for, in that samples of iodine had
been taken starting at 9:49 a.m. and had continued through- !
out the day. The inspector stated that these samples had
o .
.
20
been collected but not counted or analyzed and, therefore,
they could not be considered "an evaluation of the radiation
hazards" as defined in 10 CFR 20.201(a). The licensee also
indicated that they had anticipated an iodine problem from
the OTSGs,-as evidenced by their ALARA review requirement
to sample at the HEPA unit exhaust. However, actual iodine
levels turned out to be greater than those anticipated.
,
As indicated above, the apparent failure to survey for
airborne radioactivity is contrary to 10 CFR 20.201 and
20.203(a)(3) (289/86-05-02).
3.4.6 NRC Findings - Operational
The resident inspector reviewed the event to verify that
the licensee was performing the plant operations associated
with this event per applicable station procedures. A review
of the shutdown and degassing event was performed. This
review determined that the evolutions were performed in
accordance with applicable procedure. The plant had been
degassed down to below 5 cc/kg of hydrogen in the reactor
coolant system. However, the pace of operational activities
on March 24, 1986, appears to have been a factor in the
event. Licensee orientation that day was to establish
access to both steam generators so that contractors could
start the eddy current test process. It appears that pace
was too fast in contrast to activities over the previous
two days for support personnel; namely, Radiological Con-
trols Division, to keep up with activities in radiological
areas. On the other hand, if the Radiological Controls
Division was severely taxed, it could have stopped work for
support activities to catch up.
The inspector concluded that both the operation and
radiological controls personnel failed to effectively
.
'
communicate, which was contributory to the unplanned
exposures.
1
3.5 Conclusion j
For the reactor trip event, operator response was in accordance with
applicable emergency procedures and consistent with their performance-
oriented training. The licensee's post-trip review was thorough and
a substantial effort was made to identify the root cause.
For the degasification event, operators could have been more attentive
to prevent the relief valve lift. Further review is warranted on the
viability of the licensee's degasification methodology. I
.
'
21
For the RB iodine event, certain NRC requirements on properly evalua- r
ting' radiological conditions.were apparently violated. Contributing
to the root cause was poor licensee personnel communication with re-
spect to the pace of activities surrounding the event. Post-event
review and dose evaluation methodologies were adequate and
conservative.
4.1 Periodic Inspections and Quality Assurance Audits
4.1.1 Annual Audits
The inspector reviewed the licensee's 1985 annual audit
report, Audit No. 0-TMI-85-03, Fire Protection. The scope
of review was to ascertain that the audit was conducted in
accordance with the Technical Specification 6.5.3.2.a and
audit findings were being resolved in a timely and satis-
factory manner. No unacceptable conditions were identified.
4.1.2 Biennial Audit
The inspector reviewed the licensee's last biennial audit,
Audit Report No. S-TMI-84-03, TMI 1 and 2 Fire Protection
Program. The scope of review was to ascertain that the
audit was conducted in accordance with TS 6.5.3.1.g and
audit findings were being resolved in a timely and satis-
factory manner. The inspector also reviewed the licensee's
effort to complete Audit No. S-TMI-86-03 being performed to
satisfy the above mentioned TS requirement for the current
period. The auditers performing this audit have placed an
increased emphasis on fire brigade training since it was
previously identified by NRC as an area of potential
weakness. No unacceptable conditions were identified.
4.1.3 Triennial Audit
The inspector reviewed the licensee's triennial audit,
Audit Report No. 0.TMI-84-05, Fire Protection. The scope
of review was to ascertain that the audit was conducted in i'
accordance with TS 6.5.3.2.a and audit findings were being
resolved in a timely and satisfactory manner. No unaccept-
able conditions were identified.
4.2 Facility Tour
The inspector examined fire protection water systems, including fire
pumps, fire water piping and distribution systems, post- indicator
valves, hydrants, and contents of hose houses. The inspector toured
accessible vital and non-vital plant areas and examined fire detection
and alarm systems, automatic and manual fixed suppression systems,
interior hose stations, fire barrier penetration seals, and fire
l
i
!
< l
I
. _ _ __ _ _ _-_______
.
.
22
doors. The inspector observed general plant housekeeping conditions
and randomly checked tags of portable extinguishers for evidence of
periodic inspections.
No deterioration of equipment was noted. The inspection tags attached
to extinguishers indicated that monthly inspections were performed.
During a tour of the control room, the inspector observed that about
six 5 gallon plastic jugs full of water were on the floor by the water
cooler. The inspector pointed out to the licensee that a water spill
in the area of electrical panels is undesirable. The licensee agreed
and the excess water bottles were removed.
4.3 Fire Brigade Training
The inspector reviewed the training records of the fire brigade
members to ascertain that they had attended the required quarterly
training, participated in a quarterly drill, and received the annual
hands-on fire extinguishment practice. No unacceptable conditions were
identified, except as noted below.
4.4 Fire Drills
The inspector reviewed fire drill records and noted that, in some
instances, 12 to 20 fire fighters responded to a drill. The inspec-
tor indicated to the licensee that a drill cannot be realistic and
provide meaningful training for 20 individuals. Further, the licensee
does not have enough fire fighting equipment to supply this number of
personnel. Plant management initially agreed with the inspector that
drills under these conditions could be ineffective and committed to
review the method by which drills were conducted. Depending on the
results of that review, the licensee also agreed to submit a plan for
improvement. Subsequently, the licensee informed the inspector that
the method by which drills were conducted was adequate and in accor-
dance with the accepted program.
In light of the licensee's position on this matter, NRC Region I will
refer the adequacy of training relatively large numbers of personnel
on a per drill basis to the Office of Nuclear Reactor Regulation
(NRR). This is an unresolved item pending further NRC review of the
licensee's training methodology (289/86-05-03).
4.5 Clarification Requirement for Unannounced and Backshift Drills
The inspector reauested to review fire drill records to determir com-
pliance with the requirement to annually perform unannounced ant'
backshift fire drills for each shift fire brigade.
The licensee's records indicate that drills for each shift fire
brigade were performed. Because of the licensee's fire brigade
membership size (approximately 130-150 fire fighters), the inspector
wanted to know how many fire fighters had participated in unannounced
.
.-
23
and backshift drills. The licensee stated that they are not required
to maintain this type of records. Given the size of the shift fire
brigade of 5 fire fighters, it is possible that only 10 to 15 fire
fighters participate in these type of drills. The licensee agreed
that this is possible, but it is not the case.
!
NRC Region I will also refer this matter to NRR for review on the
acceptability of not having records to distinguish unannounced and
backshift fire drill participation and the issue will be included in
the unresolved item noted above (289/86-05-03).
4.6 Fire Drills in Adverse Weather
The National Fire Protection Association (NFPA) Standard No. 17
requires that fire drills be held occasionally under adverse weather
conditions. The inspector reviewed reports of this type of drill and
!
the licensee produced the records of 3 drills that were held under
j adverse weather conditions. The drills were held on February 14,
'
1985, August 8,1985, and September 4,1985. A check with the '
- National Weather Service identified'that for the dates indicated
the temperatures were 30 F, 87 F, and 91 F with clear skies.
The inspector asked NFPA representatives for clarification as to
i
what constitutes adverse weather. NFPA response was that heavy rain,
, snowfall, freezing conditions, and extreme heat would qualify. The
conditions under which the above drills were held do not meet the
requirements. Licensee mana'gement acknowledged the conflict and
i
agreed to review the matter. This is unresolved pending further
review of the licensee and NRC Region I (289/86-05-04).
5. Steam Generator Tube Eddy Current Testing and Emergency Feedwater
Nozzle Cracking
~
j
,
5.1 Allegation on Vendor Use of Controlled Substance
5.1.1 Background
,
On or about March 11, 1986, licensee representatives
i reported to the NRC's TMI-1 Resident Office that an indi-
vidual employed at another nuclear facility made allegations
to the cognizant NRC regional office and/or cognizant Atomic
Safety and Licensing Board on certain vendor representatives
who were to perform eddy current testing on TMI-1 steam
generator tubes. The allegation dea'.t with the misuse of
, controlled substances (drugs) at that other facility. As
yet the allegations were not substantiated, but cognizant
NRC staff are conducting a followup review. i
i
! !
l
t
l
,
i
e
.
24
For TMI-1,' licensee management became concerned and insti-
tuted action to enhance random inspections and vendor
personnel screening during the eddy current outage. During
the period from March 18 to 24, 1986, the licensee reported
i that they had the contract with the vendor revised to speci-
fically exclude the use of controlled substances by vendor
representatives and to require that vendor equipment,
facilities, and persons be subject-to random inspection
and/or testing, as applicable.
When the vendor arrived on site, the licensee reported that
they conducted a thorough search of the vendor's facilities
and equipment taken into the protected area in accordance
with the licensee's security plan. No controlled substances
were found. Also 18 vendor personnel were tested with one
having positive results, possibly indicating the recent use
of a controlled substance. The vendor management acknowl-
edged this finding and asked the licensee to terminate the
individual's access to TMI-1. The individual was not
allowed to participate in eddy current testing.
5.1.2 Scope of NRC Staff Review
In addition to receiving periodic status reports from the
licensee on this problem, the inspector verified the follow-
ing items.
--
The licensee contract with the eddy current test
vendor of equipment / facilities reflected the agreement
for licensee inspection and test of personnel for con-
trolled substances.
--
Vendor personnel were not in possession of or using
controlled substances.
--
The licensee's contract with the vendor reflected ,
appropriate regulations; such as on the reporting of '
deficiencies (10 CFR 21), on employee discrimination '
for expressing safety concerns (10 CFR 50.7) and
- having a quality assurance program (10 CFR 50,
.
Appendix B).
In addition to discussions with the Director of TMI-1 and
plant security personnel, the inspector conducted random
inspections of licensee / vendor facilities and equipment.
Selected sections of the following documents were reviewed.
--
Vendor letter to GPUN, dated April 3, 1986, on indi-
4
vidual termination
--
Contract File No. 017519, original, dated October 16,
1984, through Change No. 7, d ted March 11, 1986
,
)
.y- .-
p-
25
--
Various attachments to above contract changes includ-
ing GPUN specification GED-CS-36, Revision 1, dated
February 29, 1980, and GPUN Contract General Terms and
Conditions, Revision 2, dated January 2,1985
5.1.3 NRC Staff Findings
In general, the inspector verified the accuracy of the
reports made to the NRC's TMI-1 Resident Office. No con-
trolled substances were observed to be in place or in use
at the data collection facility on site and at the analysis
facility in Hershey, Pennsylvania.
The contract reflected provisions for the licensee to con-
duct random inspections and testing of vendor personnel for
controlled substances. The licensee reported that they
implemented that option with negative findings, except as
noted above. The contract also invoked applicable federal
regulations.
As noted below, NRC staff independently observed and con-
firmed selected vendor / licensee eddy current results for
the steam generator tubes. The results of the NRC staff's
technical review of the area are addressed below.
5.2 Eddy Current Examination of Steam Generator Tubes
The inspector reviewed data, observed examinations in progress, and
interviewed eddy current examination personnel to ascertain that the
examinations and the data interpretation were done in accordance with
regulatory requirements and ASME Code requirements. The inspector's
observations included the examination of tube 107, row 41, in the "A"
steam generator. This was a re-examination to assure that complete-
and legible data were obtained for interpretation and evaluation
purposes. Additionally, the inspector observed the interpretation of
data including tube 68, row 30, in the "B" generator, and tube 27,
row 104, and tube 1, row 73, in the "A" generator, and discussed the
interpretation with cognizant vendor Level III personnel.
A 46?; throughwall flaw was identified in "B" generator tube 65, row
30, which was not identified during the 1982 examination of the tube.
A review of the 1982 data, using the equipment available for the 1986
interpretation, indicated that the flaw was present in 1982. The low
voltage associated with the indication, the size of the indication
displayed on the screen, and the presence of non-relevant indications
in the display resulted in misinterpretation of the indication in
1982. Currently available improved equipment, including computer
enhancement capability, allowed the data evaluator to recognize the
indication as a 46?4 throughwall flaw. Additional 1982 data were
o
-
~
26
reviewed by the licensee to assure that additional flaws were not
missed. The inspector reviewed the 1982 and the 1986 data and dis-
>
'
cussed the two data sets with the Level III individual who was
responsible for the 1986 data interpretation. The inspector con-
.
'
cluded that no change in signal characteristics sere discernible in
the two data sets and that no change in throughwall dimension had
occurred since the 1982 examination. This conclusion was stated to
NRR during a telephone conversation from the site on April 11, 1986.
At the exit meeting, the licensee stated that an anticipated revised
technical specification will change the tube plugging limit from 40%
to 50% throughwall; and, based on this, the tubes which contain flaws
less than 50% throughwall would not be plugged at this time.
The inspector found that the data collection and interpretation was
done by qualified personnel and that code and regulatory requirements
were met.
!
l
No violations were identified. Licensee repair and test activities
, will be reviewed in the next inspection period (NRC Inspection Report
- No. 50-289/86-06).
5.3 Emergency Feedsater Nozzles
Each of the six active nozzles on each steam generator contains a
<
thermal sleeve held in place by a collar which is fillet welded to
the sleeve. Cracks in the fillet welds have been found at other
plants and were evaluated by B&W. The welds were scheduled by the
licensee for inspection during the December 1986 refueling outage.
Because of maintenance on the emergency feedwater spray ring, one of
the six active nozzles was required to be removed from the "A" steam
generator and the thermal sleeve / collar fillet weld was liquid
i penetrant examined. Cracks were detected which resulted in two
1 additional nozzles being removed from the "A" generator for
i examination. The additional welds displayed cracks; and, at this
point, the licensee decided to remove the remaining nozzles from the
"A" generator and the nozzles from the "B" generator for liquid
penetrant examination. The cracked welds were repaired by welding
- and re-examinatian by the liquid penetrant method was done to ascer-
tain that the repair was properly made.
The inspector observed the examination of repaired welds on nozzles
A-1 and A-2 and the initial examination of welds on nozzles B-1 and
B-3. The examinations were done in accordance with procedure
6110-QAP-7209.02, Revision 0-01, using the visible dye solvent remov-
able technique. Indications which were identified were evaluated and
dispositioned in accordance with acceptance standards as defined by
. the procedure.
.
_ _ _ _ _ _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ . _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ . _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _
-_
o
.
27
,
' The licensee reported that B&W has changed the nozzle thermal sleeve
design to correct the cracking problem and that the new design will
be installed by TMI-1 at some later date. At the time of this
inspection, the licensee had not decided when the sleeves will be
replaced.
The cracking at TMI-1 was attributed by the licensee to fatigue, but
the specific kind of fatigue was not identified. The item is con-
sidered unresolved pending the availability of documentation to
confirm that the failures at TMI-1 resulted from the failure mechanism
which was evaluated by B&W regarding the earlier cracking at other
3
'
facilities (289/86-05-05). No violations were identified. Licensee
repair and testing activities will be reviewed during the next in-
spection period (NRC Inspection Report No. 50-289/86-06)
5.4 Conclusion
With respect to alleged use of controlled substances by vendor per-
sonnel at another facility, the inspector concluded that the licensee
is taking adequate preventive and surveillance measures.
Except as noted in above paragraphs, no unresolved issues exist on
the process and collection of required eddy current testing data.
The inspector concluded that the personnel involved in obtaining and
'
analyzing the data were knowledgeable and were very familiar with the
eddy current indications expected and found at TMI-1.
Review of feedwater nozzles and repairs determined that the repairs
were performed in accordance with applicable codes. Interviews and
, discussions with different licensee line organizations indicated that
'
qualified people were used and positive control and direction were
i
'
given to this project by licensee management. Independent visual
review of the weld repairs by the inspector determined adequate weld
repairs were performed. In general, the inspector concluded the
.
licensee was proceeding in a safe manner and adequate quality assur-
ance was being implemented on the repair process.
6.0 Outage Radiation Protection
6.1 Introduction
The inspector reviewed the licensee's radiological controls imple-
mented in support of the OTSG outage. The inspector's effort was
.
1
focused in two areas:
--
0TSG work activities; and,
--
licensee posting and labeling of controlled areas.
l Within the scope of this review, one violation, a failure to con-
l spicuously post a radiation area, was identified.
!
e
.
28
6.2 Postir.g and Labeling of Controlled Areas
, 6.2.1 The inspector reviewed the licensee's program for the !
'
posting and labeling of radioactive material and radio- '
logical areas against criteria contained in 10 CFR 20.202
and 10 CFR 20.203. This review included:
--
discussion with health physics technician and super-
visory personnel;
--
inspection of the Reactor and Auxiliary Buildings;
and,
-
--
review of procedure 9100-ADM-4110.01, " Establishing
and Posting Areas."
l
6.2.2 In general, applicable NRC regulations were properly
implemented. However, on April 1,1986, at approximately
3:00 p.m., the inspector conducted a tour of the Auxiliary
Building; and, it was noted that the equipment access double
,
doors to the " hot" machine shop on the 305-foot elevation
i had been left fully open to support equipment removal and
i' no radiological posting at the access to the shop was
visible. Posting for this area is placed on the doors to
the shop; with both doors fully opened, this posting was
totally obscured. Two workers, who had been involved in
'
removal of equipment from the machine shop, were standing
in the nearby area; however, these workers gave no warning
or did not otherwise act to prevent inadvertent entry to
the hot machine shop. The inspector remained in the
.i
immediate area an additional ten minutes until the health
l physics technician supporting the equipment removal process
returned. After surveys of the area and on the removed
,
equipment, the access doors were secured.
l The definition of 10 CFR 20.202(b)(2) for " radiation area"
is any area accessible to personnel where a major portion
1 of the body could receive a dose in excess of 5 millirem in
one hour. Requirements for 10 CFR 20.203(6) are that
radiation areas be conspicuously posted with a sign bearing
the words " Caution - Radiation Area."
At the time of this incident, licensee posting on the " hot"
machine shop identified it as a'"High Radiation Area" and a
-
'
" Contaminated Area." The inspector reviewed surveys of the
area and determined that the licensee had over posted the
area; general area dose rates were less than 100 millirem /
.
hour and did not support posting the area as a high radia- l
tion area. The inspector noted, however, that general area l
dose rates exceeded 5 millirem / hour, and ranged to a !
maximum value of 20 millirem / hour. '
[
,
1
__.________.__.____m_ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ - . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ - . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . . . _ - . _ _ _ _ _ _ _ _ . - . _ _ . _ _ _ _ _ . _ _ . _ _ . _ . _ ______._ ____._ ___ m__ ______.__________
a
,
.
29
Consequently, failure to conspicuously post the access to
the " hot" machine shop is in apparent violation.of 10 CFR
20.203(b) (289/86-05-06).
The inspector noted that, although not technically meeting
the criteria for a high radiation area, the hot machine
shop was considered as such by the licensee and was posted
and controlled accordingly. Two recent additienal problems
in the posting and control of high radiation areas are
described in NRC Inspection Report 50-289/85-30. In one of
the instances, posting for a high radiation area was left
on the ground after a worker exited from the area. Sub-
sequently, another individual inadvertently entered the
area. In the second instance, the licensee did not
adequately post a high radiation area due to non-uniform
radiation fields in the area. While the licensee's posting
and boundary were adequately placed while dose rates were
measured at waist level, dose rates greater than 100 mrem /
hour could be measured outside the high radiation boundary
at the head level.
The licensee indicated that the HP technician supporting
equipment removal from the " hot" machina shop on April 1,
1986, was, in effect, acting at posting for the area while
the doors were open and should cave continually romained in
the area until the doors were secured. The licentee went
on to state that the failure to conspicuously post the shop
was therefore a personnel problem since the HP technician
left his post. The inspector agreed that had the HP tech-
nician remained in the area and issued appropriate verbal
warnings this would have constituted appropriate " posting;"
however, since no one was present, the violation remains a
failure to conspicuously post the area. The licensee's
response to this violation should also address the implica-
tions of recent high radiation area posting problems.
6.2.3 In a related matter, licensee's letter (Serial No.85-301),
dated October 14, 1985, indicated that the licensee adopted
a practice of identifying and controlling high radiation
areas based on dose rate measured at one foot from the
radiation source. The licensee's proposed change has been
reviewed by NRC regional staff and was found to be in
accordance with the applicable regulations in 10 CFR 20.
The inspector reviewed the following procedures and veri-
fied appropriate revisions reflecting the change have been
incorporated:
--
9100-ADM-4000.08, " Definitions Used in Radiological
Control Procedures;"
. __ _ .. _ . _ _ _ . _ . _ _ .. . _ . _
r
'
!
! -
.
! 30
l
l --
9100-ADM-4110.01, " Establishing and Posting Areas;"
i
--
9200-ADM-4110-06, " Control of Locked High Radiation 4
Areas;" and,
i' ,
l
--
3000-IMP-4400.01, " Radioactive Material Identification
and Handling."
! No unacceptable conditions were identified. '
! 6.3 Steam Generator Work Activity
l
l 6.3.1 The inspector reviewed licensee performance of several
! work activities in the "A" and "B" once- through steam
! generators (OTSGs), including the performance of the bubble
test and template installation. This review included:
'
--
review of RWPs covering the above activities (Nos, l
31660 and 31652);
'
i
i
--
review of associated ALARA reviews;
!
--
review of OTSG and personnel contamination surveys;
--
review of meeting minutes of critique held after l
] personnel contamination on March 24, 1986;
i
j --
discussion with HP personnel;
- :
j --
attendance at the pre-job briefing prior to template
- installation on March 25, 1986; .tnd,
i
! --
observation of "A" 0TSG template installation activ-
l ities on March 25, 1986. ;
J
) Within the scope of the above review, no violations were
identified.
'
"
6.3.2 The inspector noted that during the licensee OTSG entries
i on March 25, 1986, adequate health physics personnel were
!
i
present at the work site to control the work activity,
OTSG jumpers were equipped with breathing zone air samples
and headset communication equipment to assess and control
_
'
,
l'
personnel exposure. However, the following deficiencies ;
were noted during the "A" upper channel head jump i
- activities.
1
i
i
j --
A OTSG manway diaphragm, removed from the OTSG on l
j March 24, 1986, was noted in the general area of 1
} worker access to the "A" OTSG platform. Surveys )
j indicated 2.5 R/hr. on contact, 800 mR/hr, at 12
i
f
,
!
'
_ _ ._ . _ _ _ _ _ _ _ ._ _ _ , _ _
a
e
31
!
I
inches from the diaphragm. Yet, it did not appear
i that ALARA principles were used to minimize worker
] exposure.
i
l
--
During jump activities on March 25, 1986, face shields
j were required by licensee RWP for general access to
j the Reactor Building (RB) due to high noble gas con-
! centrations. The inspector noted several personnel
i
involved in the template jump preparation in the RB
were not yearing face shields as required.
,
l The above deficiencies were corrected by the licensee after
.
identification by the inspector. Face shields were provided
f
to all workers inside the RB; the manway diaphragm was
j turned over to reduce dose rates emanating from the contam-
.
inated side and later moved to a lower. traffic area. The
i licensae provided the following explanation concerning the
j above items.
,
Individuals not wearing face shields in the RB were
'
--
'
OTSG jumpers, who would be wearing bubble hoods during
i jump activities. It was not anticipated jumpers would
j be waiting one-half to I hour in the RB prior to jump
- activities and would thereby require face shields.
!
l
--
The manway diaphragm had not been transported out of-
due to access restrictions described in paragraph 3.4.1
) of this report.
j The inspector noted that the manway diaphragms were
i
originally removed during the afternoon of March 24, 1986,
and access to the RB was not restricted until midnight on
the same day. Additionally, no efforts such as shielding
j the manways or removing them to a lower. traffic area had
j been taken at the time of removal.
I
- The inspector had.no further comments on this area,
l
6.3.3 On March 24, 1986, at approximately 8:00 p.m., workers
were in the upper channel heads of the "A" and "B" OTSGs to '
perform a bubble test and install "bungee" cord markers for
j later phatography of the tubesheet. During these activ-
- ities, the water level in the OTSGs was 8-12 inches above
l the tubesheet. The RWP and associated ALARA review required
i
fireman's boots, wet suits, and double rubber gloves'for
} these activities. The gloves worn were standarc plant
- issue, approximately nine inches in length. All seams were
! taped, using masking tape, during the cord inst 411ation to
l the tubesheet. Both workers, one in each OTSG, had to reach
! into the standing primary coolant water that was deeper
I
!-
l
.
'
.
'
o ~
, ;
,
-.
O
32
f
than 9 inches in water. The seams allowed water to leak
in; and, consequently, both workers were contaminated on
each forearm and hand. Skin decontamination efforts were
- not completely successful; both workers left site with
detectable levels of contamination remaining on the skin,
but affected areas were bagged to prevent contamination
spread. 7
- Within the scope'of the above review, no violations were
identified. Skin contamination resulting from this event
did not have the potential for exceeding regulatory skin
dose limits, The inspector noted, however, that the appli-
s
care RWP (No. 31652) and associated ALARA review were not
completely adequate in that appropriate protective clothing
'
, was not provided for the arm extremities in anticipation of
the work activity. Leakage hazards inherent to working in
the standing primary coolant levels were obviously recog-
nized and resulted in the prescription of high-top boots.
However, it apparently was not recognized that the work
performed would require immersion of the hands past the
glove seal level.
- Additional concerns involved in this contamination incident
include the following.
'
, ,
--
No direct method of measuring water level above the
tubesheet was used prior to personnel entry;
'f
I
--
The workers in the' steam generators did not make~any
communication during the work activity concerning the
immersion of their hands past the glove seam level.
Communications to the HP personnel providing coverage
concerning the potential for contamination would have
- been appropriate.
1
The licensee held a critique on March 24, 1986, after the
personnel contamination problem was identified. At this
time, it was decided a Radiological Investigation Report
(RIR) would be generated to identify the causes and cor-
rective actions of this incident. This is unresolved
pending NRC Region I review of the scope and findings of
1
1
the licensee's RIR during a subsequent inspection
(289/86-05-07).
6.4 Conclusion
In general, the implementation of the licensee's radiation protection
program was oriented toward worker radiation protection. The licensee
continued to exhibit difficulties in completely adhering to posting
'
regulations.
i
!
.
o
e
e
33
The events of March 24-25, 1986, were caused by some inattention to
basic ALARA principles and in adhering completely to RWP requirements.
The contamination event did not result in a significant exposure,
but communication and RWP adequacy problems were noted. The inspec-
tor did not consider these problem reflective of a programmatic
problem.
7. 10 CFR Part 21 Review: Emergency Diesel Generator Scavenging Air
Blower
7.1 Background
In accordance with 10 CFR 21.21(b), on February 24, 1986, Philadel-
phia Electric Company (PECo) reported to the NRC Region I recent
failure of the scavenging air blower on Colt Industries, Fairbanks
Morse Diesel Generator installed at the Peach Bottom Atomic Power
Station (PBAPS). Colt Industries' letter to the NRC, dated February
25, 1986, identified TMI Unit 1 as having similar emergency diesel
generators (EDGs) with the same scavenging air blower system as at
PBAPS (Fairbanks Morse Diesel Model No. 3800 TD 8-1/8).
The PBAPS and TMI-1 EDGs are equipped with a turbo-blower parallel
scavenging air system. The scavenging air blower supplies atmospheric
air under pressure to the cylinders for starting and low load
conditions. The blower discharges compressed air to the suction side
of the two turbo chargers where the air is further compressed if the
engine load is approximately 50% or greater. The turbo chargers are
driven by engine exhaust. As the engine load increases, the increase
in exhaust gases increases the speed of the turbo chargers creating a
suction at the turbo charger air inlet. The pressure imbalance opens
the turbo charger inlet air check valve allowing suction directly
from the atmosphere. At that point, the scavenging air blower
becomes " unloaded" so that at 100% load the blower is "windmilling."
The scavenging air blower is a positive displacement lobe-type
blower. This blower consists of two three-lobe spiral aluminum
impellers with approximately 30 mil (.0030 inches) clearance between
lobes and the aluminum casing. Both impeller shafts are attached to
a flexible drive gear, which is driven by the EDG upper crank shaft.
At no load conditions, the blower supplies all of the combustion air
to the diesel. During no or low loads, the loss of the blower causes
the diesel to stop due to insufficient air supply.
The PBAPS blower failure was caused by seizure and fragmentation of
the aluminum lobes of the scavenging air blower and the diesel sub- i
, sequently stopped due to lack of air. The cause of the contact
,
'
between moving parts was believed to be thermal deformation of the
air blower internals and stationary parts due to operation at no load
and low load conditions. (Prior to the PBAPS DG failure, the engine
had been running continuously for fif ty one hours at 20% or low load.)
l'
- - _ _ _ _ _ _ _ _ - _ _ _ _ -
- _ _ _ __ - ____
4
0
34
T
' A similar EDG scavenging air blower failure occurred at the Duane
'
Arnold BWR in June 1984. Colt Industries failure analysis indicated
that the cause could either be thermal creep of the aluminum impel-
1ers or elastic thermal expansion of the impellers due to higher than
normal blower operating temperatures. Either cause could lead to
inneller lobe expansion and reduction of clearances.
Fairbanks Morse Engine Division Service Information Letter (SIL),
dated November 15, 1984,'for Diesel Model No. 3800 TD 8-1/8, suggests
that the following blower clearances be checked annually since recent
blower failures were due to contact of blower impeller lobes and the
blower casing: rotor to rqtor; . rotor to housing; rotor to inner end
plate; rotor to outer end plate; and rotor to rotor trailing edge.
1
- < The rotor to rotor, rotor to housing, and rotor to inner and outer
i '
cnd plate tolerances require removal of the air blower suction and
', discharge lines. The rotor to rotor trailing edge clearances can be
taken through inspection plug holes on t,he blower.
The SIL states that the most likely cause of the Duane Arnold blower
failure is the deformation of the aluminum housing due to localized
heating while running at no load condi,tions for extended periods of
time.' The SIL further cautions against running at no load conditions
longer than five minutes since the differential pressure across the
blower ischigher due to the turbo charger inlet impeller restriction
~
which may lead to excess temperature in the blower. In 1985,
Fairbanks Morse increased the clearances between the blower lobes
and casing for new blowers. '
FairbanksMorseVSII,.datedAugust 13,'1985, cautioned against running
the air blower'at no load conditions during engine break-in after
rebuilt or major repairs. The SIL also suggests that the air temper-
ature be monitored at the suction and discharge of tha blower to avoid
exceeding a temperature differential of 100 F to limit thermal expan-
sion of the blower aluminum lobes. (Neither of the two'TMI-1 EDGs
i have blower suction and discharge temperature gauges.)
7.2 Scope of NRC Staff Review
j
'
The inspector reviewed the above noted 10 CFR 21 report and vendor ;
service information letters to ascertain the nature of the problems l
(deficiencies) as related to TMI-1. Subsequently, he reviewed j
licensee corrective actions to ascertain if the licensee received
i
complete and appropriate information from the applicable vendor and
if licensee corrective actions were adequate to resolve the deficiency
consistent with vendor recommendations.
In addition to discussions with cognizant. licensee personnel, selected
sections of the following documents were reviewed:
, 1
I
,-- - - - - , .
,, - -
_ _ _ - _ , . - - . - , . _ . _ , .
. . _ _ _ . _ ..__
O
o
35
--
Colt Industries' Service Information Letters (SILs), dated
November 15, 1984, and August 13, 1985, for Diesel Model No.
3800 TD 8-1/8;
--
Abnormal Transient Procedure (ATP) 1210-6, Revision 5, dated
March 8, 1985, "Small Break LOCA Cooldown;"
--
Surveillance Procedure (SP) 1303-4.16, Revision 35, dated
December 9, 1985, " Emergency Power System;"
--
SP 1301-8.2, Revision 27, dated December 9, 1985, " Diesel
Generator Inspection and Data for 1984 and 198E "
7.3 NRC Staff Findings
i
'
7.3.1 If a loss of coolant accident (LOCA) should occur with
available offsiteLpower, the TMI-1 EDGs would start and
remain unloaded. That ATP 1210-6 states that, if after
twenty minutes the EDGs are not required (offsite power is
available), the EDGs should be secured by bypassing the
emergency start (ES) signal and then placing the DGs in ES
standby. The bypassing of the ES signal and securing the
EDG requires that the control room operator push both the
engine stop and the diesel generator emergency shutdown
pushbuttons on the control room console. (To place the EDG
back into ES standby requires that an auxiliary operator
push the reset button at the LOCA panel and the control
room operator place the diesel start switch in automatic
standby.) The inspector had no further comments in this
area.
- 7.3.2 The licensee received the November 1984 SIL after the 1984
Annual Emergency Diesel Generator Inspection. After review-
ing the SIL, the licensee determined that the intent of the
e
SIL concerning blower clearance checks had been met since a
j vendor representative was on site and observed the EDG
- inspections. The 1985 EDG inspections were also performed
with a vendor representative present.
The inspector determined that the licensee had not received
, the August 1985 SIL. By the end of the inspection period,
'
the licensee had distributed copies of both SILs and the
'
PBAPS 10 CFR 21 Report to TMI-1 engineering personnel. The
licensee is in the process of reviewing both SILs and the
Part 21 Report for applicability to maintenance and opera- '
<
tion procedure revisions.
During review of the completed ;984 and 1985 EDG inspection
packages (SP 1301-8.2), the inspector observed that scarring
and scratching were indicated for the lobes of both EDGs
scavenging air blowers. Scarring can be an early indication
- . . _- - -- -. - . . - - ,
_
o I
E
o
36
of blower clearance problems. All clearance checks during
the annual inspections appeared to be within the tolerances
specified in the November 1984 SIL. The licensee is recon-
sidering the cause of the lobe scarring as it pertains to
possible blower failure. This problem needs further evalu-
ation by the licensee.
7.4 Conclusion
, At TMI-1, the time restriction on no load operation, along with the
licensee's inspection program, appears to be adequate corrective
action to prevent the type of air blower failure described in the
PBAPS 10 CFR 21 Report. The licensee's investigation into the lobe
scarring and appropriate procedure revisions will be reviewed in a
future inspection report (289/86-PT-01). Generic implication of this
j problem is being reviewed by NRC staff.
8. Exit Interview l
1
The inspector conducted interim exit interviews on March 27, 1986, in the
area of fire protection and on April 2, 1986, in the area of outage
radiation protection. The inspectors discussed the inspection scope and
findings for the entire period with the licensee management at a final
'
exit interview conducted April 11, 1986. The following key licensee
management personnel attended the final exit meeting: ,
J. Colitz, Plant Engineering Director, TMI-1
H. Hukill, Director, TMI-1
G. Tomb, TMI-I Communications
M. Ross, Operations Director, TMI-1
C. Smyth, TMI-1 Licensing Manager, TF
R. Toole, Operations and Maintenance Director, TMI-1
L. Wickas, Operation Quality Assurance Manager, Nuclear Assurance
A representative from the Commonwealth of Pennsylvania, Mr. William
Dornsife, also attended the final exit meeting. A member of Region I
management, Mr. W. Johnston, Deputy Director, Division of Reactor Safety,
also attended the final exit meeting.
T5e inspection results, as discussed at the meeting are summarized in the
- cover page of the inspection report. Licensee representatives indicated
that none of the subjects discussed contained proprietary of safeguards
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 3.3.4, 4.4, 4.5, 4.6, and 6.3.3.
l
1
- - _ - _ _ _ - _ _ - - - _ _ _ _ _ _ - _ _ - _ _ - _ _ _ _ _ - - - _ _ _. _ _ _ _ _ .-. _ _ _ _ _ _