ML20216J581
ML20216J581 | |
Person / Time | |
---|---|
Site: | Three Mile Island ![]() |
Issue date: | 06/22/1987 |
From: | Blough A, Conte R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
To: | |
Shared Package | |
ML20216J568 | List: |
References | |
50-289-87-09, 50-289-87-9, NUDOCS 8707060201 | |
Download: ML20216J581 (80) | |
See also: IR 05000289/1987009
Text
U. S. NUCLEAR REGULATORY COMMISSION
REGION I
Docket / Report No. 50-289/87-09 License: DPR-50
Licensee: GPU Nuclear Corporation
P. O. Box 480
Middletown, Pennsylvania 17057
Facility: Three Mile Island Nuclear Station, Unit 1
Location: Middletown, Pennsylvania j
Dates: March 6 - April 17,1987 l
Inspectors: L. Briggs, Lead Reactor Engineer, Region I (RI)
D. Coe, Reactor Engineer (Examiner), RI
R. Conte, Senior Resident Inspector (TMI-1)
J. Golla, Reactor Engineer, RI
D. Johnson, Resident Inspector (TMI-1)
A. Krasopoulos, Reactor Engi.eer, RI i
K. Murphy, Technical Assistant, RI l
B. Norris, Reactor Engineer (Examiner), RI j
S. Peleschak, Reactor Engineer, RI !
S. Pullani, Fire Protection Engineer, RI l
J. Rogers, Resident Inspector (TMI-1) '
P. Wen, Reactor Engineer, RI
A. Weadock, Radiation Specialist, RI
F. Young, Resident Inspector (TMI-1)
Accompanied By: D. Kubicki, Technical Reviewer, NRR
n ec r: ~
p,
R. Conte, Senior Resident Inspector
6//9//)
'Date
Approved by: k
A. Bloup(, Chief, Reactor Section No. lA, DRP
0 F/
Date
Inspection Summary:
Region I staff conducted safety inspections (800 hours0.00926 days <br />0.222 hours <br />0.00132 weeks <br />3.044e-4 months <br />) of licensee post-
outage, heatup/startup, and power operation activities which included 24-hour
inspection coverage of licensee transition from cold shutdown to power opera-
tion. Shift inspectors focused on overall plant and personnel performance;
startup readiness, along with material condition of the plant; and, technical
8707060201 870623
PDR ADOCK 05000289
G PDR
1
1
Inspection Summary 2
support for the Cycle 6 startup. As a result of normal resident inspector
coverage and 24-hour monitoring, follow-up reviews occurred on: contamination / l
uptake events; feedwater injection event; condenser offgas monitoring operabil-
ity; condensate storage tank level oscillations; pressurizer insulation instal-
lation; steam generator level calibration problems; remote shutdown relay
deficiency; reactor coolant system (RCS) leakrate; reactor water level instru-
,
I
mentation operability; and, quality assurance involvement in Cycle 6 startup.
Review of the following additional areas occurred: post-outage review of radi-
ation protection; post-modification / refueling testing; Heat Sink Protection
System (HSPS); remote shutdown panels; low power and startup physics testing;
special and periodic reports, including the containment integrated leak rate
test report; semi-annual release and effluent data report; licensee event re-
ports; and, annual reports. A special review of related surveillances for the
RCS pressure isolation valves occurred. Substantial follow-up occurred on ,
licensee action on previous inspection findings, especially in the fire protec- "
tion areas with the inspectors assisted by a technical reviewer from the Office
of Nuclear Reactor Regulation (NRR).
Inspection Results:
In general, the licensee made personnel, equipment, and building spaces ready l
to adequately support operations. A number of licensee initiatives contributed !
to this overall readiness. However, the most significant problem was the
alignment / calibration error for the steam generator (SG) level instrumentation
channels of the Heat Sink Protection System (HSPS). This instrumentation was
not fully ready to support operation; and it, in part, caused operator confus-
ion during precritical operations. Several factors contributed to the failure
to identify the SG level instrument misalignment sooner in the prerequisite
process. Compensatory measures were established for decay heat removal con-
siderations with the reactor critical at low power and while the licensee
investigated the SG level indication problem.
Also, as reflected by the number of unresolved items in the design and engi-
neering area, the inspectors continued to note signs of weak technical support.
The open issues centered around: procurement problems; vendor interface prob-
lems; and, weak design due, in part, to apparent lack of consideration for cer- l
a
tain human factors. An apparent violation of regulatory requirements was noted
in that sufficient thermal growth was not factored into the design of the pres-
surizer platform (see paragraph 4.2.6.3). Also, the inspectors continued to
note a reliance on the test program to identify wiring errors as a result of
maintenance / modification work.
I
i
1
Inspection Summary 3 i
1
1
,
The inspectors concluded that the operators conducted themselves in a competent l
and safe manner. Extensive training provided to the operators was apparently ;
conducive to their adequate response to off-normal situations, especially for
problems noted on HSPS SG level. In general, evolutions and testing were con-
ducted in accordance with applicable station procedures. Minor nonadherences
were noted, especially when procedural steps were conducted from memory. As
equipment problems arose appropriate corrective actions, in general, were
initiated.
Overall, test procedures were adequate to perform their intended functions. In
general, the test procedures were properly implemented. For major tests, there i
was reasonably good interface between test personnel and plant staff, espec-
ially in the reactor engineering area. Minor communication deficiencies were
noted between startup and test personnel and other groups, such as operators,
maintenance, and NRC staff. The most significant event as a result of these
communication deficiencies was the inadvertent pressurization of the "B" steam
generator due to three evolutions being conducted simultaneously. The pace of
activities was a factor in this event. There was reasonably good design pre-
diction of measured reactor core parameters.
Based on this limited review, the inspectors verified proper completion by the
licensee of Cycle 6 startup and other related NRC requirements and licensee
commitments. The areas covered were: reactor water level indication system; 1
emergency feedwater system; fire protection upgrades; and, reactor coolant sys-
tem pressure isolation valve surveillance. Additional detailed NRC staff re-
view will be warranted. A large number of past inspection findings in these
areas were satisfactorily resolved by the licensee. For these past findings,
the licensee was responsive, in general, to NRC staff concerns.
Required reports were made in a timely manner. In general, they appeared to be
adequate, but more detailed NRC staff review is planned as time and resources
permit.
One violation involving failure to perform adequate radiological surveys prior
to working in a prefilter room (see paragraph 7.3.2) was identified. This
failure resulted in an unplanned intake of radioactive material by personnel.
Licensee corrective a:tions for the letdown prefilter room events extend over a.
lengthy schedule; however, a number of those events are design related and need
to be well thought out before effective solutions are implemented. The licen-
see appears to need to pursue letdown prefilter cubicle design problems. Over-
all, based on this review, the radiation protection program is considered to be
generally sound. In general, the radiation protection program appears to be
properly implemented. Since the main focus of this inspection was an event
follow-up, some negative findings were expected. A noteworthy characteristic
identified by the inspector was the quality of the licensee's self-review
process in this area.
Overall, the plant was returned to power operation in a relatively smooth
manner.
/
TABLE OF CONTENTS
Page j
i
1. Introduction and Overview . . . . . . .... ...... 4
2. Cycle 6 Startup Readiness . . . . . . . . ........ 5
3. Shift Inspection Activities . ........ .... .. 9
4. Plant Operations. .......... ... ... ... 15- !
l
5. Post-Modification / Refueling Tests . . . . . . . . . . . . . 41 )
6. Pressure Boundary Isolation Valve Surveillance. . ..... 50
7. Radiation Protection. . . . . . . . . . . . . . . . . . .. 53
8. Regulatory-Required Reports . .... .. ........ 63
9. Previous Inspection Findings in the Fire Protection Area. . 68
10. Licensee Action on Other Previous Inspection Findings . . . 79
11. Inservice Testing Program Relief Request Resolution . . .. 85
12. Exit Interview. .......... ........... 86
Attachment 1 - Activities Review
Attachment 2 - RCS Leak Rate Results (3/21-31/87) j
<
i
DETAILS
1.0 Introduction and Overview i
1.1 Background and Purpose
With the shutdown of TMI-1 on October 31, 1986, the licensee com-
pleted the first cycle of operation since the TMI-1 restart and l
entered a scheduled five-month outage for refueling and extensive l
facility modifications. Significant modification work included up- l
grading of the fire protection and emergency feedwater systems. Also,
a number of the restart commitments and TMI Task Action Plan (TAP)- i
items were due to be completed for this startup. In light of the 1
outage length, significant licensee workload, and scope of modifica-
tions, Region I conducted a special readiness assessment team inspec- I
tion (NRC Inspection Report No. 50-289/87-06). The TMI-1 Resident 1
Office continued that assessment, which included 24-hour inspection i
coverage supplemented by Region I-based inspectors for licensee
startup activities.
The purpose of the inspection was to continue to assess the licen-
see's overall readiness for startup. Also, the purpose was to assess
overall licensee control of activities during the transition period j
from cold shutdown to power operations. The main focus of the in- !
spection was on operating staff performance as a result of perform-
ance-based training and on the adequacy of licensee management reso- l
I
lution of technical and safety issues prior to and during plant
startup.
1.2 Licensee Cycle 6 Startup Activities
During the period of March 19-27, 1987, the significant TMI-1 opera-
tional milestones included: (1) plant heatup; (2) partial cooldown
from outside the control room; (3) taking the reactor critical; (4)
completing low power physics testing; and, (5) completing initial
main turbine generator testing and electric power generation up to
80-85 percent of rated power. The chronological summa ry of plant ,
'
operations during this period is listed below.
Date Time Description
3/20/87 11:51 a.m. Containment integrity set in prepara-
tion for start of plant heatup later
that day
3/22/87 9:00 a.m.- Remote Shutdown Panel Test
12:00 Noon
3/23/87 5:50 p.m. Reactor critical
,
E_____'_____________- '- ' '
5
!
Date Time Description
1
3/26/87 5:15 a.m. Main turbine generator tied to the {
regional electrical grid 1
!
3/28/87 4:15 a.m. Achieved 75% reactor power I
I
3/30/87 5:35 p.m. Achieved maximum achievable reactor
power (84%)
1.3 Operational Events
There were no significant operational events during this period.
Minor equipment actuations and problems were addressed in paragraphs
3 and 4.
2.0 Cycle 6 Startup Readiness
During the inspection, the inspectors continued the review of plant equip-
ment readiness and licensee preparations for Cycle 6 ttartup to assess the
readiness of the plant for startup. The specific areas of focus included
safety-related building spaces; outstanding licensee-identified items in
the surveillance, maintenance, and modification areas; outstanding NRC
inspection findings; and selected valve lineups. The objective was to
identify equipment operability problems that could adversely affect safe
operation of the facility. l
The results of this review are documented below.
2.1 Safety-Related Building Spaces
Periodically, the inspectors observed safety-related building spaces
to identify any loose equipment, scaffolding, or other problems such
as fire hazards / housekeeping that could adversely affect the opera-
bility of s&fety related equipment in adjacent areas. The inspection
also included a review of the related procedures which provide admin-
istrative control of such miscellaneous equipment stored in safety-
related areas.
Selected areas of the following safety-related buildings were inspec-
ted: reactor building; auxiliary building; fuel handling building;
intermediate building; diesel generator building; river water screen-
house; and, ;ontrol building.
6 i
l
l
i
i
In general, equipment storage was satisfactory and in accordance'with l
established procedures. Unrestrained equipment without rollers was l
located and positioned on the floor at a safe distance to not ad- i
versely affect safety-related equipment if it fell over. Equipment
on rollers was restrained with wire rope. Scaffolding was minimal
and, where it was installed, it was rigid and restrained to the !
building walls using wire rope and cement anchors. Loose equipment ]
on top of platforms or brackets was minimized and were assumed to '
fall and, therefore, kept at a safe distance from the safety-related
equipment. 1
During plant walkthroughs on Sunday, March 22,1987, the inspector ,
noted certain instances where housekeeping practices and storage of 1
scaffolding (taken down) were poor. Of particular note was the aux- I
iliary building (AB) ventilation room. An excessive amount of re-
moved scaffolding was temporarily placed between the ventilation
filter trains making operator access difficult. In addition, the
inspector noted minor storage problems in other areas, but they did
not adversely affect safe operations of the plant. After the dis-
crepancies were brought to the attention of licensee management, the
concerns were eventually corrected. The storage problem in the AB
ventilation room was corrected later in the week.
In general, the licensee was very successful in removing scaffolding
(with exception as noted above) and cleaning up the plant in a short
period of time. Subsequent tours later in the startup process found
housekeeping to be adequate.
The inspector also noted that the licensee's program to decrease the
number of areas that had slowly become posted radiation or contami-
nation areas produced positive results by an aggressive radiological
cleanup program. The purpose of this program was to allow access to
a large percentage of the plant without requiring substantial radio-
logical controls.
2.2 Outstanding Licensee-Identified Items
The inspector reviewed selected portions of the licensee's applica-
ble corrective action tracking systems to di:termine if any adverse
condition for safety-related equipment operability existed. The
inspector's review included tracking systems for open maintenance job
tickets, open exceptions and deficiencies (E&D's) associated with
technical specification surveillances, and open plant modification
incomplete work list items.
The inspector reviewed the list of open job tickets and discussed all
selected outstanding work items with cognizant personnel. The in-
spector determined that the work required by these jobs would not
have an adverse effect on the plant safety if not performed prior to
returning the plant to operations.
.
. .
... . . .
.. . .
7
The inspector reviewed the list of E&D's noted by the licensee for
all current surveillances. From a sampling review, the inspector
determined that the licensee was properly conducting required sur-
veillances and none of the noted deficiencies or exceptions would
adversely affect plant safety.
The modification incomplete work items list (IWL) was also reviewed.
The inspector found no condition that would adversely affect plant
safety in this area.
2.3 Outstanding Inspection Findings
During the readiness assessment team inspection, the inspectors .re-
viewed the Region I file of outstanding inspection findings to iden-
tify any equipment operability problems adversely affecting sa fe
operation of the facility. No conditions adverse to nuclear safety j
were identified. i
?
2.4 Valve Lineup Verifications
The inspectors independently verified the position of safety-related
valves with the aid of auxiliary and licensed operators. The follow-
ing operating procedures were reviewed:
--
Operating Procedure (OP) 1104-1, Revision 17, dated December 8,1986,
" Core Flooding System;"
--
OP 1104-11, Revision 25, dated December 3,1986, " Nuclear Services
Closed Cooling Water System;" and,
!
--
OP 1104-30, Revision 25, dated January 19, 1987, " Nuclear River ]
Water." I
Additionally, the inspectors independently verified that containment
integrity had been established by witnessing a supervisory lineup of
containment isolation valves and verifying that proper conditions
existed inside the containment. This was accomplished in accordance
with Operating Procedure (OP) 1101-3, Revision 43, dated January 14,
1987, " Containment Integrity and Access." In addition to verifying
valve lineups and proper capping of containment penetrations, other
items were verified to be acceptable on a sampling basis.
--
Proper attachment of removable gratings
--
Missile shields in place
--
Spoo'. pieces removed where applicable
a
i
.
..
.
.. .
. . . .
.. . .
. . . . .. .
. .
. . . . . . . .
.- .
. . . .- .
. . . . -
. . - - - . . - .. ..
8
--
Verification of 30 degree maximum opening for AH-V-1A, B, C, and
D containment purge isolation valves
i
--
Containment sump accessible and free of debris
Several minor problems, such as loose gratings, leaking test connec-
tion caps, and a main steam line snubber which appeared to be blocked,
were noted which were corrected during the lineup or were noted for
later correction or evaluation by licensee personnel. Subsequent 1
inspections of the containment verified that these items were cor- I
rected. The inspectors concluded that proper containment integrity ,
had been established for the components inspected or action initiated !
to correct documented discrepancies.
Also, the inspectors observed the licensee's administratively-con-
trolled valve alignment check for the high pressure injection (HPI)
and low pressure injection (LPI) systems. No unacceptable conditions l
were identified.
1
In general, the valve lists were determined to be accurate and valves '
checked were in their proper position. However, system lineup for
reactor coolant system (RCS) did not reflect the current condition of
several RCS instrument isolation valves and the operator performing
the valve lineup was unaware of the cause of the misalignment. Sub-
,
j
sequent discussions with the Instrument and Control (I&C) lead fore-
man indicated that instrument testing was being performed at that
time and the valves were in the position desired by plant staff.
These valves were later repositioned to support startup.
2.5 Engineering Analysis to Support Startuy
The readiness assessment inspection report identified that a number I
of licensee engineering analysis for the Heat Sink Protection /Emerg-
ency Feedwater (HSPS/EFW) systems remained to be completed and :
issued. The licensee committed to complete this action before Cycle l
6 startup (Unresolved Item Nos. 289/87-06-08 and 09).
The licensee completed these analyses and issued appropriate internal
documentation. The inspector performed a cursory review of these
documents to assure that the licensee completed the stated action and
to assure that there were no obvious safety issues.
The inspector noted last minute modification actions occurred as a
result of these analyses. For example, a single failure suscepti-
bility was identified at the common instrument air supply line for
.
'
the pressure regulator for the turbine-driven emergency feedwater
pump. This was corrected before Cycle 6 startup. Overall, the
analyses substantiated the design criteria for the systems and no
major modification work was needed.
j
1
'
9
The reports varied in length, detail, and thoroughness. The NRC
staff will use additional specialist personnel to complete a more
thorough review of the HSPS/EFW analyses. Accordingly, the adequacy i
of these analyses remains unresolved pending further NRC Region I '
review. l
i
During the course of this inspection, the inspector became concerned i
that the main feedwater isolation on high steam generator level func-
tion of HSPS might be susceptible to a single failure causing main i
feedwater (MFW) isolation to a steam generator. After discussions I
with licensee engineering personnel and review of design drawings,
the inspector concluded that the function was not subject to a single i
failure. In a feedwater path, HSPS trains "A" and "B" control each l
of the associated regulator and isolation valves. Accordingly, both '
Trains "A" and "B" must actuate in order for the valve to close. A !
single failure in either train would not cause the valve to inadver- )
tently close. l
Also, on April 14, 1987, prior to startup, Region I personnel met
with licensee personnel on the deficiencies identified by NRC staff
in the environmental qualification area (NRC Inspection No. 50-289/
87-01). The licensee reported that these deficiencies were reviewed
and none warrart hardware changeouts. Region I personnel were satis-
fied - with the licensee's presentation and a confirmation follow-up I
inspection will occur after Cycle 6 startup. l
l
2.6 Readiness Summary
i
No conditions were identified that would preclude safe operation of !
the facility. Additional issues were identified, in retrospect, con- l
cerning readiness of equipment to support startup and these were
discussed in Section 4.
3.0 Shift Inspection Activities
3.1 Scope of Review and Observations
The NRC's TMI-I Resident Office monitored licensee heatup, startup, i
and power ascension activities on a 24-hour basis from 3:00 p.m. ,
March 19,1987, to 6:00 p.m. March 27, 1987. The NRC shift inspec-
tors assessed the adequacy and effectiveness of operating personnel
performance based on the inspectors' observations of preoperational
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;
-
-
_ - . .
10
i
--
procedures are used and followed as required by plant policy;
--
equipment status changes are appropriately documented and com-
municated 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;
--
logkeeping is timely, accurate, and adequately reflects plant
activities and status;
--
operators follow good operating practices in conducting plant
operations; and,
--
operator actions are consistent with performance-oriented !
training. l
Shift inspectors also assisted in the overall restart readiness re-
view (described in Section 2) by conducting selected valve and
breaker lineup verifications of safety-related systems.
1
The shift inspectors' observations included, but were not limited to,
those shutdown facility operations, reactor plant startup and testing
operations, periodic surveillance activities, and preventive and
corrective maintenance activities listed in Attachment 1.
The shift -inspector assured that any potential safety concern or
regulatory finding was promptly identified to the licensee's shift
supervisor. Those items requiring additional staff review or
follow-up are described in Section 4 of this report.
3.2 Shift Inspection Assessment
i
3.2.1 Plant Operations
The plant operations department continued to maintain good
overall command and control of functions in the control
room. Operators conducted business in a professional man-
ner and the control room layout remained conducive toward
this goal with the segregation of. where work was performed.
Operations department management insisted and, in most,
cases achieved a relatively quiet control room despite a
relatively large number of personnel present at times (up
to two dozen during significant evolutions ur major
testing).
i
j
11
In general, operators imparted a cautious approach to major
evolutions and testing. Ample test briefings were conduc-
ted. Additional crews were used to support complex test or
evolutions. During test delays, operations department took
advantage of that time to run standby equipment as a final
check out of readiness.
Overall, procedures were properly implemented. For major
evolutions or tests, strict adherence was noted. For ad-
ministrative procedures and other technical procedures con-
ducted from memory, errors were made as noted in other sec-
tions of this report. The quality assurance department was
very active in verifying proper adherence to procedures. ,
There appeared to be a respect for the use of procedures.
i
3.2.2 Operator Performance / Training l
Overall, operators were knowledgeable of plant design
(including new modifications) and frequently changing plant
status. However, the inspectors noted a weakness or a lack
of full understanding by the operators associated with Heat
Sink Protection System (HSPS). During the steam generator
(SG) level drain down on March 23, 1987, the "B" SG startup
control valve controlled at 30 inches on the startup range;
yet, operators apparently did not notice (in that the evo-
lution was allowed to continue) that the control should
have been 18 inches (see paragraph 4.2.7 on why 30 inches
was set). Also, a number of alarms came in that appeared
to be subctantially off required settings and, again, there
was no stopping of the evolution. The evolution for the
"A" SG was eventually stopped at 18 inches on the startup
range when the operators thought the SG was beginning to
boil dry (see also paragraph 4.2.7). The "A" and "B" SG
levels were restored to 50 inches on the startup range to
be conservative until additional support arrived during the
daylight hours.
Further, on the night of March 23, 1987, with the reactor
critical, due to technician error, two-of-four operating
range level transmitters were inadvertently equalized, in-
dicating high level; and this resulted in isolation of the
'
startup and main feedwater valves for the "B" SG. Also,
during the above-noted inadvertent feedwater isolation,
i
there appeared to be some initial confusion on the part of
the control room operators on how to defeat both HSPS actu-
ations logics for the hign level isolation function in
order to restore main feedwater. With management guidance,
operators quickly restored main feedwater to the "B" SG,
preventing a plant trip.
12
Overall, the inspectors found the plant staff to be appro-
priately trained and prepared for the plant startup. The
above discrepancies can be expected because of operator
working knowledge unfamiliarity compounded by last minute
engineering changes reflecting poor technical support. In
general, operators were conservative in their response
actions to off-normal events.
The NRC's readiness assessment report noted extensive !
classroom and in plant training for the HSPS and fire pro-
tection modifications. Discussions with plant operators I
about plant modifications and evolutions demonstrated that
the operators were cognizant about these various modifica-
tions along with last minute changes. That training proved
to be fruitful based on operator performance during the
remote shutdown panel test and in response to various pro-
blems with steam generator level indications (addressed in l
other sections of this report). The discrepancies as noted !
above could be expected.
3.2.3 Maintenance and Material Condition
Several problems occurred during the heatup and startup
sequence that required involvement by station maintenance.
The inspectors noted on a tour of the remote shutdown
panels that one RCS pressure meter was pegged high. Also
the same pressure instrument recorder in the control room
was noted to be very erratic. Licensee Instrument and
Control (I&C) personnel were aware of this condition and
the instrument was successfully adjusted to eliminate the
anomaly. This was verified by the inspectors on subsequent
tours.
Further, a tour of the reactor building (RB) identified a
small packing leak on the Pressurized Operating Relief-
Valve' (PORV) pilot valve. This was evaluated by plant
maintenance personnel as acceptable. Prior to startup, the
emergency feedwater flow indication read 80-100 gallons per
minute (gpm) with no flow. The licensee corrected this
problem during final startup preparations.
Also noted was a problem with the indication of flow
through the high pressure injection (HPI) lines to the
reactor coolant system (RCS) with no actual HPI flow. These
flow transmitters and their inaccurate indication were the
subject of a previous inspection findings (289/86-17-04).
13
l
1
The licensee adjusted the subject flow transmitters during :
the outage, but the instrument continued to indicate posi- l
tive flow in two of the lines when no flow existed. A j
Permanent Change Request (PCR) was issued to accept this j
condition but the 10 CFR 50.59 did not explain why it i
existed. When the RCS system was pressurized for heatup {
and startup, the erroneous flow indication was reduced, but j
a positive flow indication still existed. It appeared to I
the inspector that these flow transmitters still require
licensee action to resolve the improper indication. This ,
anomaly is one part of the overall concern for delta-P j
transmitter calibration problems subject to static pressure l
shift as discussed later in this report for the Once- l
Through Steam Generator (OTSG) level transmitters (para- i
graph 4.2.7).
The other maintenance activities during the startup were
completed in an orderly manner. Examples of these were
recalibration of NI-1 (source range), which had exhibited
erratic behavior and repair of leaking fitting on a RCS
flow transmitter.
Generally, maintenance activities during the startup were
accomplished in an acceptable manner, except as noted in I
paragraph 4.2.7.
3.2.4 Surveillance Activities
The inspectors witnessed various surveillance activities
throughout the heatup and startup process. Generally, the
surveillances were accomplished in accordance with accept-
able procedures in a safe manner.
The inspector witnessed portions of the control rod drive
mechanism (CRDM) " Drop Testing" accomplished in accordance
with Surveillance Procedure (SP) 1303-11.1, Revision 14,
dated February 27,1987, " Control Rod Drop Time." During
this evolution, the "B" loop RCS flow transmitter indica-
tion in the controi room was out of service for repairs to
the transmitter. Part of SP 1303-11.1 required recording
.
of RCS flow just prior to insertion of the particular rod
group being tested. This data could not be obtained and an
Exception and Deficiency (E&D) sheet was generated. T51s
was corrected by deleting the requirement to record the
flow, as operators could verify that full flow through the
core existed as evidenced by the operation of all four
reactor coolant pumps. The inspector verified that the E&D
was properly dispositioned. The inspector noted that the
procedure prerequisites could have been more complete by
including a requirement that instrumentation for required
data be in service prior to the start of the test.
I
14
!
The inspector also observed portions of the HSPS surveil-
lance procedure which were being implemented via the ;
Special Temporary Procedure (STP) method. The licensee had 1
generated four SP's for quarterly testing of the HSPS logic. ;
These procedures, SP 1303-11.26, 27, 28, and 29 in draft
form, were accomplished as STP's. Changes to the draft
SP's were made, as required, to ensure that later perform- l
ance when the plant was at power would not result in any
adverse transients from inadvertent emergency feedwater
(EFW) automatic initiation.
l
During one portion of SP 1303-11.27, emergency feed pumps ;
EF-P-2A and EF-P-1 automatically started due to technician !
error (see also paragraph 4.2.7). ~
Although in normal practice the STP's would require an
approved change when a discrepancy was noted, these STP's
were specially written to allow changes to be made during 1
performance. The shif t supervisor was required to approve I
any changes that could affect plant operations. Except for '
the automatic start of the emergency feedwater pumps
(EFP's), no major problems were noted. The Plant Review
Group (PRG) reviewed the completed STP's and verified that
the surveillances for the HSPS logic were completed satis-
factorily prior to plant criticality. The inspectors re- i
viewed the completed STP's to compare them against the new l
HSPS technical specifications and determined that the HSPS !
technical specification surveillances were adequately com- '
pleted. l
l
The inspectors also witnessed portions of the Engineered
Safeguards Actuation System (ESAS) logic testing per SP
1303-5.2. During the performance of this test, a timer
relay for ESAS block RC-2A would not reset. This delayed
testing for a shift while troubleshooting was in progress.
The licensee successfully identified the faulty component -
and replaced the relay and completed the _ test. The inspec-
tors witnessed the remaining portions of the test to verify
that repairs were made correctly.
Generally, surveillance testing during the plant heatup and
startup period was conducted adequately. The licensee used
extra crews of operations personnel to accomplish complex
surveillances while the normal crews maintained plant para-
meters.
-
_ _ _ _ _ _ _ _ _ _ _ _ _ _
_
l
15
,
l
3.2.5 Technical Support
Overall, shift inspectors noted a substantial amount of
support from corporate and site engineering personnel for
the startup. The project engineer for the Heat Sink Pro-
tection System (HSPS) was on site on a daily basis and was
available to resolve last minute problems and questions.
He was actively involved in the resolution of the steam
generator (SG) level calibration problem. Startup and Test
(SU&T) personnel, along with nuclear engineers, worked well
with the shift crews to assure appropriate test briefings
and overall coordination of test activities. The shift
inspectors noted relatively few interface or communication
problems between technical support personnel and the opera- 1
ting crews.
'
Design weaknesses noted during shift inspection coverage ;
were addressed in other areas of this report. !
3.3 Shift Inspection Summary
In general, there was overall good command and control of evolutions.
Operators and technicians were knowledgeable of plant design and
status. In general, procedures were properly followed. Errors were
made, but they were relatively minor and they had minimal impact on
plant operations.
Operations were appropriately supported by engineering personnel.
As equipment problems occurred, support personnel and management were
active in their resolution of those problems.
Additional insights are addressed in paragraph 4.3.
4.0 Plant Operations
4.1 Routine Review
In addition to the shift coverage noted above, the resident inspec-
tors 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, includ-
ing plant housekeeping and fire protection measures;
--
control of ongoing and special evolutions, including control
room personnel awareness of these evolutions;
_ _ _ _ _ _ _
l
16
--
control of documents, including logkeeping practices;
--
implementation of radiological controls; and, >
t
--
implementation of the security plan including access control,
boundary integrity, and badging practices.
The inspectors focused on the following areas:
--
control room operations during regular and backshift hours, in-
cluding frequent observation of activities in progress and per-
iodic reviews of selected sections of the shift foreman's log '
and control room operator's log, and selected sections of other
control room daily logs;
--
areas outside the control room;
--
selected licensee planning meetings; and,
--
various activities listed in Attachment 1.
As a result of this review and shift inspector observations, the
inspectors reviewed - specific events in more detail as noted below.
4.2 Findings / Conclusions J
4.2.1 General
Licensee site and corporate management continued their
detailed attention and involvement in plant operations.
The inspectors noted a number of corporate-based engineer- 1
ing and management personnel on site just prior to and dur- I
ing the startup process to assure last minute problems were
resolved and their respective departments were ready to
support operations. As equipment or test problems arose, 1
site management was active in their resolution.
The most noteworthy startup problem was the calibration / ,
alignment errors associated with the HSPS SG level instru- 1
antation (details in paragraph 4.2.7). It caused a delay
the scheduled criticality time on March 23, 1987, and a
delay in escalating above 5 percent power. That instrumen-
tation was not ready to support startup because of several '
.
factors addressed in that section. Without the root cause
of the problem identified, the licensee made the reactor
critical and simultaneously started a substantial trouble-
shooting effort on SG level instrumentation. However, com-
pensatory measures were. established to assure reactor decay
heat removal capabilities through the steam generators by
conservatively resetting the low level limits function.
i
i
17 !
l
l
I
Overall, housekeeping and fire protection were adequate to i
support startup. A number of problem areas, as noted in {
paragraph 2.1, were discussed with licensee management. In '
general, . the licensee was responsive to NRC staff concerns
in this area.
As reflected by the number of unresolved items listed be-
low, the inspector continued to note signs of weak tech-
nical support for plant operations. The open issues cen- j
tered around: procurement problems; vendor interface prob- '
lems; and, weak design reflected an apparent lack of con-
sideration for certain human factors.
Overall, there was a relative smooth transition from cold
shutdown to power operations.
'4.2.2 Reactor Coolant System (RCS) Leak Rat _e
The inspector selectively reviewed RCS leak rate data for
the past inspection period. The inspector independently
calculated certain RCS leak rate data reviewed using licen-
see input data and a generic NRC " BASIC" computer program
"RCSLK9" as specified in NUREG 1107. Licensee (L) and NRC
(N) data are tabulated below.
I
l
l
1
_ _ _ _ - _ _ .
18
TABLE 1
RCS LEAK RATE DATA
All Values (GPM)
DATE/ TIME (NUREG 1107) CORRECTED
DURATION Lg Ng Ng N L
U U
.3/21/87 1451 0.1642 0.16 -0.03 0.07 0.0717
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/22/87 0003 0.3999 0.40 0.24 0.34 0.3419 l
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/22/87 1550 0.1993 0.19 0.01 0.11 0.1283
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/27/87 0143 0.2752 0.27 0.18 0.28 0.2837
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/27/87 0714 0.1719 0.18 0.07 0.17 0.1713
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> (Lic. Manual Calc.)
3/27/87 2006 0.2103 0.21 0.09 0.19 0.1812
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/28/87 1017 0.4173 0.41 0.32 0.42 0.4232 l
'2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> '
3/28/87 1943 0.2674 0.27 0.18 0.28 0.2827
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/28/87 1947 0.2637 0.28 0.19 0.29 0.2747
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> (Lic. Manual Calc.)
3/29/87 0750 0.2395 0.25 0.14 0.24 0.2408
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/29/87 1653 0.2021 0.20 0.12 0.22 0.2237
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
3/30/87 0700 0.1302 0.13 0.01 0.11 0.1188
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
4/21/87 0106 0.1319 0.13 0.05 0.15 0.1611
2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />
G = Identified gross leakage U = Unidentified leakage
L - Licensee ca.lculated N = NRC calculated
--
.
I
l
19 I
Columns 2 and 3; 5 and 6 correlate + 0.2 gpm in accordance with NUREG 1107.
(Nu is corrected by adding 0.1044 gpm to N u
due to total purge flow through the
the No. 3 seal from RCP's.) '
The inspector concluded that the licensee leak rate deter-
minations were in good agreement with those calculated by
the NRC staff program. However, the inspector pursued an
anomaly in the data brought to the attention of the inspec-
tor by licensee management. Upon review of this data, cer-
tain tests produced results which do not seem to be real-
istic. Certain leakage plus losses values were smaller j
(although minutely) than the unidentified leak rate value.
The inspector tabulated additional data reflected in
Attachment 2.
For the period April 1 to 14,1987, licen'see test data did i
not yield the anomalous results at least until April 21, l
1987. ;
As of April 3, 1987, licensee engineering personnel re-
viewed the anomalous conditions and then tentatively con-
cluded that the effect was due to upgrading of the packing
system for certain valves in the RCS and make-up systems
and capping the leakoff line on several system valves.
However, the document reflecting engineering review of
April 3,1987, did not explain the stopping of this random
anomalous condition, at least until April. 21, 1987
At the end of the inspection period, the Plant Operations
Director noted that the calculation appears to be sensitive
to minute changes in the No. 3 seal purge flow (the 0.1044
gpm term noted in Table 1 above, which is checked on a
weekly basis. Drif ting of the set flow has occurred and
can be correlated to RCS leak rate anomaly. Licensee
engineering personnel stated that they still had this mat-
ter under review. This is unresolved. pending further
licensee and NRC staff review (289/87-09-01). ,
The inspector also noted that invalid leak rate test re-
suits were also retained to the licensee's file.
4.2.3 Condenser Offgas Monitoring
The licensee has modified the radiation monitoring capabil-
ity of the main condenser vacuum pump discharge. This
modification added an additional low range noble gas mon-
itor (RM-A-15) to serve as a back up for RM-A-5. Also, an
instrument was installed to monitor.the discharge flow rate
of the combined exhaust of the three condenser vacuum
pumps.
R
20
Further, the licensee installed an iodine / particulate samp-
ling system for the combined exhaust of the main and auxil-
iary vacuum pump. This modification was installed in re-
sponse to an NRC staff identified concern about the ability
to monitor for radiciodines being released through the con-
denser offgas system in the event of an OTSG tube failure
at power (re: NRC letters, dated January 2, 1985, and
March 7, 1986).
On a sampling basis, the inspector reviewed the proper
installation of these modifications. The inspector re-
viewed completed test data for the new instrument calibra-
tion, applicable surveillance procedures that required
changes as a result of the modification, technical specif-
ications revised as a result of the new system, and con-
ducted a walkdown of the systems to ensure that installa-
tion was in accordance with applicable requirements. The
following documentation was reviewed:
--
TP 366/15, Revision 0, "RM-A-5 Functional Test;"
--
TP 366/16, Revision 0, "RM-A-15 Functional Test;"
--
OP 1101-2.1, Revi.sion 21, dated February 13, 1987,
" Radiation Monitoring System Setpoints;"
--
OP 1105-8, Revision 32, dated December 29, 1986, " Rad-
iation Monitoring System;" ;
!
--
SP 1301-5.9, Revision 18, dated March 23, 1987, " Con- !
denser Vacuum Pump Release Sampling;"
--
- SP 1303-4.15A, Revision 3, dated April 9,1986, " Rad-
iation Monitoring System Monthly Test - Atmospheric
Channels;" and,
--
Technical Specification (TS) Table 4.22-2, " Radio-
active Gaseous Waste Sampling and Analysis Program."
The inspector verified that these systems were installed as ;
required and in accordance with licensee plans and instal-
lation documents. The iodine / particulate sampler had been
installed during a previous outage and had been in service
for most of Cycle 5. Routi_no sampling and analysis of this
pathway had been previously examined by NRC inspectors and
has now been forn' ally incorporated into technical specifi-
cations. No problem with this installation were identified.
- _ _ _ _ - - - _ _ _
21
l
The installation of the noble gas monitors in the condenser i
offgas pathway has experienced some problems. Moisture {
tends to accumulate in the sample lines and this appears to 4
interfere with sampling analysis data. As a example, the I
indications for RM-A-15 have consistently been a factor of
5-10 times ac, high as RM-A-5. This may be due to a water
buildup in the detection chamber. The licensee has insti-' ;
tuted a program to continuously blow down a moisture sepa- j
rator that was installed on the detector inlet line. Re-
. suits to date have not improved significantly.
In addition, the licensee has not installed an associated )
recorder for RM-A-15 due to procurement problems.
The inspector verified that all applicable licensee sur-
veillance and operating procedures were updated to reflect
the status of the new monitoring systems and that new set-
points , vere incorporated into the procedures. Temporary
Change Notices (TCN's) were processed in most cases and !
have yet to be formally incorporated on a permanent basis.
Licensee action to correct the discrepancy between RM-A-5
and 15, install proper recorders, and complete procedure
changes remains an unresolved item (289/87-09-02).
4.2.4 Once-Through Steam Generator (OTSG) Overfill During HSPS
Testing
I
On March 8, 1987, during preoperational testing for the i
HSPS logic channels for the main feedwater (MFW) isolation i
on high OTSG 1evel, the "B" OTSG was inadvertently filled
and pressurized. Licensee SU&T personnel were attempting
to cycle the various combinations of MFW regulating and
block valves. In order to prevent feeding the OTSG, the
main feedwater to the OTSG's supply was required to remain
isolated. This was necessary because one main condensate
pump was running and flow was being maintained through the
feedwater heaters for systein cleanup. The feed and conden-
sate systems were being ; leaned up at the time. The plant ]
was less < 200 F at the time. '
Test personnel were coordinating with operations personnel
in the control room to ensure that when the MFW regulating
and startup regulating valve were opened the associated
block valves were closed. At the time, plant maintenance
was working on MW regulating. block valve FW-V-5B to cor-
rect some spring packing and indicating light problems.
Control room operators opened FW-V-16B and FW-V-17B at the
request of SU&T personnel . This should not have resulted
in flow to the OTSG. Apparently, the operators were not
-
22
fully aware at the time that the breaker for. FW-B-5B had
been closed under " blue tag" control by maintenance per-
sonnel and that, when FW-V-16B reached 80 percent open, an
open signal was given to FW-V-58. This resulted in a flow
path to the "B" OTSG through the MFW regulating valve FW-V-
178. The OTSG and steam header up to the main steam isola-
tion valves were completely filled and pressurized to 125
psig. The operators, when they recognized the problem,
took action to shut the feedwater control and block valves.
The OTSG was depressurized and drained and, subsequently,
refilled. Chemistry parameters were restored to in-specif-
ication values as the water that entered 0TSG was not ac-
ceptable for make-up.
The problem appears to have been caused by a lack of com--
munication between operations and maintenance personnel in
conjunction with the SU&T personnel (performing the valve
testing). The testing was initially planned to have been
completed prior to use of the feed and condensate systems
for plant heatup/startup evolutions. Due to delays in the
completion of the HSPS work and testing, the pre-heatup
activities were being accomplished in parallel with testing
of the same systems.
The inspectors verified that no adverse affect occurred to
the OTSG as a result of the overfill and pressurization.
The OTSG chemistry was returned to correct specifications
by draining and refilling. Subsequent testing of the MFW
regulating and block valve circuitry for high OTSG level
isolation was delayed for approximately three ' days due to
the discovery of an unrelated wirirg problem in the status
light and block / enable switch cirraitry in the control room.
(Final testing of this portion of the HSPS system is dis-
cussed in Section 5 of this report.) The inspector con-
cluded that this problem resulted from the increased pace
of activities due to the compression of the test and start-
up schedule resulting from work and test delays on the HSPS
system.
4.2.5 Condensate Storage Tank (CST) Level Oscillations
In accordance with Restart License Condition No. 3.a, GPU
Nuclear was required to install a safety grade- low-low
level alarm prior to startup following Cycle 6 refueling.
In addition, NRC Order, dated July 18, 1985, and Regulatory
Guide (RG) 1.97 required GPU Nuclear install safety grade
!
23
condensate storage tank level indication prior to startup l
following cycle 7 refueling. The licensee opted to install I
one plant modification that would provide both requirements ;
and committed to complete both requirements prior to start- ;
up following Cycle 6 refueling. i
The licensee installed new safety grade level transmitters
on the condensate storage tank (CST) suction lines to sup-
ply the necessary input signal for both level indications
and low-low level alarm. After installation of the modif- )
)
'
ication, the licensee noted large oscillations in level
indications in the control room.
An engineering evaluation of the problem determined that
fluctuations in the condensate storage tank suction press-
ure induced by fluid flow dynamics caused the noted level
oscillation in the control room. The licensee conc uded l
,
that the modification required rework and would be per- j
formed during the next outage. The licensee also deter- i
mined that.this did not affect the operability of the low-
low level alarm, which still functioned off the new trans-
mitters.
l
The licensee briefed the NRC staff on this problem and
committed to upgrading the tank level indication as re- ;
quired by Cycle 7 startup. The previous tank level inai- !
cation system (which is supplied directly off the. tank) was 1
re-installed. This non-safety related indication was
placed back in operation in the control room. The licensee {
j
notified NRC staff of this problem by letter, dated March
21, 1987. The Office of Nuclear Reactor Regulation will be ,
reviewing this letter.
The inspector noted that the safety grade transmitter in-
stallation was poorly designed. The inspector reviewed the
licensee's evaluation and interim corrective action. The
inspector concluded that previous tank level indication was
more reliable and accurate and, thus, gave better indica-
tion of tank level. It was also determined the low-low l
level alarm was functional and the installed modification '
met the intent of the license condition. This was based on
the understanding that if the licensee was in a condition
that the tank was supplying water to EFW pumps and the
alarm was received, prudent operator action would be taken
to shift to an additional water source.
24
Further, during steady-state flow conditions, the pressure
transmitter oscillated approximately 0.5 feet. This was
confirmed during the remote shutdown panel testing using
EFW with flow past the safety grade transmitters. The
inspector discussed the relaxation.of the previous commit- l
ment with NRR. NRC:NRR staff concurred that the above- I
noted licensee decision met the intent of RG 1.97 and its
related order, along with the Restart Condition 3.a. This
is unresolved pending the satisfaction of RG 1.97 require-
ments for the CST indication prior to Cycle 7 startup
(289/87-09-03),
4.2.6 Installation of Pressurizer Insulation
4.2.6.1 Background
During this outage the licensee elected to replace the
existing metal reflective-type insulation on the pressur-
izer with flexible blanket-type insulation. The new insu-
lation was being insta'11ed to reduce the heat loss from the
pressurizer. To protect the new insulation on top of the
pressurizer from maintenance-related work, that is performed
in that area, a work platform was installed. The platform
was secured to the shell of the pressurizer. The design,
fabrication, and installation of the new insulation and
platform was' accomplished.by a vendor.
On March 17, 1987, while performing a general tour of the
reactor building, the inspector noted that the pressurizer
,
I
spray line, ' aligned across the top of the pressurizer, was
in physical contact with the new platform. From the ar- i
rangement, it appeared the platform may even be lifting the '
spray line. The inspector noted this concern to on-site
plant engineering, who stated that the concern would be
evaluated.
4.2.6.2 Licensee Findings / Actions
On-site plant engir.eering reviewed the noted discrepancy
and determined that - the design and installation had not
factored in the thermal growth of the pressurizer from cold
shutdown temperatures to normal operating temperatures.
All inspections of 'the work were performed while the plant
was in cold shutdown. Measurements by the licensee deter-
mined that thermal growth of the pressurizer had caused the
platform to lift the pressurizer spray line by approximately
1/8 inch. Plant engineering evaluation concluded that the
spray . line had not been subjected to stress that would
adversely affect the spray line.
_ _ - _ ___ -________ __ _ _ . ..
.
.. . .. ..
. . . . . -.
1
l
25
i
The licensee removed part of the work platform, which was ;
under the spray line, to eliminate the interference between
the spray line and the platform. ]
4.2.6.3 NRC Review / Findings
After the initial walkdown, the inspector re-reviewed the l
pressurizer insulation replacement design and installation I
package A25A-53182 and supporting documentation. The in-
spector determined that the licensee was correct in the
assertion that the original design did not properly address
the thermal growth of the pressurizer. Additional licensee /
vendor reviews performed as part of the modification " turn-
over" process did not address this problem. The inspector
concluded that the original vendor design review and subse- !
quent licensee reviews were weak to not identify the prob- ~1
lem.
Failure to properly review and adequately verify the design
associated with the pressurizer platform installation is
considered an apparent violation of 10 CFR 50 Appendix B, j
Section III and the licensee Operational Quality Assurance l
Plan, Section 4.2.12 (289/87-09-04). 1
The inspector reviewed the licensee's immediate evaluation
and corrective action and determined them to be acceptable.
In addition, the inspector performed a second walkdown to
inspect the final field modification and found the pressur- i
izer work platform had been properly modified as required. i
Measuras to prevent recurrence of the event remain to be
addressed.
l
4.2.7 Steam Generator Level Instrumentation Calibration
4.2.7.1 Background
During the period of March 20-23, 1987, a number of events
occurred which were related to the readiness of the Heat
Sink Protection System (HSPS) and, in particular, its asso- i
ciated new steam generator (SG) level instrumentation chan-
nels to support operation. During the recent five-month i
refueling and maintenance outage, eight new SG level trans-
mitters were installed on each of the two SG's. Four of
these level transmitters were installed for the startup
(SV) range and four for the operating range (0P).
,
!
26
The distance between. tube sheets in a SG is approximately
626 inches from bottom tube sheet to top tube sheet. The
SU range covers approximately 400 inches from close to the
bottom sheet (actual level tap is about 6 inches above bot-
tom tube sheet). The OP range indication is 0-100 percent,
but actually covers from 102 inches to 394 inches with
respect to bottom tube sheet.
Both SU and OP range instruments are temperature / pressure
compensated sir ce the reference legs are outside of the
SG's. An uncompensated full (FU) range (2 per steam gener-
ator) covers essentially the distance between tube sheets
and the FU range is primarily used for cold shutdown activ-
ities. The FU range does not interface with HSPS. 1
All channels of FU range are indicated in the control room
(CR). Only two channels of SU and OP ranges, respectively,
are indicatea in the CR due to a second highest reading
channel auctioneering circuit to a CR recorder. All four j
channel trip indicators ' lights for HSPS are indicated in
the CR, Many CR annunciators have shared functions such as
SG's low level /EFW pump start indications.
Prior to the outage, the low level limit (LLL) controllers
for each SG were set at 30 inches on the SU range. Since
this instrumentation was uncompensated, the real level in
i
j
the SG was approximately 18 inches. Since the HSPS SU i
range was temperature compensated, licensee engineering i
personnel decided to set the controller -for the. integrated i
control system (ICS) at 18 inches (for main feedwater con- '
trol). However, the EFW controller setting remained at 30
inches with the new compensated input from the SU range.
Because the LLL during low power operation was so close to
the EFW pump start setting of 10 inches, the licensee ob-
tained NRR permission o(' y TS) to defeat the EFW low level
actuation at < 30 percent power. For post-trip response
considerations, the licensee had provisions to raise the
ICS LLL controller to 30 inches during the power escalation
process.
4.2.7.2 Events
On March 20, 1987, prior to criticality, the licensee was
implementing the (draft) quarterly surveillance procedures
for HSPS by special temporary procedures (STP's). (The EFW
system was required to be operable above 250 F RCS, but
HSPS was not required to be operable until criticality.)
--
.
,
27
During the day shift and while troubleshooting an identi-
fied problem, a- technician inadvertently grounded the
actuation circuit for EFW pump start and the EF-P-2A
started along with MS-V-13A opening that admitted steam to
the turbine-driven EFW pump (EF-P-1). Operators responded
by confirming the inadvertent actuation and secured the
equipment that had started. The surveillance procedure was
later satisfactorily completed.
During the mid-shift of March 23, 1987, as a part of the
rear'or startup procedure, and while decreasing the SG
leveis, the "B" SG reached the ICS controller setting of 30
inches and the controller adequately maintained SG 1evel. !
Further, the "A" SG reached the control setpoint, which was
set at 18 inches of level,-and the SG pressure was noted to a
decrease by approximately 35 psi. (The discrepancy between
level settings was later reviewed by NRC staff as noted
below.) The pressure decrease caused the shift supervisor
to be concerned that primary-to-secondary heat transfer was
inadequate. With the level control setpoint at 18 inches,
a concern was that, with a possible level transmitter error,
the SG may have been boiled dry. Operators responded by
reseting both SG levels to 50 t .hes until day light hours
when additional technical support personnel arrived.
'
During the day shift, with licensee management present,
operators, subsequently, slowly lowered the SG level down
to 18 inches and then confirmed the 18-inch setpoint to be
adequate for proper heat transfer. But, there were signif-
icant discrepancies between SU level channels that caused l
licensee personnel to suspect the. instrument channel cali-
brations. Licensee management decided that, despite the
discrepancies, the reactor could be safely made critical
with a conservative LLL setting of 30 inches cn ICS con-
trollers.
The licensee also initiated troubleshooting efforts on each
SU channel instrumentation and this included procuring the I
services of the vendor (Foxboro) to assure proper setting
and alignment of the instrumentation. While low power
physics testing was conducted, the licensee concurrently
conducted troubleshooting efforts on SU (and OP) range ;
channels in order to get more consistent readings between '
channels. i
I
l
i
I
__
28
i
During the swing shift of March 23, 1987, at approximately
6:00 p.m., a technician was isolating an OP range transmit-
ter for the above noted troubleshooting effort when he
inadvertently equalized two transmitters sharing common '
reference and variable legs. With the transmitter legs
equalized in pressure, this simulated high level in the SG l
and the two-out-of-four HSPS logic was satisfied to isolate l
MFW to the "B" SG. Operators responded to defeat both HSPS
trains for this function and they restored MFW to the SG l
without causing a reactor trip. At approximately 7:30 p.m., ;
a similar event occurred causing operators to defeat the
actuation train and restore MFW to the "B" SG, again, with-
out causing a plant trip or significant plant upset.
4.2.7.3 Licensee Review / Findings
The licensee was in the process of writing a post-outage
review report and they plan to address many of the problems l
associated with the readiness of SG level instruments to i
support startup. Licensee tentative findings listed below
were based on information gathered during the week of
March 23, 1987, and they centered around the SG level
instrumentation channel discrepancies.
A detailed engineering analysis was performed by the licen-
see and documented in an internal memorandum (Serial No.
3300-87-0065), dated March 26, 1987. It was determined ;
that the SG pressure dip which -initiated the concern for l
reduced heat transfer was, in fact, caused by an increase 1
in feedwater flow and not be ' decreased heat transfer. The I
increased feedwater flow was in response to indicated SG
level going below the LLL (18 inches) of the ICS control- ,
ler. The licensee concluded that the "A" SG did not boil l
dry.
The licensee's initial review determined tha; the primary
cause for the errors in the level sensing systems appears
to be the static zero shift associated with the level
transmitters. This occurs when the instrument must operate
at static system pressure (1000 psig), which is substan-
tially different from the static pressure during calibra-
tion (essentially atmospheric, 0 psig). This is accounted
for by static pressure alignment procedures. The initial
instrument calibration data for all sixteen installed
transmitters were found to be well within the 0.5 percent
acceptance criterion,
j
i
29
However, the licensee determined that during the initial
calibration, no static alignment was performed on site.
Licensee SU&T personnel believed that, since the static
alignment was specified on the purchase order, it was per-
formed by the vendor; and, since no components in the
transmitters had been changed, no further static alignment i
was needed. The vendor manual supports this assumption.
'
Also, vendor representatives stated static alignments are
performed prior to shipping new transmitters.
Subsequent checks following the identification of the inac-
curate level indications showed that some static alignment
zero shifts had occurred. Following the zero shif t correc-
tion and recalibrations, satisfactory calibrations were
obtained. But in some instances a zero shift resulting
from the altering of 'a magnetic field was noted. This
magnetic field, associated with the transmitter operation,
appeared to be altered after the topworks cover was rein-
stalled for which this had to be compensated.
One transmitter in the OP range (LT 1041) sustained some
damage, apparently during or after installation, and a
lengthy repair and recalibration effort was required. As-
sistance from the vendor in the repair of this transmitter
and in resciving other difficulties was required.
4.2.7.4 Scope of NRC Review
The NRC shift inspectors followed licensee actions as they
occurred. In addition, a region-based inspector followed
licensee troubleshooting actions and their review of the
above-noted events to assure that underlying causes and
corrective action (lessons learned) were taken. Additional
resident inspector review occurred subsequent to the com-
pletion of the power escalation test program.
4.2.7.5 NRC Findings
4.2.7.5.1 Human Factors / Design
The inspector noted that the HSPS channel indicators for SG
1evel do not easily lend themselves to cross checking with
only two channels repeated in the control room. Al so, the
recorded channel scale is difficult to read. Of the four
SU channels, one has a' digital indication and only one of
the remaining three indicates and records, as determined by
a mid-channel select switch. The OP range has a similar
arrangement. Also, two additional channels per SG indicate
i
l
30
l
'
1
uncompensated SG (reads higher than compensated level) and
may have been used by the operators during the level reduc-
tion (being easier to read and trend), adding to confusion ,
on what the proper alarm settings were. Further, the chan- )
nel feeding the ICS system may not always be indicated in 1
the CR. Also, the chart posted on the center console for
correlating SG level ranges appeared to be outdated at the
time of startup. Compounded by significant differences in
channel indicators, operator confusion resulted. The
licensee's design review process was weak in not identify- 1
ing these deficiencies earlier.
Further, subsequent to the identification of the above-noted
calibration problems, the inspector reviewed the surveil-
lance tests associated with the weekly channel checks of
the SG level and EFW valve controllers. Both of these pro- ,
cedures were noted to require some changes in order to '
clarify their intent. During inspector witnessing of a-
check, an operator inadvertently compared a channel to it-
self but later corrected himself. It was noted that the 1
'
test compares only the two channels indicated in the CR.
The NRC staff gave the licensee a position that all four
channels of HSPS must be channel checked in order to satisfy
the TS requirement. The licensee representatives initially
disagreed, but they complied with the NRC staff position at
least until full confidence in the instrumentation has been
established. This entails using test equipment for four
channel readings. Surveillance procedures associated with !
required weekly checks were improved. ,
The adequacy of the licensee's two-channel check methodology
is unresolved pending completion of licensee review and/or
licensee decision to relax his current procedures for a
four-channel check and subsequent NRC Region I review (289/
87-09-05)..
Improved operator awareness of the CR indications has re-
sulted from the licensee's - review of the level problem.
The new LLL setpoints associated with the instrumentation
(30 inches) were evaluated and found to be conservative and
acceptable provided the level transmitters meet their 0.5
percent acceptance criterion.
4.2.7.5.2 Steam Generator Level Installation and Alignment
The overall findings resulting from the licensee's effort
to establish accurate level indication systems are that a
number of factors appear to have contributed to the ob-
served level errors.
i
31 i
There are:
--
the failure to perform a zero static alignment;
--
the apparent failure to fully account for the zero
shift due to cover reinstallation;
--
all the alarms and actuation occurred within the lower 1
5 percent of the instrumentation range where a zero
static shift error would have most effect; and,
--
last minute design setpoint changes for OTSG LLL set-
ting in which the reason for these changes were not
clearly understood and/or warranted.
The initial calibrations and startup testing performed was
believed to be adequate to assure instrument accuracy.
l
Initial calibration data and performance of a quarterly '
surveillance test show all acceptance criteria were met.
The quarterly Technical Specification surveillance test
associated with steam generator level, Procedure 1303-11.37,
"HSPS-0TSG Level and Pressure Channel," was performed prior
to plant startup and the results were acceptable. The
static zero shift and apparent magnetic effects occurred i
after the calibration / initial alignment. The vendor repre-
sentative apparently " fine tuned" the licensee's calibra-
tion methodology to assure that all channels gave consis-
tent and accurate readings. This included a static zero
check and allowance for zero shift due to cover reinstalla-
tion.
The inspector noted that the licensee has had a history of
instrument alignment problems, apparently due to zero static
shift. The NRC Inspection Report Nos. 50-289/85-21 and
85-28 documented a similar problem with the EFW flow in-
struments. (The licensee corrected a recent reading of
80-100 gpm with no flow in the system just prior to the
Cycle 6 startup.) Further, the high pressure injection
(HPI) flow instruments give a positive reading periodically
with no flow in the lines (re: NRC Inspection Report No.
50-289/86-17). This reoccurred during this startup.
. __ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
1
32
I
The adequacy of licensee calibration of differential press- l
ure transmitters used in high pressure service is unresolved I
(289/87-09-06) pending: '
--
licensee review and incorporation into their proced-
ures of a methodology that produces consistent and
accurate instrument readings for the above-noted
problems;
--
incorporation of any lessons learned into RG 1.97
modifications scheduled for Cycle 7 and 8 refueling
outages (RG 1.97 requires the upgrading of instruments
used to monitor accident conditions); and,
--
NRC staff specialist review of this area.
4.2.7.5.3 Work Control
The licensee controlled the troubleshooting work associated
with the accuracy of the level instruments using Job Ticket
(JT) CM 260, which included a general maintenance procedure
1430-Y-17, " Differential Pressure Transmitter Repair and ;
Calibration." The vendor instruction manual was also used J
(see paragraph 4.2.11 on Qualification Assurance (QA) find- ;
ings in this area). This procedure appeared to be adequate {
to control the activities being performed. However, due to i
technician error, two inadvertent feedwater isolations on
SG 1evel occurred. The above-noted procedure appropriately 1
described how transmitters are to be removed- from service.
The technician apparently implemented this procedure from
memory
In general, the licensee's control of safety grade work
associated with the checking and recalibration of the level l
instruments was adequate.
The inspector also reviewed why the ICS controller for the
"B" SG level was set at 30 inches, while the controller for
the "A" was set at 18 inches. The licensee established
that an engineer and technician performed an apparently
unauthorized change to the "B" controller setting just
prior to March 23, 1987, in the interest of troubishooting.
The particular controller is in a non-safety-related system.
The low level limit for the SG is important to safety in
that it provides heat sink protection. The HSPS is the
safety grade system that backs MFW to perform that func-
tion. Had the ICS controller been nonconservatively set,
a challenge to a safety-related system could occur.
i
33
The apparent technician / engineer error reflects a poor at-
titude toward control of work processes for non-safety-
related systems. This attitude is also repetitive of other
apparently isolated instances where work was not proced-
uralized. The licensee's lack of a document review of this
event also reflects poorly on overall licensee corrective
action systems. Routine review of this area will continue
by the NRC TMI-1 Resident Office.
1
4.2.7.5.4 Modification Completeness '
During the above-noted SG 1evel alignment problems, the
licensee revealed a concern about placing the HSPS MFW iso-
lation function in " enable." Apparently, as a result of
the licensee's review of the Davis-Besse Loss of the Feed-
water event of June 1985, the licensee became concerned
that steam pressure oscillation may also cause significant
oscillation in the HSPS SG (OP range) level indication sys-
tem. This could result in an inadvertent MFW isolation on
a post-trip situation, thereby compounding the operator
post-trip response actions. Electronic filters for the SU
and OP ranges were ordered but not received prior to Cycle
6 startup, and they were not expected on site until the end
of April 1987. The electronic filters were supposed to 'I
damper these oscillations.
As a result of the above concern and in response to NRC
staff questioning, the licensee conducted a 10 CFR 50.59
review for leaving the HSPS MFW isolation in defeat. The
licensee's evaluation indicates that the function is anti-
cipatory to prevent main steam line flooding. The licensee
concluded that they were justified in placing the HSP3 MFW l
isolation function in defeat. This evaluation was for-
warded to NRR by Region I for a more detailed review.
Licensee representatives gave an initial position that the
function could be defeated for the rest of the operating '
cycle (ends July 1988). The NRC staff disagreed and stated
that once the ' electronic filters were installed and opera-
ting properly, the MFW isolation should be " enabled." The
licensee's position was apparently based on the fact that
no TS were issued on this system. The inspector also noted
that no TS were issued on this function. However, it was
clear to the inspector that the TMI-1 hearing record sub-
stantiated that the safety grade function be installed and
operable. Based on a limited review, its safety function
appeared to an anticipatory.
__
34
The return of the HSPS MFW isolation function to " enable"
is unresolved pending completion of licensee action as noted
above and subsequent NRC Region I review (289-87-09-07).
The adequacy of the licensee's above-noted 10 CFR 50.59
safety evaluation and current TS on the HSPS MFW isolation i
function is unresolved pending additional NRC staff review
(289/87-09-08).
4.2.7.5.5 Event Review and Follow-up
The inspector noted five significant events that occurred
during the startup dealing with HSPS system.
(1) OTSG level initial calibration and alignment problem
identified during startup operations on March 23, 1987.
(2) On March 23, 1987, the "B" 0TSG LLL controller was set
at 30 inches with the "A" controller set at 18 inches.
(3) Three inadvertent actuation events, one on March . 21,
1987, and two similar events on March 23, 1987.
Licensee management acknowledged that there were lessons
learned from item (1) above. They stated that a number of
issues will be documented in their post-outage review re-
port.
Based on discussions with licensee personnel, it was not
clear whether or not the above-noted report would include ,
the issues /iessons learned / corrective actions from items '
(2) and (3) above. Item (2) is addressed in paragraph I
4.2.7.5.3.
With respect to item (3), the inspector questioned the re-
portability of these events with respect to 10 CFR 50.72,
inadvertent engineered safety features (ESF) actuations. l
The licensee took a position that EFW/HSPS actuations are l
not reportable because they are not listed in their Final !
Safety Analysis Report (FSAR) as ESF systems. Based on
discussions with NRR staff and Region I management, the
inspector disagreed and stated that these systems are engi-
neered features that perform a safety function and that
inadvertent actuation of EFW/HSPS are of interest to NRC.
As one of the users of information provided by 10 CFR 50.72,
NRC Region I management indicated that NRC considers these
reports important to NRC in fulfilling its responsibili-
ties. The inspector stated that the FSAR may be outdated
with respect to current NRC staff positions. This is unre-
solved pending further NRC staff review (289/87-09-09).
-_
"
35
4.2.8 Inadvertent Failure of Class J Control Relays l
)
4.2.8.1 Background
On April 2, 1987, Region I received a 10 CFR Part 21 notif-
ication from a vendor, Telemecanique Corporation, that
Class J control relay magnet block (J20M) manufactured dur- l
ing the eighth week of 1987 did not meet the manufacturer's
design specifications. These relays were supplied to TMI-1
for the remote shutdown panel (RSP) installed for 10 CFR 50
Appendix R fire protection rule. During shop manufactur-
ing, the air gap associated with the magnet coil was too
large, leading to high currents and, perhaps, early failure
of the relay.
The manufacturer had noted this failure from relays that
had been installed in a fossil-fueled plant. Due to the 3
large , air gap, when energized, the relays draw more current. )
After prolonged use under full current, the relay could i
fail and open the closed contacts. l
On April 3, 1987, the resident inspector informed the l
licensee of the potential early failure of this type relay. j
The licensee responded that there are a total of sixty-
three J20M relays used in these RSP cabinets that are. nor-
mally de-energized and energized when control is trans-
ferred to the RSP. The licensee immediately began a review
of the concern. A test of four spare relays was set up to.
determine the electrical characteristics of the defective
relays and the time for the failure to occur. l
4.2.8.2 Licensee Findings / Actions
The licensee's discussions with the vendor determined that ;
the relays installed at TMI-1 were the defective relays. !
Typical current usage by a J20M relay when energized was !
.15 to .19 amperes. From the licensee's test setup, the
four spare relays were drawing approximately .2 amperes.
The increased current confirmed that these test relays do
have an increased air gap associated with the magnetic coil.
In addition, the licensee confirmed that failures of the
relays may occur as early as six full days after energiza-
tion. A relay failed to the de energized state within a
week on the test stand.
During a conference call with Region I on April 14, 1987,
the licensee committed to the following:
,
36
--
replace the relays during the next unscheduled shut-
down which would place the plant in cold shutdown;
--
on a sampling basis, veri fy the replacement relays
meet proper electrical specifications in order for the
relay to be considered qualified; ,
--
determine which relays would be required to operate if
the licensee was to be in a long-term core cooling
from the remote shutdown panel and determine how to
mechanically override the relay to remain in the ener-
gized position and to formalize this action in a pro-
cedure and,
--
brief operating crews on the potential problem.
Operating areas had been briefed. At the close of this
report period, the licensee had obtained replacement relays
and was in the process of revising operating prscedures to
address how to override the relays in case of failure.
4.2.8.3 NRC Findings / Reviews
The inspector with regional assistance reviewed the licen-
see's corrective action and interim measures to ensure that
their actions did not adversely affect plant safety. These
reviews included the review of the informal testing on
spare relays, temporary changes to procedures as described
above, and discussions with operators. The inspector con-
cluded the licensee's approach was proper and proposed
interim measures were acceptable. Plant operators had been
briefed and were knowledgeable about the additional actions
that were required if a relay would fail while operating
the RSP.
A review of the licensee's acceptance criteria and field
inspection of the new relays determined that the noted dis-
crepancies with this type of relay was an isolated case and
the new components met the applicable standards. This area
is unresolved pending completion of licensee action as noted
above and subsequent Region I review (289/87-09-10).
4.2.9 Reactor Water Level Indication
The Order for Modification of License for TMI-1, dated
December 10, 1982, to accommodate additional instrumenta-
tion to detect inadequate core cooling required the licen-
see to design and install a post-accident reactor water
level indication system, licensee designated " reactor cool-
ant inventory tracking system" (RCITS), to conform to the
37
i
)
.
design specified in NUREG 0737, Item II.F.2. Prior to using
the installed RCITS as a basis for operator decisions or .
actions, final documentation shall be submitted for NRC - i
review and approval. The licensee has installed their RCITS l
and submitted applicable documentation. The NRC staff is j
in the process of reviewing the design. The licensee has '
placed the RCITS in operation.
The inspector reviewed the status of the RCITS and associ-
ated procedures to ensure that the system was functional
and presently no operator action was being taken solely on
information obtained from RCITS. The inspector reviewed OP
1103-1, Revision 2, " Reactor Coolant Inventory Tracking
System," and PCR 7-05-87-0298 and SP 1301-4.1, Revision 36,
" Weekly Shif t Checks," to determine how the licensee had
incorporated the use of RCITS. In all instances, the
inspector noted that operation decisions were not based on
RCITS indications.
Based on the above, the inspector concluded that the system !
was functional. Discussions with licensed operators and !
computer technicians found personnel knowledgeable and
aware of the restricted use of RCITS. Final review of the
installation and procedure implementation on the use of
RCITS will be conducted after the NRC Safety Evaluation
(SE) on RCITS has been issued.
Also, during the design review of RCITS, the Office of
Nuclear Reactor Regulation (NRR) identified an apparent
single failure susceptibility because all channel indica-
tions are processed through a single computer interface
" multiplexer" for control room indication. Accordingly,
the NRR staff obtained a commitment from the licensee to
proceduralize their ability to take safety grade voltage
readings at the (local) signal conditioning cabinets (up-
stream from the multiplexer in the instrumentation loop).
By letter, dated March 19, 1987, the licensee committed to
proceduralize these local voltage readings to reactor water
level by October 1, 1987.
The inspector questioned why it would take so long to
change a facility procedure to provide the correlation that
should already be known, assuming operability of the RCITS.
Licensee representatiaves indicated that a procedure cnange
request (PCR) was being processed and that it should be
issued by the end of May 1987. The inspector has no addi-
tional comments.
_ - ___-__ _ _
l
1
38 j
4.2.11 Quality Assurance Department Monitoring
On a sampling basis, the inspector reviewed quality assur-
ance (QA) monitoring activities to assess the extent of
their involvement in monitoring licensee activities in pre-
paration for an during the Cycle 6 process. Particular
focus occurred on QA review of procedure adherence. 'In
addition to observing QA monitors in the plant, the inspec-
tor reviewed a sampling of QA monitoring reports for the l'
outage and startup period. Particular documents reviewed
are listed below.
--
Licensee Internal Memorandum No. 6112-87-054, dated
March 23, 1987, "6R Outage Monitoring Summary"
l
--
Quality Deficiency Report (QDR) No. HRH-24-87, dated ;
April 1, 987, on apparent failure to follow, in part, l
administrative controls for generic test procedures )
--
QDR No. HFT-025-87, dated April 2,1987, on apparent
failure to follow guidance when silica and cation con-
ductivity parameters exceeded specifications for steam
generator chemistry
--
QDR No. HRH-026-87, dated April 3, 1987, on apparent
failure to properly control vendor manual . and work in ;
accordance with procedure for steam generator level ;
calibrations '
--
QDR No. HRH-028-87, dated April 7,1987, on apparent
failure to follow engineering administrative control
procedures for using approved supplier of services and
in not properly classifying related procurement docu-
ments for the procurement of vendor services for steam
generator level calibration
--
Selected Quality Assurance Monitoring Reports (QAMR) i
for the period March 15 to April 13, 1987
--
Index of QAMR's for the period March 15 to April 13,
1987, reflecting tho completion of approximately 125
QAMR's
The monitoring activities included several functional areas
besides plant operations, and they included: maintenance;
surveillance; startup and test; security; and, overall
administrative control implementation. The focus of these
39
monitoring reports was to assure procedure adherence, not
so much as to question the adequacy of procedures, although
monitors opted to comment on such things. The QA personnel
were on 24-hour coverage during the transition from cold
,
i
shutdown to power operations.
Throughout the outage, there was evidence of backshift and
weekend coverage by this group also.
A large number of procedural steps were verified to be pro- 4
perly completed. For the above-noted one month period, '
there were approximately 125 monitoring reports each with
one to five individual procedural step verifications. Minor i
nonadherences were corrected "on the spot" and the more !
significant or uncorrected ones were documented by quality )
deficiency reports as noted above. The inspector estab-
lished confidence that the QDR process was a reasonable
indicator to measure enhanced performance with respect to
licensee corrective actions (currently incomplete) for the
procedure adherence problem noted in past inspections. It,
also, signifies a detailed QAD involvement in licensee
activities.
The above-noted QDR's for the SG level problem have been
noted independently by the NRC inspectors. In fact, all of
the above-listed QDR's are of significance and their reso-
lution is of interest to the NRC staff since the licensee
is to be credited for self-identification of these issues.
Accordingly, the resolution of the above-noted QDR's is l
unresolved pending completion of licensee action and NRC !
Region I review (289/87-09-11).
Based on the above review and on a licensee internal report,
it appears that corrective actions to date were somewhat
effective in reducing the relative number of nonadherences.
In fact, the nonadherences, which appear to be few, seem to
be centered around not properly following administrative
requirements or in plant operational guidance. In fact,
the NRC staff, during its 24-hour coverage period, noted
strict adherence to technical requirements such as evidenced
in operating, surveillance, or maintenance procedures in
distinction to contrary findings during past intense per-
iods of NRC staff review. Nonetheless, although improve-
ment was noted, there were still signs that when it became
difficult to comply or procedure _ requirements impeded pro-
gress, the requirements were unintentionally bypassed rat-
her than properly changing the requirement if needed as
reflected in the above-noted QDR's. The NRC staff will
continue to monitor licensee efforts in the procedure ad-
herence area.
\
40
l
i
Also, the inspector reviewed the quality control group's i
involvement in Cycle 6 startup. Their activities centered
on the identification and closecut of inservice inspection
open items and outstanding material nonconformance reports
that were needed. to be resolved by startup. Also, there
was a " hold point" to sign off various modification " tie-
in" documents. attesting to the readiness of the modifica-
tions to support system operations.
The inspector reviewed a sampling of plant inspection re-
ports (PIR's) as evidence for that involvement in startup
activities. The PIR's have minimally required review ele-
ments for the quality control inspector; such as: proper
signatures authorizing work; controlled document usage;
calibrations traceable to national standards; QC acceptance
of replacement parts, if applicable; and, proper procedure
adherence. Other technical criteria are added commensurate
with the scope of review. The inspector identified no con-
dition adverse to quality assurance requirements.
Further, the inspector reviewed the audit group's involve-
ment in Cycle 6 startup. The NRC's readiness assessment
team inspection addressed this area with respect to modifi-
cations. There was no direct involvement.by this group in
the startup, since that is generally reserved for the modi-
fication/ operations and quality' control (QC) groups. How-
ever, the aud,it group indirectly supported startup by doing
a refueling audit, which independently confirmed proper
core configuration.
The inspector reviewed the audit. It was quite detailed
and thorough and auditor comments were appropriately ad-
dressed. There were no " audit findings". The inspector
identified no conditions adverse to quality assurance re-
quirements in this area.
4.3 Plant Operations Summary
In general, the inspectors concluded that the operators conducted
themselves in a competent and safe manner. Licensee management and
the Quality Assurance Department continued their presence and in-
volvement in plant activities. In general, evolutions and testing
were conducted in accordance with applicable station procedures.
Minor nonadherences were noted, especially when procedural steps were
conducted from memory. As equipment problems arose, appropriate cor-
rective actions were initiated. Extensive training provided to the
operators was apparently conducive to their adequate response to off-
normal situations, especially for problems noted on the H5PS SG level.
<
!
I
l
41
)
)
Based on this limited review, it appears that the licensee met com-
mitments made for Cycle 6 startup. The NRC staff's detailed review
of certain related modifications remains to be accomplished.
The new SG level instrumentation was not ready to support operations.
The reactor was taken critical without the licensee effectively l
understanding and resolving the problem. Two primary factors con-
tributed to the failure to identify the misalignment sooner: poor j
implementation of measures to assure proper alignment and calibration I
and/or the lack of an integrated functional testing of the new level
instrumentation. The pace of activities may have contributed to the
first factor.
)
1
As reflected by the number of unresolved items noted above, the in- i
spectors continued to note signs of weak technical support for plant j
operations. The open issues centered around: procurement problems; j
vendor interface problems; and, weak design reflected an apparent 1
lack of consideration for certain human factors. Also, the inspec- l
tors continued to note a reliance on the test program to identify l
wiring errors as a result of maintenance / modification work. Addi- l
tional experience in the maintenance / surveillance areas should reveal
if problems were effectively identified and resolved by the licensee.
A number of licensee initiatives were noted. Additional crews were
used to assist in complex evolutions or testing. The quarterly sur-
veillance procedures for HSPS and the RSP test were satisfactorily ,
conducted prior to startup and they provided a learning experience j
for plant staff. Plant equipment was rotated as final check on read- {
iness for equipment normally in a standby mode. Continuous coverage I
was provided by the QA department. l
l'
Overall, there was a relatively smooth transition from cold shutdown
to power operations.
5.0 Post-Modification / Refueling Tests 5.1 Integrated Test for the Remote Snutdown Panels (RSP)
By letter, dated February 19, 1987, the licensee outlined their plan
to perform a limited integrated test for the Remote Shutdown Panels ;
(RSP's) to meet their commitment as documented in NRC Inspection '
Report No. 50-289/86-14, paragraph 3.2.5. The licensee prepared Test
Procedure (TP) 683/1, Revision _0, " Remote Shr.down Outside Control
Room," and performed this test on March 22, 1987. The inspector (s)
reviewed this procedure and witnessed the test. The details are
discussed below. '
j
-.-
!
.i
!
42
l
5.1.1 Procedure Review
The scope of review was to ascertain that the procedure met
the licensee's commitments documented in NRC Inspection
Report No. 86-14, paragraph 3.2.5 and further discussed in
their letter, dated February 19, 1987. The procedure meet
these commitments. Specifically, the objective of the test
as stated in the letter was to demonstrate the ability of
operations personnel to transfer plant cooldown controls
from the control room to the RSP's and maintain stable hot
shutdown conditions at approximately 532 F using reactor
coolant pump heat. The reactor coolant system (RCS) tem- 1
perature was then to be reduced approximately 30 F and l
maintained for thirty minutes using controls on the RSP's. l
The procedure also met the prerequisites, test conditions, !
and other test details discussed in the letter..
No unacceptable conditions were identified.
5.1.2 Test Witnessing
The licensee performed this test on March 22, 1987, and
four NRC inspectors witnessed the test. The scope of wit- l
nessing included a verification that the test was conducted 1
in accordance with the approved procedure and that the test
results were within previously established acceptance cri-
teria. The acceptance criteria included the ability of
operations personnel to transfer plant cooldown ' control s
from the control room to the RSP's and maintain stable hot
shutdown conditions at approximately 532 F for thirty min-
utes and, then, reduce the temperature by approximately i
30 F using RSP controls. The second criterion was to
demonstrate the feasibility of performing the above opera- -
tions using the normal shift crew (five people). The third 1
criterion was to demonstrate the adequacy of plant proced-
ures used to perform the shutdown outside the control room.
The inspectors concluded that the licensee met all accept-
ance criteria and also the commitments in the February 19,
1987, letter.
No unacceptable conditions were identified.
5.2 Heat Sink Protection System (HSPS)
5.2.1 Criteria and Scope of Review
During this inspection, the insped.or completed the review
of preoperational test procedure TP 332/3, Revision 0,
"HSPS Integrated Functional Test," as discussed in NRC
i
- --
- - - = - - - - ' ' '
'
43
Inspection Report No. 50-289/87-06. The same criteria dis-
cussed in paragraph 5.2.1.1 of that report were used for
this procedure review. In addition, the inspector wit-
nessed a portion of the performance of TP 332/3, utilizing
the criteria stated in paragraph 5.2.2.1 of that report.
5.2.2 Findings and Conclusions
5.2.2.'1 Test Procedure 332/3 Review
l
The inspector found the test procedure to be technically
adequate to functionally test the HSPS system with one ex-
ception. During the review, the inspector noted that the
backup power supply was not being functionally tested.
This item was previously discussed in paragraph 5.2.1.2 of
NRC Inspection Report No. 50-289/87-06. During that in-
spection, the test engineer had agreed to incorporate back-
up power supply testing into TP 332/3. j
i
During discussions with Startup and Test (SU&T) personnel, )
it was pointed out that' the backup power supplies supplied
power to the "A" and "B" trains only and not the whole sys-
tem as previously thought. No logical reason for the power
supplies could be provided at that time. Previous thoughts j
were that they would prevent spurious trips of the HSPS l
systems. This did not now appear to be the reason since l
the trains must energize to actuate and the channels de- i
energize. The backup power supply would, in fact, allow
HSPS to actuate if more than one vital power supply was
lost and the backup supply remained energized.
l
The inspector requested copies of the latest System Design i
Description (SDD) and the latest installation specification
for the HSPS. The inspector also contacted the HSPS pro- 1
ject engineer and requested any information available that !
would identify the purpose of the two backup power supplies
and what testing should be performed to verify their func-
tion. From the information provided and discussions, the
inspector was able to determine that one of the supplies
(T-1181) provides backup power to trains "A" and "B". The
purpose of that backup power supply is to prevent loss of
the signals passing through the train "A" cabinet to the
integrated control system (ICS). The second backup power
supply supplies power to HSPS console indicating lights and ;
the local alarm panel. This information had been issued on '
December 4, 1986, via Field Change Request (FCR) No.
C051206. ,
.
i
1
_
44
The licensee representative agreed that the backup power
supplies should be tested to verify that they functioned
per design. Testing was satisfactorily completed by the
licensee on Saturday, March 21, 1987.
The results of the above testing will be reviewed as part
of the regular NRC inspection program after licensee ac-
ceptance of the results. The above licensee actions satis-
factorily resolve, in part, unresolved item (289/87-06-07). j
5.2.2.2 HSPS Test Witnessing
On March 19, 1987, the inspector observed a portion of per-
formance of TP 332/3. The inspector found the licensee to
be following the test procedure as written and meeting the
criteria stated in paragraph 5.2.2.1 of NRC Inspection
Report No. 50-289/87-06. The inspector also noted that
continuous quality' assurance (QA) coverage was being pro-
vided.
5.2.3 preoperational Test Summary
-
Overall, test procedures were adequate and properly imple-
mented. In general, the licensee has a knowledgeable SU&T.
staff. However, an increase in communication between the
SU&T department and Technical Functions is needed to ensure
that items such as the backup power supply testing are not
overlooked. Further complication of testing issues was due
to the large SU&T workload experienced during this outage.
5.3 Cycle 6 Startup Physics Testing I
l
5.3.1 Test Results and Reload Safety Evaluation Reviews Prior to
Cycle 6 Startup
In preparation for Cycle 6 startup, the inspector verified
the satisfactory completion of the following documents and
procedures:
--
Three Mile Island Unit 1 (TMI-1) BAW-1977, dated
October 1986, " Cycle 6 Reload Report;"
--
B&W Technical Report, 61-1000733-19, " Physics Test
Manual for TMI-1 Cycle 6;"
--
SP 1303-7.1, Revision 13, dated October 14, 1985,
" Intermediate Range Channel Test," (test completed on
March 20, 1987);
I
i
A
45
l
'
--
SP 1303-7.2, Revision 17, dated October 2, 1985,
" Source Range Channel," (test completed on -March 20,
1987); j
t
--
SP 1302-5.13, Revision 8, dated May 19, 1986, " Control !
Rod Absolute and Relative Position Indication Equip-
ment Surveillance Calibration," (test completed 'on
March 8, 1987); and,
'
/
--
SP 1303-11.1, Revision 14, dated February 27, 1987, {'
" Control Rod Drop Time," (test completed on March 22,
1987).
]
l Findings
i
The reload report performed to support this cycle's opera-
-tion concludes that Cycle 6 can be safely operated at the
rated core power level of 2535 MWt and the core design and
associated- TS changes pose no significant hazards as de-
fined in 10 CFR 50.92. This reload submittal was reviewed $
by NRR and was found acceptable (letter from J. O. Thoma,
NRC, to H. D. Hukill, GPU, dated March 20,'1987)..
l
'
The inspector independently verified that the predicted
values and acceptance criteria- of the procedure for "Zero l
Power Physics Tests," RF-1550-02, were consistent with the i
fuel vendor (B&W) supplied values (Physics Test Manual for
, _
TMI-1 Cycle 6). No discrepancies were noted.
The above-listed surveillance tests .were: satisfactorily
performed prior to criticality. Test results were accept-
.able, i
5.3.2 Startup Physics Testing Witnessing and Data Review
5.3.2.1 Test Witnessing
At .various times 'during the inspection period, the inspec-- :
tor witnessed most portions of Zero Power Physics Tests
l. (ZPPT) in progress. The tests witnessed included:
--
initial criticality;
i
l --
reactimeter checkout;
-
--
all rods out boron concentration measurement;_
l
l
I
j
l
46
--
control rod worth measurement; and,
--
differential boron worth measurement.
Tests were observed for the following areas.
--
Zero power physics tests were conducted in accordance I
with the approved test procedure, RF-1550-02, Revision {
8, "Zero Power Physics Testing" i
l
--
Prior to performing each test, briefings with the test j
crew and operations personnel were conducted and the i
briefing was adequate 1
--
Test prerequisites and initial conditions were met
--
Operator actions were correct
--
Summary analysis was made upon completion of each test.
5.3.2.2 Test Results Review i
l
Physics test results at the hot zero power and subsequent
power escalation test at 30 and 75 percent power plateaus
were reviewed and compared with TS and with acceptance cri-
teria detailed 'in the test procedures. The details and
findings of the review are described in the following.
5.3.2.2.1 Initial Criticality
Initial criticality of Cycle 6 was achieved on March 23,
.1987, with reactor coolant system boron concentration of
1453 ppm and group 7 position at 85 percent withdrawal
(WD). The predicted boron concentration based on the same
group 7 position is 1384 ppm. The measured deviation from
prediction is, therefore, 1453 - 1384 = 69 ppm. This re-
sult 1. within the acceptance criteria of 100 ppm.
5.3.2.2.2 Reactime*ar Checkout
The inspector independently verified that the reactimeter
was adjusted with the correct inputs of delayed neutron
fraction and decay constants and noted that the results of
the " cold" calibration checks were satisfactory.
l
)
47
,
The reactimeter was further checked using Intermediate
Range NI-3 input when the reactor reached criticality.
Comparisons of predicted and measured reactivities, based
on doubling time measurements, were generally acceptable.
However, during one of the react 1 meter checkout tests (add-
ing a positive reactivity), the inspector noted that the
test engineer had difficulty in obtaining accurate doubling
time. This was because the test engineer read the flux
change from the NI-4 console indication. The NI-4 indica-
tion was a logarithmic scale and was difficult to read
quickly (about 40 seconds) and accurately. In another re- I
actimeter checkout test (adding a negative reactivity), the
test engineer utilized the same practice to collect the
test data but used a much longer test duration (about 160
seconds). This test experienced no problems and the meas-
ured reactivity agreed closely with the predicted values.
This subject was discussed with the licensee's lead nuclear I
engineer. He indicated that in the next cycle's ZPPT, the
flux change would be monitored by using the flux strip ;
chart recorder instead of the NI-4 console indication dur- j
ing the reactimeter checkout test. The inspector may re- i
view this in the next cycle startup physics test inspection. l
5.3.2.2.3 All Rods Out ( AR0) Critical Boron Concentration
The inspector reviewed test data and noted the following
results.
Predicted Valua Measured Value
Configuration (ppm) _ (ppm) _
ARD, Hot Zero Power 1394 1 100 1449 ,
!
Test results met acceptance criteria. l
5.3.2.2.4 Isothermal Temperature Coefficient (ITC)
The licensee measured the ITC by performing two heatup and
one cooldown tests. The inspector noted that 'the licen-
see's test result was based solely on the cooldown test
data, as follows:
Predicted Value Measured Value
Configuration (pcm/F) _. (pcm/F) _
ARD, Hot Zero Power 0.698 1 4 0.515-
j
__
48
The corresponding moderator temperature coefficient (MTC) was deter-
mined to be as follows:
Measured Value TS Limit
Configuration (pcm/F) _ (pcm/F)_
ARD, Hot Zero Power 2.165 < 5.0
The inspector independently evaluated the test data and obtained the
following results:
Measured Value - ITC MTC TS' Limit
Test Condition (pcm/F) _ (pcm/F) (pcm/F)_
Heatup (Slope at 532 F) 0.376
Cooldown (Slope at 532 F) 0.515
Average 0.446 2.096 < 5.0
ITC test results were within acceptance criteria and the correspond-
ing MTC value met the TS requirement.
5.3.2.2.5 Control Rod Worth Measurement
The licensee measured the regulating group rod worth in
accordance with the procedure RF 1550-02. The following
results were noted.
Predicted Value Measured Value
Regulating Group (pcm) _
(pcm) _
Group 7 994 15% 963.5
Group 6 847 15% 756.5
Group 5 1477 15% 1530.5
Groups 5 - 7 3318 10% 3250.5
Test results were within acceptance criteria.
The corresponding differential boron worth as derived from this test
was 9.918 pcm/ ppm. This value agreed well with the predicted value
of 9.308 pcm/ ppm and was within the acceptance criteria of 15 per-
cent.
_ _ -
1
49
)
5.3.2.2.6 Core Power Distribution Verification
The detailed core power distribution at the 75 percent power
plateau was measured by the licensee per procedure RF 1550-
08, " Core Power distribution Verification, " Revision 7.
The inspector noted the following results.
--
The measured quadrant power tilt. was 0.51 percent,
which was within the TS 3.5.2.4 limit of 4.12 percent;
--
The measured radial peaking factor for each fuel as-
sembly was consistent with the analytically predicted
value. The comparison of the highest measured radial
peaking factor (1.281) at core location L-13 agreed
closely with the predicted value of 1.260.
--
The measured total peaking factor in each fuel assem-
bly also agreed consistently with the predicted value.
The highest measured total peaking factor (1.514) at
core location L-13 agreed well with the predicted
value of 1.510. j
--
The measured linear heat rate at each axial location
was well within TS 3.5.2.7 limit.
I
All results were acceptable.
5.3.2.2.7 Power Imbalance Detector Correlation Test ,
The relationship between the indicated out-of-core offset
and the indicated incore offset was determined at the 75
. percent power plateau. Reasonably good linear relationship
for all four. power range channels (NI's 5-8) were observed.
The scaled difference amplifier gain factor K was deter-
mined to be 3.684 from the test results. The inspector
verified that this new K factor was entered into the plant
Reactor Protection System by I&C per SP 1303-4.1, Appendix
B, " Procedure for Changing Scaled Difference Amplifier l
Gain" on March 30, 1987. '
5.3.3 Quality Assurance / Quality Control (QA/QC) Interface with !
the Startup Physics Testing program
Throughout the ZPPT period, the inspector observed QA/QC
personnel witnessing test performance and reviewing test
procedures while in progress. The QA/QC provided a 24-hour
coverage during this period.
I
_ - _ _ _ _ - -
.. .. .. . .
. ..
.
. . .. .. . ..
50
During the performance of ZPPT, QA/QC verified that test
conditions, procedure compliance, and acceptance criteria
were being met. No outstanding items in the startup
physics testing area were left open.
Based on inspector observations of QA monitoring and docu-
ment review of QA monitoring reports, the inspector' con-
sidered the monitoring program to be thorough and compre-
hensive in this area.
5.4 Testing Summary
For the items reviewed, preoperational testing was properly completed
and the procedures were overall adequate to perform their intended
functions. The licensee adequately met their commitment to perform
an integrated functional test of the Remote Shutdown Panels despite
initial reluctance when NRC, in early 1986, initiated discussions in
that regard. An integrated functional test of the steam generator
level indicators was not done and it could have been a measure to
avoid plant power escalation delay and operator confusion in the
startup process (see paragraph 4.2.7). Also, despite commitments in
a previous inspection report, the HSPS function test procedure was
not revised until this inspection to adequately test the HSPS cabinet
backup power supplies.
The startup physics testing program, up to 75 percent power plateau
test, was accomplished in accordance with approved test procedures.
All test results were within test acceptance criteria. This indi-
cates that the characteristics of the Cycle 6 core are consistent
with the design predictions. The licensee performed very well in i
this area.
Because of fouling problems on the secondary side of the steam gener-
ators, TMI-1 was being restricted to 83 percent power level during
this inspection period. Physics testing at 100 percent power level
has yet to be completed.
6.0 Primary Coolant System Pressure Isolation (Event V) Valves
6.1 Acceptance Criteria / Scope of Review
An inspection was conducted on March 17, 1987, at TMI-1 concerning
an NRC Temporary Instruction 2515/84, " Verification of Compliance
with Order for Modification of License."
The inspection objective was to ensure .that actions required by a
1981 NRC Order for modification of license concerning primary coolant
system pressure isolation (WASH /1400 - Event V) valves have been
implemented.
-__
i
51
i
l
Reference was made to a past special inspection conducted by Region I
resident inspectors involving Event V concerns (re: NRC Inspection
Report No. 50-289/85-17).
6.2 Review
The April 20, 1981, Order from J. F. Stolz to H. D. Hukill, entitled
" Order for Modification of License Concerning Primary Coolant System l
Pressure Isolation Valves," was reviewed along with the associated
Technical Specification (TS) Section 3.1.6 (Amendment No. 47). Also
reviewed was the leak test procedure, 1300-3T, that implements the i
leak test requirements. The procedure tests each of the four Event V !
valves (CF-V5 A/B and DH-V22 A/B) individually with correct regard to l
test pressure differential and calculated leak rate adjustments based
on the observed pressure differential. In addition, the procedure
provides correct leak rate acceptance criteria. Observations regard-
ing the recording and evaluation of "as-found" leak rate data are
discussed below. The inspector determined that the plant documenta-
tion associated with the Order was acceptable.
The leak rate test records of the six past tests were reviewed to
determine that the procedure was being adequately irplemented since
1981. Table 2 provides a summary of the test results.
TABLE 2 - RESULTS OF LEAK TESTING !
Procedure: Three Mile Island Nuclear Station Unit No. 1 SP 1300-3T,
" Pressure Isolation Test of CF-V4A/B, SA/B, and DH-V22A/B"
Test Revision Test
Date Number Results Comments
4
8/28/81 0 Zero Leakage
8/30/83 4 22A and B Failed 22A seated after flush and 22B
seated after delta P applied.
"As-left" leak rates acceptable
5/22/84 6 Acceptable Leakage
4/12/85 7 22A and B Failed 22A seated after delta P applied
and 22B seated after a number of
seating attempts including flush-
ing. "As- left" leak rates ac-
ceptable
1/28/86 8 5B Failed SB passed after retest
11/1/86 9 Acceptable Leakage
- _ - - _ _ - - - - ,
'
52
I
The table indicates failed "as-found" tests. Engineerir.g staff mem- 1
bers stated that these failed tests were caused by the check valves j
not being fully seated due to lack of adequate pressure differential, (
slight misalignment, or small amounts of debris on the seating sur- i
faces. In all cases of "as-found" failures, reseating of the valve l
or flushing the valve resulted in adequate seating and "as-left"
leakage rates. The inspector determined, based on the records re-
viewed, that all six tests were conducted as required and that the
records adequately documented the tests.
6.3 Findings and Conclusions
The inspector made two observations concerning the recording and
evaluation of "as-found" test failures. These observatic > * were dis-
cussed at the exit meeting.
(1) The supplemental data sheets that provide the leak rates of past
tests are somewhat misleading in that they record only the "as-
left" leak rate. No indication of "as-found" failure is re-
corded on these summary sheets. The detailed test records were
quite good and they appropriately reflected the "as-found" fail- '
ures. The inspector was concerned that these summary sheets may
be misleading if only used as a bases for future decisions with
respect to the detailed test records. The licensee noted the
concern and indicated that the data sheet would be reviewed and
modified to assure that it would not be incorrectly used. l
(2) The procedure calls for " backup test methods" if valves fail the
leak tests. These methods involve reseating the valves by ap-
plying a pressure differential and/or flushing the valves. The
inspector raised the concern that these backup methods allow
multiple attempts at seating the valve and, if used excessively
or without engineering judgement, may allow a degraded valve to
pass the test instead of identifying a defect that should be
corrected.
4
The licensee stated that engineering staff pay close attention to the
"as-found" test failures and usually are directly involved in the
testing. The inspector interviewed the engineering staff members
directly involved in testing and was satisfied that adequate evalua-
tions have been made. The engineering management and staff were
questioned regarding the major objectives of the Order. Their re-
sponses indicated a clear awareness of the importance of assuring
over pressure protection of the low pressure decay heat removal (DHR)
piping.
6.4 Summary
It was concluded by the inspector that the 1981 Order has been pro-
perly implemented at TMI-1.
i
j
53
I
7.0 Radiological Controls
Licensee performance in the radiological controls area was reviewed
against applicable 10 CFR 20 and license criteria. Several recent per-
sonnel contamination events, resulting in uptake of radioactive matericl, j
were reviewed during this inspection. Licensee posting, labeling, and l
post-outage ALARA results were also included in this review. :
i
7.1 Contamination / Uptake Events
The inspector reviewed events associated with three personnel contam-
ination incidents, resulting in uptakes of radioactive material,
occurring on February 24, March 7, and March 12, 1987.
Licensee performance in association with these events was reviewed by
the following methods:
i
--
discussion with health physics, maintenance, and radwaste per-
sonnel;
--
review of the following documentation:
--
Procedure 9100-ADM-4110.04, " Radiation Work Permit;"
--
Radiation Work Permit (RWP) 32751, " Install Convex Mirror,"
and associated radiological surveys;
--
RWP 32830, " Inspect and Place Filter Housing in Makeup
Filter Cubicle," and associated radiological surveys;
--
RWP 32789, "Decon Areas...," and associated radiological
surveys;
--
Radiological Investigative Report (RIR) Nos. 87-0189,
87-0192, and 87-0193;
--
Radiological investigative critique minutes for each inci-
dent;
--
Licensee Memorandum 9100-87-0131, dated March 27, 1987,
" Makeup Filter Chronology";
--
Licensee Memorandum 9100-87-0141, dated April 2, 1987,
" Letdown Pre-Filter Lessons Learned"; and,
--
Licensee Memorandum 9100-87-0146, dated April 3, 1987,
" Awareness of Radiological Conditions".
Within the scope of the above review, one . apparent violation and one
unresolved item were identified. A description of each incident is
given below.
. .
.. .. ._ . _
_. ._ -_ _- _
. - . . . . .
. . . _..
.
. ,
,
. .
. .
. . .
.. . .. .
.. .
!
54
!
!
7.2 February 24, 1987, Event
On February 24, 1987, two plant maintenance workers drilled holes in
the west wall of the letdown pre-filter cubicle to support the in-
stallation of a mirror. A licensee pre-work survey identified con-
tamination levels of < 1000 disintegrations per minute (beta gamma)
per 100 cm2 (dpm/100 cm 2
) at the drilling site on the wall and ap-
proximately 30,000 dpm/100 cm2 beta gamma in the immediate floor
area. Significantly higher levels of contamination were present in i
the approximately 5 ft. x 7 ft, cubicle. The pre-work survey indi- !
cated 440,000 dpm/100 cm 2 (beta gamma) on the adjacent filter l
housing. Contamination in the mrad smearable levels were also pre- !
sent on the cubicle floor due, in part, to a spent filter that was !
dropped in July 1986.
Workers performed the drilling in accordance with RWP No. 32751, i
which did not require respirators. Subsequent to the work operation,
both workers were found to have facial contamination, ranging from
4,000 to 5,000 dpm by direct frisk. BZA air sample results led to j
the assignment of 7.3 MPC hours for each worker.
7.2.1 Licensee Followup
l
Investigational post-work surveys were performed to iden-
tify the cause of the airborne activity. Smears taken on I
the drill bit used showed negligible activity; smears taken ,
in the cubicle indicated contamination levels in the immed- 1
iate floor area ranging from 100,000 to 200,000 dpm/100 cm2
(beta gamma).
A critique 'was held on February 26, 1987, to identify
causes of the contamination. Critique minutes state that
work practices were good and did not contribute to the con-
tamination/ uptake event. Instead, the critique identified
a problem with smear sample analysis methods as a cause of
the occurrence. Contamination in the letdown cubicle was
composed mostly of Co-58, which, due to its low beta abund-
ance, was higher by a factor of 8 - 10. The licensee's
smear counting equipment was not as efficient as needed for
Co-58. Licensee immediate response at this point was to
concentrate on the Co-58 problem by identifying other plant
areas where this might be a concern and by developing new
counting efficiencies.
A Radiological Investigative Report (RIR No. 87-0189),
dated March 26, 1987, was also generated concerning the
above event. This investigation recognized the high levels
of contamination present on the floor of the cubicle as a
potential cause of the airborne activity, indepecdent of
55
the Co-58 characterization problem. This RIR . recognized
that the post-work increase in floor contamination levels
seen in the immediate work. area was probably due to workers
" tracking" the contamination around inside the cubicle.
The RIR recommendations included posting the letdown pre-
filter cubicle "high contamination," " airborne radioactiv_
ity area," and " respiratory protection required." ,
!
7.2.2 Conclusions '
The inspector determined, based upon review of pre work and
past work surveys and licensee followup documentation, that
the licensee's corrective actions for the above event were
incomplete. Identification of the Co-58 characterization
problem was a significant step; however, it was not until
the RIR was generated on March 26, 1987, that it formally
recognized the high potential for airborne activity in the
letdown filter cubicle based on high contamination levels. ;
Consequently, the followup actions recommended in the RIR :
(i.e., posting the cubicle " respiratory protection i
required") were not in place in time to prevent the inci-
dent described below.
7.3 March 7, 1987 Event j
Changeout of spent letdown pre-filters is routinely accomplished from
the roof of the letdown pre-filter cubicles, working through an ap-
proximately 8-inch diameter access hole in the ceiling located above
each filter housing. Workers on the roof remove the roof shield plug
and, using long-handled tools, open the filter housing and pull the
spent filter up into a shielded transfer pig. A temporary ventila-
tion line takes a suction on the cubicle through the unused 8-inch
diameter access hole. Workers are not present inside the cubicle
during this operation due to the high dose rates generated by the
filter.
On March 7, 1987, a mock-up work evolution was conducted to see if
spent pre-filters could be temporarily stored inside the cubicle.
Two radwaste workers entered the cubicle to place a filter " carousel"
holder inside the cubicle. Two maintenance workers, located on the
roof of the cubicle, used long-handled tools and a clean filter,
which was passed in . through the ceiling access holes, to determine
if filters could be placed into the carousel from above. Both the
practice filter and filter carousel were clean prior to placement
into the cubicle. Also, no system breaches or work with an actual
"used" filter was included as part of this operation.
. _ _ - - - _ _ _ - _ - - _ _ _ _ _ _
.. .. .. .
. .. .. ..
, .. . .
i
56
Pre-work surveys indicated floor contamination levels between the "A"
and "B" filter housings (the intended location for the carousel) from l
12 - 32 mrad / hour smearable beta. Highest floor contamination level
in the cubicle was approximately 40 mrad / hour beta. The pre-work
survey did not sample contamination levels inside the ceiling access j
holes through which the clean filter was passed.
'
Controlling RWP 32830 required a wet suit and respiratory protection
for any work in the cubicle other than visual inspection. However,
at the pre-job briefing, it was determined by radiological controls
supervision that the placement of the carousel inside the cubicle did
not constitute airborne generating work and could be performed with-
out respirators. Consequently, none of the involved workers wore
respirators during this operation. The scope of the work was also
discussed at the pre-job briefing; the-radwaste workers indicated the
work scope included placing ;he carousel in one spot in the cubicle
and then standing by to see if the filter could be placed into the
carousel from above.
During the operation the filter carousel was repositioned three or
four times inside the cubicle as the clean filter was lowered in and
out of the cubicle. The radwaste workers indicated this reposition-
ing did not require climbing or worker movement all over the cubicle;
the small size of the cubicle allowed carousel repositioning while
the workers remained - in one area. Both radwaste workers remained
inside the cubicle to observe efforts to place the filter in the
carousel. Workers indicated that temporary ventilation was not used
during this operation, since both ceiling access holes to the cubicle
were being used.
All workers indicated they frisked out clean after the operation. I
BZA air sample results for the two radwaste workers indicated air- '
borne activity levels of 67.6 MPC (10.1 due to beta gamma, 57.5 due
to alpha activity). Subsequent whole body counts of the four workers
indicated intakes of radioactive material for three of the four
workers.
7.3.1 Licensee Follow-up
A post-job critique was held on March 10, 1987. A RIR (No.
87-0192) was also generated concerning the event. The RIR
identified two potential causes of the airborne activity:
--
multiple placement of the filter carousel stirred up
activity from the floor; and,
--
movement of the clean filter through the ceiling ac-
cess holes dislodged contamination from the inner
edges of the holes.
1
1
- _ - _ _ _ _ _
57 i
Subsequent post-work survey of the ceiling access holes
indicated extremely high contamination levels (approxi- {
mately 200 mrad / hour B) inside the access holes. The in- )
spector noted that airborne activity levels of 50.1 MPC l
(mixed beta gamma and alpha activity) were generated by the
technician during the performance of the post work inves- l
tigational survey. i
7.3.2 Conclusions
The 10 CFR 20.201(b) requires that each licensee make such
surveys as may be necessary to comply with all sections of
Part 20 and are reasonable under the circumstances to eval-
uate the extent of radiation hazards. The 10 CFR 20.201(a)
defines " survey" as an evaluation of the radiation hazards
incident to the production, use, release, disposal, or
presence of radioactive materials or other sources of rad- {
iation under a specific set of conditions. W
Contrary to the above, the licensee failed to perform an
adequate survey (evaluation) to accurately evaluate the
extent of radiation hazards involved with the letdown fil-
3
j
ter cubicle operation performed on March 7,1987. Specif- ;
ically, the licensee failed to: (i) anticipate, survey, l
and evaluate the potential for airborne activity generated
by dislodging contamination in the access holes; and, (ii)
adequately assess cubicle previous history and floor con-
tamination levels, which indicate a high potential for air- }
borne activity due simply to worker presence and traffic in '
the cubicle.
This failure to adequately survey (evaluate) radiological
conditions resulted in the unplanned intake of radioactive '
material by three workers and constitutes an apparent vio-
lation of 10 CFR 20.201(b) (289/87-09-12).
Licensee investigation and proposed corrective actions, as
identified in the formal event critique, RIR 87-0192, and 3
several " lessons learned" memoranda appear comprehensive. I
Credit cannot be given to the licensee for adequate self- !
identification of the apparent violation, however, as com- 4
prehensive and timely corrective actions were not completed l
in response to the February 24, 1987, incident. !
I
Additional concerns identified by the licensee associated I
with the March 7, 1987 incident include the following. j
i
1
i
.1 'l
_ _
l
58
--
The original scope of work cs discussed in the pre-
work briefing was expanded upoc (carousel moved three
times instead of placed once in csbicle);
--
The radiological controls technician providing cover-
age for the job was not present at the pre-work brief-
ing. The technician did receive a briefing from the
radiological control supervisor, however, prior to
covering the work.
--
The BZA air sample taken while the radwaste workers
were in the cubicle was not counted for alpha activity
until approximately 4.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> after the workers exited
the cubicle. Beta gamma activity was assessed within
half an hour after worker exit. The alpha component
of the air sample activity was found to predominate;
57.5 of the overall calculated 67.6 MPC of air activ-
ity was found to result from alpha activity.
The inspector noted Licensee Memorandum 9100-87-0141, dated
April 2,1987, addressed the above concerns and was trans-
mitted to all Unit 1 Group Radiological Control Supervisors.
l
The inspector also noted the high levels of airborne activ- !
ity apparently generated solely by technician survey in the i
cubicle on March 7,1987, (approximately 50 MPC). The in- !
spector verified the technician wore respiratory protection
for this entry. The inspector also questioned the licensee
as to whether technicians routinely wore respirators while
surveying the cubicle or if the potential had existed for
unrecognized intakes of radioactive material during pre-
vious surveys of the cubicle.
The licensee indicated that worker entries into the cubicle
'
were rarely made- and that contamination surveys inside the
cubicle were consequently rarely performed.
The inspector reviewed letdown filter cubicle surveys from
July 1986 to the present and noted only one instance prior
to the entries discussed in this report where a survey
requiring full entry into the cubicle was performed (Survey
No. 08587J0208, performed January 8,1987). All other sur-
veys appeared to have been conducted from the roof or door-
way of the cubicle using- a radiation detector with exten-
sion capabilities. The technician performing the January
8, 1987, survey inside the cubicle was not available for
interview during the current inspection and, consequently,
,
--___
. .
59
could not be questioned concerning the use of respiratory
protection. The licensee indicated that a thorough review
would be performed to determine if (1) additional entries /
surveys in the cubicle had taken place; and, (ii) whether
retroactive MPC/ hour assignments need to be made. The re-
sults of this investigation will be reviewed during a sub- I
sequent inspection in conjunction with NRC staff follow-up j
to the above-noted apparent violation (289/87-09-12). i
7.4 March 12, 1987 Incidert
On March 12, 1987, two radwaste workers were contaminated while pre-
paring a newly decontaminated high integrity container (HIC) to re-
ceive spent filters. The workers were signed in on standing RWP No.
32789, "Decon Areas / Equipment .., " which did not require respiratory-
protection. Subsequent whole body counting indicated the workers had
received intakes of radioactive material. No air sample was taken in
the work area during the operation.
Preparation activities included taping a "herculite" cover to the HIC
and connecting a highly contaminated splash ring to the HIC. The
splash ring was contained in a taped-up yellow bag but was not
labeled as radioactive material. Licensee Procedure 3000-IMP-4400.01
indicates a radioactive material label is not required for material
within a controlled area. The splash ring was contained in a posted
contaminated area during this evolution.
Licensee investigation determined the personnel contamination and
uptake resulted when the workers removed the plastic bag containing
the splash ring and when one of the workers reached through the
splash ring to cut holes in the "herculite" cover. During NRC inter- ,
view of one of the workers and the radwaste supervisor, both indi- I
viduals indicated they knew the splash ring was contaminated.
Licensee RWP Procedure 9100-ADM-4110.04 states that a standing RWP
shall not be used when contaminated or potentially contaminated sys-
tems are opened or when the task is expected to cause significant
change in the static radiological conditions in the area. Licensee
RIR No. 87-0193, dated March 12, 1987, identifies as one cause of the i
above incident the failure to generate a separate specific RWP to
cover opening of the yellow bag containing the splash ring.
The above failure to generate a specific RWP for work involving the
highly contaminated splash ring constitutes an apparent violation of
the RWP procedure. However, in an effort to encourage licensee self-
identification and correction of problems, NRC Enforcement Policy (10
CFR 2, Appendix C) allows the mitigation of violations if they were
1
1
60 l
i
identified by the licensee and certain criteria were met. This ap-
parent violation, therefore, will not be cited at this time but will-
remain unresolved pending review, during a subsequent inspection, of
the implementation of their proposed corrective actions (289/
87-09-13).
7.5 Licensee Overall Review of Filter Operations
Subsequent to the occurrence of the three incidents described above,
the licensee's radiological engineering group produced memorandum
9100-87-0131, titled "MU Filter Chronology," dated March 27, 1987.
This memorandum lists a chronology of problems associated with the
changeout and disposal of make-up filters, which have occurred from
October 1985 to the present.
The inspector reviewed the above memorandum and noted it provides an
effective overall review of problems associated with this operation
and highlights the need for effective corrective action. The memor-
andum also summarizes the results of a licensee meeting held on
February 26, 1987, concerning resolution of this problem area. Several
long-range action items are identified, including the following:
--
evaluate letdown lines for potential shielding above the make-up i
filter cubicle; and, i
--
evaluate plant ventilation system around the make-up filter
cubicle.
The licensee also indicated that an additicr,al management meeting to
discuss the same topic was scheduled for the near future. Due to the
significant number of radiological problems which have occurred in
association with the make-up filter changeout and disposal operation,
the licensee is requested to include in their response to the appar-
ent violation described in paragraph 7.5 of this report a description
of their planned long-term corrective actions planned for this area
(289/87-09-12).
7.6 Whole Body Count Results
The inspector reviewed whole body count data and preliminary airborne
activity exposure estimates for those individuals involved in the
incidents described above. Although the licensee's review and
analysis of this data was not complete, the inspector was able to
determine the following:
--
all airborne activity exposures (MPC/ hours) appear to be well
below regulatory and licensee administrative limits; and,
61 ,
1
--
appropriate steps are being taken to determine MPC/ hours expos- .
ure from whole body count data for those individuals for whom no !
air sample exists. Estimates of alpha activity, based on con-
tamination beta gamma / alpha ratios, are being calculated and
credited towards the workers exposure.
Licensee finalized exposure estimates will be reviewed during a sub-
sequent inspection as part of the followup of :he above-noted viola-
tions (289/87-09-12).
7.7 Posting and Labeling
3
The licensee's program for posting and labeling radiological areas
and radioactive material was reviewed against criteria contained in
10 CFR 20.203 and license TS. Licensee performance in this area was ,
evaluated by the following methods: i
--
discussion with radiological controls personnel; and,
--
tour of the auxiliary building controlled areas.
Technical Specifications, Section 6.12, require that each high radia-
tion area (HRA) be barricaded and conspicuously posted. On April 20, y
1987, the inspector noted the door to the waste evaporator cubicle in
the contaminated waste processing (CWP) area, lower level, auxiliary
building, was open with no barricade in place to prevent entry. This
cubicle was posted as a HRA.
The inspector questioned an operator who was working inside the cub-
icle and an area radiological controls technician who both indicated i
that the door to the cubicle was often left open, particularly when 1
personnel were working in the room. Both individuals indicated no
temporary barrier was utilized during those occasions when the door >
was left open. I
I
An NRC independent survey inside the cubicle indicated a HRA did not !
exist at the time the door was discovered open; maximum area dose i
rates ranged from 35 - 40 mR/ hour. Consequently, the cubicle was
conservatively posted and no violation of TS requirements occurred. {
'
The inspector noted, however, that the apparent routine practice of
leaving the cubicle door open may indicate a lack of awareness of TS j
HRA requirements. Previous examples of HRA posting concerns have
j
been noted in NRC reviews in this area (tee NRC Inspection Report j
Nos. 289/85-30 and 86-05). t
I
l
l
1
\
l
62
Upon identification of the above concern, the licensee modified the
area posting such that a rupe barrier delineated access to the HRA.
The licensee indicated that aggressive efforts have been recently
made to increase workers sensitivity to proper HRA posting, particu- ,
larly in light of NRC concerns. These have included placing alarming '
rope barrier connectors at certain HRA access points. These connec- ;
tors alarm if the barrier is not replaced after an individual passes
through.
The licensee also indicated that, in response to the above concern,
additional training concerning HRA posting and control would be dir-
ected towards the radiological controls, auxiliary operators, and
mechanical maintenance groups. The effectiveness of licensee efforts
in this area is unresolved pending NRC staff review in a subsequent
inspection (289/87-09-13).
7.8 Post-Outage ALARA Results
Licensee preparations and performance in the radiological controls
area for the 6R Outage was reported in previous NRC inspections
(50-289/86-19 and 86-21). The inspector reviewed overall ALARA per-
formance and " lessons learned" for the recently completed outage by
the following methods:
--
discussion with cognizant ALARA engineers;
--
review of the following material:
--
Licensee Memorandum 9100-87-0167, dated April 22, 1987,
" Radiological Controls for 6R Outage;" and,
--
cyclic training, " Outage Recap Outline." j
Overall job-specific outage exposure was lower than pre-outage esti-
mate. Overall outage exposure was approximately 30 percent lower
than estimated (239 versus 350 person-rem). The licensee indicated
that overall dose rates were lower than anticipated based, in part, l
on the following source-term reduction techniques utilized by the '
licensee:
--
primary system clean up;
--
extensive use of temporary shielding; and,
--
scheduling of reactor building integrated leak rate test (ILRT) I
at the start of the outage to' allow for the decay of short-lived
isotopes prior to working in the reactor building,
l
1
l
l
!
4
63
Numerous " lessons learned" were identified in the outage report,
indicating that the licensee critically evaluated their performance
despite overall success in bringing outage exposure in under esti-
mate. Significant " lessons learned" included:
--
significant exposure was accrued due to the use of a 0.540-inch
diameter probe, which became stuck in the tubes during OTSG eddy
current testing; and,
--
problems with the fuel upender were identified as significantly
contributing to the high levels of contamination discovered in
the deep end of the fuel transfer canal.
The licensee is also effectively utilizing the cyclic technician
training program to ensure all radiological controls technicians are
briefed concerning the outage and all significant " lessons learned"
are captured. The radiological engineer who had overall responsi-
bility for outage coordination is currently giving a 1.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> talk
to each class of technicians concerning outage performance and en-
countered problems. Technicians are encouraged during the cycle
training to formally identify specific problems they encountered to
the radiological engineering staff or Field Operations supervision.
Future licensee response to " lessons learned" during the 6R outage -
will be reviewed as part of the normal inspection program in the
ALARA area.
7.9 Radiation Protection Summary
' In general, the radiation protection program appears to be properly
implemented, except as noted above. Since the main focus of this
inspection was on event follow up, the numb'er of negative findings
was somewhat expected. A noteworthy characteristic identified by the
inspector was the quality of the licenste's self review process in
this area. The licensee needs to pursue the letdown prefilter prob-
lems. Overall, based on this review, the radiation protection pro-
gram is considered to be generally sound.
8.0 Regulatory-Required Reports
8.1 Introduction
The licensee issued a number of regulatory-required reports during
this period, as described below. A preliminary in office review of
these reports (unless detailed below) indicated proper reporting in
_
accordance with applicable technical specification requirements.
Unless reviewed within this per.iod or to be reviewed by the resident
inspectors, the adequacy of these reports will be unresolved pending
an in plant review by Region I specialists as indicated below.
-_ _. .. .-_
64
8.2 Licensee Event Reports (LER's)
The inspector reviewed the LER's listed below, which were submitted
to the NRC Region I office pursuant to 10 CFR 50.73. Based on resi-
dent office review of the LER's, the inspector determined that cor-
rective action discussed in the licensee's report was appropriate and
that there were no generic issues. In addition, the inspector deter-
mined that the event is not appropriate either for classification as
an Abnormal Occurrence or for Licensing Board, Appeal Board, or
Commission notification.
--
LER 86-012, dated October 6, 1986, on September 4, 1986 "Inop-
erable Fire Detector" - This LER was submitted because a fire
detector in a ventilation exhaust duct had been isolated and i
rendered inoperable. This was a violation of the technical
specifications. The duct had been blocked off for ventilation
balancing. The plastic material was removed and engineering
personnel conducting this testing were . briefed on the require-
ments for evaluating and controlling temporary bypasses or {
Jumpers.
l
--
LER 86-013, dated December 8,1986, on November 7,1986, "Reac-
tor Building Purge Exhaust and Penetration Pressurization Valve
Leakage" - During the containment integrated leak rate test
(CILRT), which was completed on November 9,1986, leakage was i
detected from both a main purge valve and the associated pene- l
tration pressurization (PP) system check valves. This leakage i
caused the failure of the "as-found" CILRT. The leakage from
the PP check valves was isolated and the valve seat on the main ~ <
purge valve was adjusted to reduce the leakage. The "as-left" )
CILRT was completed successfully. (This event was discussed.in '
NRC Inspection Report No. 50-289/86-21.)
--
LER 87-001, dated February 2,1987, on January 9,1987, " Diesel
Generator Automatic Start Due to Defective Work Instructions" -
This LER was written due to an automatic start of the "1B"
emergency diesel generator (EDG), resulting from work in the
"1E" 4160 volt a.c. emergency bus. Inadequate work instructions
caused a worker to actuate the bus overcurrent relay and de-
energize the bus. The licensee reviewed the work procedures for
remaining work on these relays. All remaining work was com-
pleted with the bus work de-energized. This event was reviewed
in Inspection Report No. 50-289/86-22.
65
In some instances, the information provided in the LER's did not
specifically satisfy the requirements of 10 CFR 50.73 or comply with
guidance for LER preparation provided by NRC Office of the Analysis
and Evaluation of Operational Data (AE00) report on TMI-1 LER's, l
dated November 20, 1986. Both LER's86-012 and 86-013 had titles i
that did not provide either root cause or result and a link between i
the two. Also, LER 86-012 did not provide a summary of the correc-
tive action in the abstract. LER 87-001 was adequate in all of these l
categories. l
!
These improvements were discussed with licensee personnel and the
licensee was given a copy of the AEOD LER SALP (Systematic Assessment
of Licensee Performance) input for TMI-1 LER's. This report con- ;
tained detailed evaluations of ten LER's from TMI-1 and it should be '
of assistance in the improvement of LER's. Generally, LER's at TMI-1
are above average in content and analysis. Commencing with 1987
LER's, the NRC staff will evaluate LER's with the guidance given in
the AEOD report. LER's that do not meet those criteria will require !
a revision to correct the discrepancies. This area is unresolved
(289/87-09-15), 1
i
i
8.3 Containment Integrated Leak Rate Test Results Evaluation
l
In accordance with the reporting requirements specified in paragraph
V.B.3, Appendix J, 10 CFR 50, the licensee summarized leakage test i
results from the November 1986 Type A test and Type B and C tests j
since the last containment integrated leak rate test (CILRT). The l
results are presented in a summary technical report, entitled "Reac- ;
tor Containment Building Integrated Leak Rate Test," cover letter,
dated February 9, 1987.
This evaluation of Type A test results is based on the summary tech-
nical report submitted to the NRC by the licensee. The report con-
tains plant general and technical data, discussion of Type A test
prerequisites and procedures, presentation of test results, and Type
B and C leak rate histories. Pertinent Type A test parameters and
results are presented below. Both mass point and total time calcula-
tional methods were employed for the November 1986' CILRT. Note that
the mass point calculational method of ANSI /ANS 56.8-1981 is not
endorsed by 10 CFR 50, Appendix J. These results are presented here
for information. The total time calculational method ANSI N45.4-1972
is consistent with Appendix J requirements and is therefore the
method of record for the test.
I
66
The purpose of the test was to demonstrate that leakage through the
reactor containment building and systems penetrating the containment j
building do not exceed that allowed by plant technical specifications. i
The test was conducted with containment isolation valves (CIV's) and
containment pressure boundaries (CPB's) in an "as-found" condition.
The "as-found" test was declared a failure by the licensee due to
excessive local leakage through reactor containment building purge '
exhaust valves ( AH-V-1A/18). The test was reinitialized after isola-
tion of the leaks and conducted in an "as-left" condition. The test i
was witnessed by two NRC regional inspectors as a routine safety I
inspection. Inspection findings are documented in NRC Region I I
Inspection Report No. 50-289/86-21.
l
1
!
._ _ _ _ _ _ _ _
67
TABLE 3
CONTAINMENT INTEGRATED LEAKRATE TEST RESULTS SUMMARY ;
A. Type "A" Test Parameters and Acceptance Criteria
1. Te s t Me thod. . . . . . . . . . . . . . . . . . . . . . . . Ab s ol ute
i
2. Calculational Method............... Total time (per ANSI
N45.4-1972) Mass Point (per ;
ANSI /ANS 56.8-1981) )
l
3. Test Duration I
I
Stabilization Period. . . . . . . . . . . . . . 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> '
Data Gathering for Leakage. . . . . . . . 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />
Calculation
Verification Leak Rate Test....... 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />
4. Test Pressure................ . . . . 50.6 psig (full !
pressure test)
5. Maximum Allowable................. 0.075 Wt.%/ day or
Leak Rate (At the .75 La (la = 0.10
Upper Bound of 95% Wt.%/ day) ,
Confidence Limit) I
i
B. Test Results
Type A Test Results, Wt.%/ Day
"As-Found" "As-Left"
Acceptance criteria, 0.75 L 0.075 0.075
a
(Maximum Allowable Leak Rate) l
Measured Leak Rate, L
am
Mass Point Approx. 0.10 0.03169
Total Time > 0.10 0.04231
Leak Rate at the Upper Bound
of the 95% Confidence Interval
.-
Mass Point > 0.10 0.03428
Total Time > 0.10 0.06682
Conclusion Unacceptable Acceptable
t
,i
! 68
I
C. Verification Leak Rate Test (VLRT)
1. Test Method.... .................. Superimposed Leak
l 2. Calculational Method............... Total Time
3. Test Duration..... ... ........... 6 Hours
4. Superimposed Leakage (Lg ). . . . . . . . . . 0.09174 Wt.%/ Day (5.58
SCFM)
5. The VLRT acceptance criteria require that the VLRT result
not deviate from the total leakage (superimposed, L,
g plus
measured leakage, L,,) by more than 25% of the maximum
allowable leak rate (.25 La ). Therefore, the result must
be greater than the lower limit and less than the upper
limit, as shown below:
Uppe r Li mi t . . . . . . . . . . . . . . . . . . . ... 0.15905 Wt.%/ Day
(L g +L am + 0.25 La )
Lower Limit. . . . ................. 0.10905 Wt.%/ Day k
(Lg + L,, - 0.25 La )
!
VLRT Result........................ 0.14367 Wt.%/ Day
The inspector concluded that, based on a review of the total time
results, the containment has passed its acceptance criteria for the
"as-left" condition. Failure in the "as-found" condition has been
acknowledged by the licensee. The inspectors informed the licensee
of their requirements with respect to reportability and test fre-
l
quency.
1
1
1
__ - ._ _
69
8.4 Other Reports
The following additional reports were recently issued by the licen- )
see:
(1) Licensee letter (Serial No. 4410-87-L-0030), dated February 27,
1987, 20 CFR 20.407, " Personnel Exposure Reports - 1986."
(2) Licensee letter (Serial No. 5211-87-2038), dated March 2,1987,
"1986 Annual Report," covering various TS-required annual re-
ports.
(3) Licensee letter (Serial No. 5211-87-2041), dated March 2,1987,
" Semi-Annual Ef fluent and Release Report."
The adequacy of these reports is unresolved pending further Region I
specialist review (289/87-09-16).
8.5 Report Summary
The above-noted reports were made in a timely manner. Certain re- i
ports were reviewed for adequacy with results noted above. Other l
reports will require further review by NRC Region I as time permits.
9.0 Previous Inspection Findings / Issues in the Fire Protection Area
9.1 (Closed) Unresolved Item (289/85-25-04). Adequacy of Hydrogen /
0xygen Storage in the Emergency Feedwater Area
The licensee is storing oxygen and hydrogen gas cylinders in the ]
intermediate building at the 295/322 foot elevations for the Contain- I
ment Hydrogen Monitor System. The NRC questioned this practice and
raised the concern that the fire loading for these particular areas
,
could be increased unacceptably.
The licensee performed a fire hazard analysis and a 50.59-type review
which determined that the cylinder storage in fire areas is accept-
able and in accordance with the applicable-Occupational Safety and
Health Administration (OSHA) and National Fire Protection Association
(NFPA) guidance. The inspector reviewed this analysis, toured. the
area, and concurred with the licensee. ;
This item is resolved.
1
!
70
9.2 (Closed) Violation (289/86-01-07). Fire Fighters Respond to Drills
Without Respiratory Protection
The NRC identified as a violation the practice of fire brigade mem-
bers responding to drills without respiratory protection.
The licensee disagreed with the violation and requested a meeting to
resolve this issue. Following the meeting of August 12, 1986, the
licensee committed to require that at least. two members of the bri-
gade utilize protective clothing and respiratory protection during
the drill .
The inspector verified that the licensee included this requirement in
the training procedures and fire protection program. The inspector
also observed a fire drill to assess the brigades effectiveness.
Several fire fighters responded to the drill using the proper pro-
tective equipment. The licensee's responses to this violation, dated
May 2, 1986, and September 17, 1986, were acceptable. This item is
resolved and the violation stands.
9.3 (Closed) Unresolved Item (289/86-01-08). Inadequate Fire Brigade
Training Records
The inspector reviewed the training records of the fire fighters to
verify that they participate in the required number of drills and
meetings. From this review, the inspector determined that fire
fighters participate in two or more drills per year. The inspector
also noted that even though the above requirements are satisfied, it
is possible that two drills could occur within days or weeks of
another and he suggested that that practice be discontinued to assure
more uniform refresher training over the year. Licensee representa-
tives acknowledged the inspectors' comments. This item is resolved.
9.4 (Closed) Unresolved Item (289/86-01-09). Inoperable Fi re Doors.
During a previous inspection, the inspector observed that some fire
doors were not closing automatically. This was attributed to ai r
balance problems and improperly adjusted automatic closers.
The licensee adjusted the closers, thus making the fire doors fully
functional. The inspector toured the facility and randomly checked !
select fire doors for operability. No unacceptable conditions were ,
identified.
This item is resolved.
1
J
-
,
71
9.5 { Closed) Unresolved Item (289/86-01-10). Fire Drill and Page System
Inoperability
During a fire drill, a membar of the brigade did not respond because
he was working in an area. ere the paging system did not work. The
NRC questioned the licensee on the ability to communicate during a
fire emergency. The licensee explained that during an actual emerg-
ency the "Whelan" system (siren) is activated prior to announcing the
emergency. During a drill, the Whelan is not activated. If an oper-
ator hears the siren and does not hear the announcement, then, by
training, he calls the control room for information. He can commun-
icate with the control room in a number of ways.
The communication capability consists of a "Gaytronics" page phones
(the gray phones), the red phones for emergencies, hand-held radios,
and the regular telephones throughout the plant, In addition, the
paging system is under the preventive maintenance program so that
when a page phone is identified as malfunctioning, a work order is
issued to fix the phones. The inspector randomly checked on the
operability of the page phones and did not identify any malfunctions.
This item is resolved.
9.6 (Closed) Unresolved Item (289/86-01-11). Fire Pump Used for Utility
Purposes
The NRC raised the concern that running the fire pumps for utility
purposes may cause problems with the pumps and also deny the opera-
tors the secondary indication of a fire associated with a pump start.
The licensee acknowledged that the maintenance of the electric-driv o
pump has increased, but the overall ability to provide fire service
water to the system is not reduced since the licensee maintains three
additional fire pumps. Because the licensee's technical specifica-
tions allow plant operation with only two of four pumps operable,
there is sufficient system redundancy even if one fire pump is out of
service because of other uses.
With regard to the secondary indication prior to using the fire pump,
the control rooms and auxiliary operators for both Units 1 and 2 must
communicate with each other to identify when an electric-driven pump
is to be used for utility purposes.
This item is resolved.
a
l
72
l
l
9.7 {0. pen) Unresolved Item (289/86-03-08). Overcurrent Breaker
Coordination Study for Non-Safety Loads.
The NRC raised the concern that the licensee's associated circuit
overcurrent protection coordination analysis may not have been ade-
quate.
Following the NRC evaluation of the associated circuits study, the ,
licensee committed to perform additional studies to assure that fire I
induced failures of associated common enclosure circuits will not i
prevent safe shutdown. By letter, dated March 20, 1987, the licensee
informed the NRC that these studies have been performed and did not
identify any noncompliances. NRC/NRR will review these studies to
verify compliance. !
This item remains unresolved pending completion of that review.
9.8 (Closed) Unresolved Item (289/86-05-03). Adequacy of Drills When a
large Number of Fire Fighters Participate and the Maintenance of
Records for the Type of Drill Performed
The inspector questioned the adequacy of fire drills when a large
number of fire fighters participated. Subsequent to a management
meeting that was held at the NRC Region I offices on August 12, 1986,
the licensee revised Procedure No. 6210-ADM.2620.03, which is the !
fire protection training program procedure, to include the require-
ment that not more than ten persons participate in the drill unless
the drill scenario requirec a large participation.
With regard to the type of drill record keeping, the licensee main-
tains the necessary records in the nuclear generation training system
computerized file, t/aich allows easy retrieval.
This item is resolved.
9.9 (Closed) . Unresolved Item (289/86-05-04). Fire Drills Performed in
dverse Weather
The NRC raised the concern that although the licensee is committed
to perform fire drills in adverse weather, none were performed.
The licensee resolved this concern by stating that drills are sched-
uled in advance and are not cancelled because of. bad weather. .Addi-
tionally, the licensee now maintains a matrix that includes the wea-
ther or climatic conditions for reference purposes. The inspector
determined that this is acceptable, j
This item is resolved. !
l
!
!
!
I
73
9.10 .(Closed) Unresolved Item (289/86-14-04). Integrated Test for the
Remote Shutdown Panels (RSP's)
By letter, dated February 19, 1987, the licensee outlined their plan
to perform a limited integrated test for the Remote Shutdown Panels
(RSP's) to meet their commitment as documented in Inspection Report
No. 50-289/86-14, paragraph 3.2.5. The letter also outlined the ob-
jective and a summary description of the test. The licensee prepared
a test procedure TP 683.1, Revision 0, " Remote Shutdown Outside Con-
trol Room," and performed this test on March 22, 1987. The details
are documented in paragraph 5.1 of this report. This meets the
licensee's commitment.
This item is resolved.
9.11 (Closed) Unresolved Item (289/86-23-01). Revision to Emergency
Procedure 1202-37
During the 10 CFR 50 Appendix R inspection in December 1986, the
inspection team had identified a need to revise the draft Emergency
Procedure 1202-37, Revision 26, "Cooldown From Outside the Contral
Room." This was because of:
--
the draft nature of the procedure; and,
--
a need to streamline the procedure, especially in the area of
minimizing the number of keys required to operate various equip-
ment.
These two concerns were satisfactorily addressed in the revised and
approved version of the procedure (Revision 29). Further, the unre-
solved item included the licensee's commitment to perform an inte-
grated test for the RSP's. This commitment is satisfied as discussed
in this section (Unresolved Item 289/86-14-04).
This item is resolved and closed.
9.12 (Closed) Violatinn (289/86-23-02). Emcrgency Lightino in the Control
Room Not Installed in Accordance with the 10 CFR 50.48 Schedule
The licensee did not provide emergency lights with an eight-hour
battery power supply in the control room on a time schedule commen-
surate with 10 CFR 50 Appendix R. The reason for not providing these
lights was that the licensee considered the control room lighting to
be as reliable as battery packs since the control room lighting is
powered by two power trains from emergency diesel generators.
l
-
f
74 1
I
i
A review of this system, however, determined that, in two areas.of
the plant, the power feeds to the control room lighting pass through
the same areas. A fire in these areas may cause loss of lighting in
the control room.
The licensee addressed this concern by rerouting the "B" train out-
side of the potentially af fected area. The inspector reviewed the
modification packages and did not identify any unacceptable condi-
tions. The licensee's actions as described in their response letter,
dated March 13, 1987, were acceptable. j
l
This item is resolved. J
9.13 (Closed) Unresolved Item (289/86-23-03). Lack of Emergency Lighting ,
in Areas Where Safe Shutdown Actions are Required {
During an inspection of the licensee's compliance with the require-
ments of 10 CFR 50.48 and 10 CFR 50 Appendix R requirements, the
concern was raised that some plant areas were not provided with
emergency lights as required by Appendix R,Section III.J.
The licensee stated that emergency lights were installed only in
areas where manual actions must take place and these actions require
NRC approval.
The inspector verified -that upon receipt of the approval from NRC
staff, the licensee proceeded to install the required emergency
lights in the affected areas noted during the previous . inspection.
This item is resolved.
9.14 (Closed) Unresolved Item (289/86-23-04). Oil Collection System for
the Reactor Coolant Pump
During the inspection to ascertain licensee's compliance with the
requirements of Appendix R to 10 CFR 50, the inspector raised the
concern that some oil piping used for instrumentation purposes on the
reactor coolant pumps (RCP's) may not be seismically supported and a
collection system for this piping not provided.
,
]
TM licensee performed a seismic analysis of the piping involved and .
al'o modified the oil collection system'under the piping in question. 1
The inspector reviewed the seismic analysis and reviewed the modifi-
cazion packages to verify that the collection system satisfies the ,
Appendix R requirement.
No unacceptable conditions were identified. This item is resolved.
i
i
75
9.15 Fire Protection Exemptions
During review of the licensee's program to adhere to the requirements
of 10 CFR 50 Appendix R, " Fire Protection," prior to Cycle 6 startup
(see NRC Inspection Report Nos. 50-289/86-23 and 87-05), the inspec-
tor identified the following items needing further action:
--
The Office of Nuclear Reactor Regulation (NRR) requested from
the licensee a letter on Appendix R associated circuits compli-
ance prior to startup; t.nd,
--
various fire protection procedures did not include the follow-
ing: actions to be take, for a fire in the Emergency Safeguards
Actuation System (ESAS) room (CB-FA-3c) as specified in the
revised response to Generic Letter (GEL) 81-12, dated February
10, 1987
--
Emergency procedures to connect an external air or nitrogen sup-
ply for WDL-V-1 or WOL-V-3, 4, or 5 for fire compensation; plant -
cooldown rates; and, Enclosure 3 of EP 1202-37 for substantial
loss of control building ventilation needed to be issued. Also, I
new repair procedure 1420-Y-30, " Repair of Appendix R Cold Shut- I
down Circuits," was to be issued prior to startup.
--
Exemption Request (5211-87-2021), dated February 2,1987, needed
NRR approval prior to startup. Exemption Request (5211,87- {
2033), dated February II, 1987, needed to be withdrawn and a '
letter sent to NRR detailing the specifics of the twenty- minute
roving fire watch program prior to startup.
5
--
Formal operator training on the actions required for a fire in
the ESAS room.
,
The inspector reviewed the following documents concerning the above
items for accuracy and completeness:
--
Emergency Procedure (EP) 1202-31, Revision 27, dated March 12,
1987, " Fire;" j
--
EP 1202-37, Revision 29, dated March 20,1987, "Cooldown From
Outside the Control Room;"
--
Corrective Maintenance Procedure 1420-Y-30, Revision 0, dated
March 20,1987, " Repair of Appendix R Cold Shutdown and Remote
Shutdown System Circuits;"
--
Operating Procedure (OP) 1105-20, Revision 0, dated January 21,
1987, " Remote Shutdown Systems;"
,
.
- _ _ _ _ _ _
l
!
76
--
GPUN Letter No. 5211-87-2070 to the NRC, dated March 20, 1987,
concerning the results of studies completed for Appendix R- t
Associated Circuits Adherence; l
--
GPUN Letter No. 5211-87-2047 to the NRC, dated February 28, 1987, 1
concerning the withdrawal of the ventilation exemption request, 1
dated February 11, 1987, and detailing the 20 minute roving fire l
watch for those areas of the plant in which a fire could affect i
safe shutdown required ventilation; and, j
--
NRC letter, dated March 19, 1987, to GPUN concerning NRR ap-
proval of exemptions requested to the technical requirement of
Section III.G.2 of Appendix R.
Based on the above document review and on discussions with cognizant
licensee personnel, the various open issues identified by the inspec-
tor in IR 50-289/87-05 have been resolved. Additional inspector fol-
,
l
low-up on NRC letter, dated March 19, 1987, occurred as noted below. !
9.15.1 Exemption - Procedural Compensatory Measures I
The NRR exemption approval letter identified several items
that the inspector considered were in need of verification
of licensee actions. The inspector reviewed various docu-
ments to verify that the following items are satisfactorily
resolved.
l
(1) The licensee committed in their letter, dated February j
10, 1987, to revise the post-fire safe. shutdown pro- !
cedures to include the immediate dispatching of an l
operator to the remote shutdown panels (RSP's) if a
fire occurs in a fire area / zone requiring manual ac-
tions within thirty minutes. During review of EP
1202-31, " Fire," the inspector confirmed that for each
fire area / zone requiring manual actions within thirty
minutes, an operator is immediately dispatched to the
RSP's. The operator is to provide mitigating action
under direction from the control room.
(2) The licensee committed to prepare post-fire shutdown
procedures in conformance with tt.e guidance provided
in Generic Letters 81-12 and 86-10. In review of the
fire protection procedures, the inspector verified
that all procedures appeared to be in conformance with
the aforementioned generic letters.
I
- - - _ _ _ - - _ . _ - - _ - _ _ - . . - - - . _ - - . - - 'A
77 !
(3) The licensee committed to utilize portable ventilation
equipment to compensate for fire damaged heating, ven-
tilation, and air conditioning (HVAC) for a fire in -
the screenhouse. EP 1202-31, " Fire," states that if a '
i
loss of ventilation occurs in the screenhouse and con-
trol room and if manual actions cannot restore venti-
lation, then provide portable ventilation as provided
in Appendix D. Appendix D to EP 1202-31 has screen-
house layout figures depicting the placing of portable
ventilation, equipment required, and equipment storage
locations. The procedure for loss of screenhouse
.
j
ventilation due to a fire appears adequate to ensure i
that screenhouse safe shutdown equipment remains oper-
able on loss of ventilation. ,
9.15.2 Exemption - Twenty-Minute Roving Fire Watch
In GPUN Letter No. 5211-87-2047, dated February 28, 1987, i
to NRC concerning their revised ventilation exemptions, the l
licensee committed to provide a twenty-minute roving fire l
watch for those areas of the plant containing cables and l
components where damage could result in a loss of HVAC, in I
place of protection required by Section III.G of Appendix R I
(except the screenhouse as discussed above). NRR approved
the twenty-minute roving fire watch for those areas affec-
ted in its letter, dated March 19, 1987.
The inspector reviewed the above revised ventilation exemp- I
tion request and GPUN memorandum (No. 3210-87-0083), dated I
February 27, 1987, concerning the HVAC roving fire watch
for completeness and accuracy. Both documents include i
plant layout diagrams detailing the roving watch path and '
HVAC components / cables to be observed in the "1B" diesel,
intermediate, and control buildings. The licensee has
installed two cameras with remote screens to observe two
cables in the auxiliary building (fire zone AB-FZ-7 Nuclear .
Services Water Area). The licensee installed the cameras l
to prevent the roving fire watch from having to enter a '
radiation area every twenty minutes to observe these
cables. I
Security personnel are performing the twenty-minute fire
watch. One security individual performs the watch for one
hour during an eight-hour shift. The inspector has verif-
ied that security personnel have had the licensee's fire
watch training and that there are enough security personnel
per shift to complete normal security duties along with the
dedicated roving fire watch. The inspector held several
l
)
_ _ _ _ _ _ .
78
discussions with security and fire protection concerning
the adequacy of training, fire watch manning, general fire
protection practices, and fire protection knowledge of the
individuals involved (including. roving watch standers).
No adverse conditions were identified.
Also, the inspector independently walked down the path
taken by the roving watch and verified that one tour can be
completed in less than twenty minutes. The inspector had
no further comments in this area.
9.15.3 Exemption - RCP Seal Cooling Area
On March 10, 1987, GPUN letter No. 5211-87-2059 informed
the NRC that if Reactor Coolant Pump (RCP) seal thermal
barrier cooling and seal injections are lost due to an
Appendix R fire while the RCP's are on, seal integrity
could not be guaranteed for longer than ten minutes. This
ten-minute limit was specified by the vendor on March 2,
1987. This information modified the exemption request of
February 2, 1987, which stated that at least thirty minutes
was available for manual actions if a fire occurs in fire
areas CB-FA-2d and CB-FA-2f to restore seal thermal barrier
cooling and seal injection. GPUN states in their March 10,
1987 letter that they will procedurally restore seal ther-
mal barrier cooling immediately or trip all four RCP's.
The licensee also committed to include the fire areas in
question in the twenty-minute roving fire watch and to dis-
patch an operator to the remote shutdown area to take
appropriate manual actions for a fire in these two areas.
The inspector reviewed EP 1202 .! and EP 1202-37 and con-
firmed that procedurally an operator would be dispatched to
the remote shutdown area to restore seal thermal barrier
cooling and seal injection after fire damage. In NRC let-
ter, dated March 19, 1987, NRR concurred that the twenty-
minute roving fire watch will detect fires early in their
formative stages and allow time to extinguish the fire and/
or take appropriate manual actions. Since the fire areas
CB-FA-2d and CB-FA-2f were previously part of the roving
twenty- minute fire watch route, the inspector had no fur-
ther questions on this issue.
- -
79
l
9.16 Fire Protection Surveillances
4
The inspector reviewed the following fire protection surveillance
procedures for incorporation of Appendix R modifications, completion
of required surveillances prior to startup, conformance to technical
specifications, and for overall accuracy and completeness.
--
SP 1303-12.9, Revision 13, dated February 27, 1987, " Fire Bar- j
rier Seal Inspection;"
--
SP 1303-12.23, Revision 5, dated March 6,1987, and performed )
March 19, 1987, " Visual Inspection of Fire Dampers;" j
1
--
SP 1303-12.8 A,B,C,D,E,F,G,H, and I, " Fire Protection Instru- )
mentation Functional Test," for the control building, diesel,
generators, screenhouse, reactor building, intermediate build- !
ing, auxiliary and fuel handling building;
--
SP 1303-12.17, Revision 9, dated October 20, 1986, and completed I
March 23,1987, " Fire System Testing Miscellaneous Deluge Func- '
tional Test;
--
SP 1301-15.1, Revision 0, dated March 17, 1987, " Appendix R Cold
Shutdown Repair Material Inventory" and completed March 22, 1987,
and GPUN memorandum No. 3330-87-0017, dated February 17, 1987,
" Staging of Required Appendix R Repair Material;"
--
SP 1303-12.16, Revision 12, dated March 14, 1987, and completed
March 19, 1987, " Fire System Testing Air Tunnel Deluge Functional
Test;"
--
SP 1303-12.22, Revision 6, dated February 13,1987, " Fire Door
Inspection - Screenhouse," (daily check);" and,
--
SP 1303-12.21, Revision 7, dated February 18,1987, " Fire Door 4
Inspection - Primary Side," (daily check). '
In reviewing the above surveillances, the inspector noted no viola-
tions or deviations. New surveillances, surveillance changes, and
surveillances requiring completion during 6R appear to have been
completed in a timely and adequate manner. I
9.17 { Closed) NRC 3taff Temporary Instructions (289/85-15-61 and 62):
10 CFR 50 Appendix R Review
The above-noted Office of Inspection and Enforcement temporary
instructions (TI's) were completed for TMI-1 with NRC Inspection
Report N ' *289/86-23. Residual outstanding inspection findings
.
were addrei. _ above and closed. Accordingly, these TI's are admin-
, 1stratively closed for TMI-1 based on the above review.
j
80
{
9.18 Fire Protection Summary
i
Although a good number of outstanding inspection findings were iden-
tified in 1986; overall, the licensee was responsive toward the sat-
isfactory resolution of related issues. The inspector noted that
some items could have been precluded, in part, had better communica-
tions occurred between respective organization representatives and,
in part if licensee personnel would have given a more complete
initial review of the issues.
The items in IR 50-289/87-05 identified as requiring further inspec-
tion prior to startup have been resolved. The NRR will perform an j
overall safety evaluation of the licensee's compliance with Appendix
R at a later date.
The last minute change from a maximum of thirty minutes to ten min-
utes required to restore RCP seal thermal barrier cooling and seal
injection during an Appendix R fire suggests that the licensee's
engineering review process along with vendor support appeared to have
been inadequate, in part, to support startup.
10.0 Licensee Action on Other Previous Inspection Findings
10.1 (Closed) 10 CFR Part 21 Report (289/86-PT-01). Diesel Generator
Blower Inspection
This item concerned the inspection of the scavenging air blowers for i
both of the emergency diesel generators (EDG's) for further evidence 1
of lobe . scaring and clearance tolerances in accordance with Colt l
Industries' Service Information Letters (SIL's), dated November 15,
1984, and August 13, 1985 (see NRC Inspection Report No. 50/289/
86-05). Also, the item involved the inclusion of the SIL's recom-
mendations for blower ' inspection into the appropriate procedures.
The inspector reviewed Surveillance Procedure (SP) 1301-8.2, Revision
29, " Diesel Generator Annual Inspection," completed for both EDG's
during the 6R outage for incorporation of SIL recommendation, lobe
tolerances, and indications of further wear. SP ' 1301-8.2 has been
revised to include the tolerance inspections specified in the SIL's
and all tolerances appear to be within specification. Through inde-
pendent verification, the inspector verified that there are no new
indications of scaring on the blower lobes. The inspector had no
further questions and this item is closed.
81
10.2 (Closed) Unresolved Item (289/85-20-02): A Low Leakage Fuel Design
to be Implemented in the Cycle 6 Core
As part of pressurized thermal shock issue, the licensee previously
committed to implement a low leakage fuel design in the Cycle 6 core.
The inspector reviewed the licensee's Cycle 6 reload submittal (GPU
letter 5211-86-2182) and noted that the low leakage core design f0r
an "in-out-in" fuel management strategy was incorporated in this
cycle's fuel design. The inspector further reviewed the licensee's
refueling procedure, RP 1507-12, " Visual Inspection," and verified
that the as loaded Cycle 6 core agreed with the intended low leakage
loading scheme. This item is closed.
10.3 (Closed) 289/85-22-02: Excessive Challenges to the Emergency
As noted in Inspection Report No. 50-289/85-22, the licensee noted
that the start of the turbine-driven EFW pump caused the reliefs as-
sociated with the turbine to lift. The licensee committed to evalu-
ate the situation and develop a permanent solution. The licensee's
solution included replacing the relief valves with a higher setting
(approximately 300 psig), installing a different pneumatic pressure
controller for MS-V-6, and installing a time delay control logic for
steam supply valves MS-V-13A and B. To ensure that the modifications
resolved the issue, functional testing was performed to ensure no
relief valve lifts occurred.
The inspector reviewed the logic changes and testing (applicable sur-
veillance procedure at the time of plant shutdown October 31, 1986).
Based on this review and -previous review of test data, the inspector
concluded that the licensee's proposed resolution adequately ad-
dressed the original concern.
10.4 (0 pen) Unresolved Item (289/85-24-01). Training Feedback
An ASLB Partial Initial Decision (PID), dated May 3,1985, required
the licensee to develop a method to provide supervisors with a means
to give feedback into the training programs by directly evaluating
the effect of training on the actual job performance of trainees
under their supervision (performance-based training evaluation). The
licensee used an existing procedure which provided for direct trainee
feedback into the training program, modified it to incorporate super-
visory feedback, and issued it as 6200-ADM-2682.10 " Trainee Evalua-
tion - Once Back on the Job," which became effective September 13,
1985. At the time of NRC Inspection No. 50-289/85-24, this procedure
had not been implemented for licensed operators.
_ __ _ _----
l
82
Implementation for licensed operators was documented by an October
31, 1985, memorandum to supervisors directing their evaluation and
input for all licensed personnel. The results of this evaluation
were summarized in a May 29, 1986, memorandum to the Manager of
Operator Training. A second evaluation for all licensed personnel
was directed by memorandum on November 6,1986, but results were not
yet available.
The examiner reviewed the ASLB PID of May 3, 1985, and subsequent
licensee proposal to meet these concerns in addition to the NRC's 1
approval of this proposal. The results of the evaluation directed t
by the October 31, 1985, memorandum were reviewed.
The applicable procedure has been fully implemented for licensed
operators on an annual basis and for newly licensed operators on a
one-time basis six months after assignment to licensed duties. This
procedure provides for both trainee feedback (on a form labeled
Exhibit 1) and supervisor feedback (on a form labeled Exhibit 2).
The focus of this review was on the use of form Exhibit 2.
Exhibit 2 contains sixteen questions for the supervisor to answer
regarding the performance of the operator / trainee. Additionally,
there are two questions which imply that the operator / trainee may
voice concerns or offer feedback into the training program and this
feedback is then documented on the form by the supervisor. A review l
of completed Exhibit 2 forms show that the overwhelming source of
feedback from this process in terms of suggested changes or improve-
ments to the training programs is coming from the operators / trainees.
The licensee has fully implemented an NRC-approved plan to meet the
ASLB concerns stated above. Supervisors are being given the oppor-
tunity to evaluate operator / trainee job performance and provide feed-
back into the licensed operator training programs. However, there
exists an appearance that supervisory feedback is taking place on a
form specifically developed for supervisors when, in fact, the feed-
back is coming from the operator / trainees.
The fact that supervisors are finding no major problems with employee
job performance may be quite valid, but the mechanism for supervisor
evaluation was intended to be separate and distinct . from operator /
trainee inputs. The licensee representatives acknowledged the above
and they agreed to separate or distinguish supervisor feedbacks from
operator / trainee feedbacks.
This item should remain open until the licensee separates supervisory
input from operator / trainee input and evidence exists that super-
visors are giving a thoughtful assessment of job performance from the
standpoint of evaluating training effectiveness,
i
I
.
. . . . . .
.. . . .
.
.
i
_
.
!
I
i
!
83 l
10.5 (Closed) (289/86-09-03): Review of IST Relief Request for Testing of
Boric Acid Pumps
) 1
The licensee had requested an exemption from the ASME Code for quar- j
terly testing of the boric acid pumps CA-P-1A/B. The basis for this j
request was that sufficient instrumentation was not installed on the !
pump piping to take the required data and no acceptable flow path was ]
available for testing. NRR denied this request in a letter, dated
March 19, 1987, and the licensee will be required to make piping-and
instrumentation modifications and commence testing of the pumps -fol- ]
lowing the 8R outage. The completion of this modification will be
followed by an unresolved item (see Section 11) which included other l
scheduled modifications that the licensee must make to resolve other
NRR denied exemption requests. This item is closed.
10.6 (Closed) Unresolved Item (289/86-10-01). Review Licensee Evaluation
of Reactor Vessel Internals Vent Valve Exercise Procedures
{
j
The licensee committed to review SP 1301-10.1, " Internals Vent Valve i
Inspection and Exercise" prior to the next performance of the proced- I
ure. NRC concerns were that an individual using a manual lifting
device would exceed the 400 pound maximum allowable force required by j
TS. The licensee used NUREG/CR 3517 PNL-4865 and determined that l
there was good assurance that an individual using the manual lifting i
tool (backup method) for this job would not exceed the 400 pound
lifting requirement.
A procedure note was also added to only allow one individual to per-
form the lift. Additionally, a requirement was added to record data
for the scale that is used when the polar crane is used (normal met-
hod) to perform the check valve lift.
The inspector reviewed the latest completed version of SP 1301-10.1,
Revision 10, dated December 8,1986, which was completed on December
23, 1986, and verified that the appropriate changes were made. Dur-
ing the performance of this test, the backup method was used to per-
form the verification of check valve operability. The inspector had
no other concerns and this item is closed.
10.7 (0 pen) Violation (50-289/86-12-02): Inadequate Safety Evaluation for
Change to Procedure Described in FSAR and Modifications and Single-
Failure Analysis for Backup Instrument Air
A part of this violation concerned a procedure for testing the motor-
driven emergency feedwater pumps (SP 1103-11.42). The Final Safety
Analysis Report (FSAR) required that the discharge cross-connect '
valves remain open for the surveillance test, but the procedure had
the valves (EF-V-2A/B) closed. The licensee changed the procedure,
l
___
-
_. .. .
- - - -
-
-
- - - -
-
J
i
l
84
Revision 8, dated October 17, 1986, to correct the position of these
valves for the test. The inspector reviewed the procedure change and
a subsequent completed surveillance procedure to veri fy correct
implementation of the change. This portion of the violation is
closed.
10.8 (0 pen) Inspector Follow Item (289/86-13-02): Emergency Feedwater
Pump Bearing Temperature
This item concerned the review of the engineering evaluation and
licensee actions for resolving possible long-term bearing and turbine
degradation to the turbine-driven emergency feedwater pump (EF-P-1)
apparently due to steam valve leakage into the turbine casing. The
inspector reviewed the following documentation to determine if appro-
priate measures have been taken by the licensee to resolve _ the -issue.
--
GPUN Memorandum (No. 3310-86-0171), dated September 8, 1986,
"EF-P-1 Bearing Temperatures / Steam Leakage;" and,
--
SP 1300-3G A/8, " Turbine-Driven Emergency Feedwater Pump Func-
tional Test," completed February 28, 1987, and March 13, 1987.
The above memorandum is the engineering evaluation concerning equip-
ment operability with higher than ambient bearing and casing temper-
atures. Plant engineering considers that the steam leakage does not
pose a continuous erosion problem if the affected steam piping re-
mains drained. The inspector confirmed that the affected steam pip- ,
ing is drained at least once per shif t. Also, during the just com-
]
pleted outage, the licensee inspected the steam inlet valves (MS-V- ]
13A/B) for possible steam cutting. No indications were found. Prior
to plant startup, the licensee successfully tested EF-P-1 twice (SP
1300-3G A/B).
However, after startup the inspector noted that the same condition
exists as described in IR 50-289/86-13 suggesting possible steam cut- l
ting of (MS-V-13A/B bypass) inlet valves MS-V-10 A/8. The inspector '
also noted a chattering noise coming from the area of MS-V-9A, a
check valve in the steam supply line f rom the "A" 0TSG to EF-P-1.
Subsequent investigation revealed that steam was emitting from the
vent of the EF-P-1 turbine exhaust. The vent is located outside the
intermediate building. It appeared that steam leaks in the system
had increased to the point to cause this check valve to move slightly
off its seat, indicating that leakage had increased since the last
cycle. The resident inspectors will follow further licensee actions '
concerning repair of valves MS-V-10 A/B.
-
85
10.9 (Closed) Unresolved Item (289/86-19-04). Discrepancy Between Chest
and Upper Arm Dosimetry Monitoring Results During Spent Filter
Disposal Operations
Licensee investigation into the above event included the interview
of involved radwaste personnel and the performance of extensive rad-
iation surveys to " profile" the high radiation beam emanating from
the top of the spent f11ter cask when the shield plug is removed.
The licensee concluded the desimetry discrepancy was an isolated
event which resulted from the performance of a non-routine activity
(wiping up spilled water droplets) during the transfer operation.
Based upon the results of their radiation dose rate profile survey,
the licensee also took the following steps to minimize exposure and
ensure proper monitoring during the filter transfer operation.
--
Scaffolding around the cask was lowered approximately 30 inches
to take advantage of shielding provided by the cask itself;
--
Whole body thermoluminescent dosimeter (TLD) badges were placed
at the head and ground levels. The licensee's study indicated
that either the head or ground regions, depending on the work-
er's specific function, is the limiting portion of the whole
body.
The inspector reviewed licensee survey results and a summary of the
investigation and found the above conclusion reasonable. The inspec-
tor alse verified that dosimetry placement as described above had
been incorporated into ALARA Review No. 87-03-11, "MU-F1A/1B/2A/28/
4A/4B Changeout and Disposal."
10.10{0 pen)UnresolvedItem(289/87-06-07): HSPS Test Results
Evaluation
Additional review occurred for this item in preparation for startup
(see paragraphs 3.2.4 and 5.2.2.1). This remains open pending licen-
see completion of their test results evaluation for the HSPS preoper-
ational test procedures and pending Region I verification of all
documented test results to satisfy TS requirements.
10.11 (0 pen) Unresolved Items (289/87-06-08 and 09): Licensee
Complete Engineering Evaluations for HSPS
This was addressed in paragraph 2.5. This item remains unresolved
pending additional specialist review by Region I.
-
86
10.12 Past Inspection Findings Summary
In general, the licensee was responsive to the issues raised by NRC
staff. The items that remain open were expected by NRC staff, since
it was NRC staff's intention to update this area.
11.0 Inservice Testing Program Relief Request Resolution
On March 19, 1987, NRR issued a letter to GPUN to disposition remaining
unresolved relief requests for the second ten year inservice testing (IST)
program at TMI-1. In this letter, a schedule was established for the
licensee to incorporate procedural changes, make plant modifications, or
provide justification in order to bring the IST program into compliance
with appropriate ASME Codes. Accordingly, the proper implementation of
these required actions will remain the subject of three unresolved items.
The first unresolved item (289/87-09-17) is opened to track license action j
required prior to either August 19, 1937 or September 19, 1987. NRR re- i
quired the licensee to procedurally incorporate the testing of some spent i
fuel cooling system valves into the IST program. Relief was granted until
the end of August 1987.
Also, the licensee was required to incorporate the testing of diesel gen-
erator air start valves EG-V-17A/B prior to September 19, 1987, or provide
justification that the testing was not required. The incorporation into
the IST program of the above-listed valves will be reviewed by the inspec-
tors when the licensee completes the appropriate changes and commences
testing of the subject valves.
The second unresolved item (289/87-09-18) will track the completion of
plant modifications that NRR required the licensee to complete prior to
startup after the 8R outage.
The licensee is required to install or modify systems to allow for quar-
terly testing of boric acid pumps CA-P-1A/B and boric acid recycle pumps
WDL-P-13A/B. The inspectors will review the completion of these modifica-
tions and subsequent testing at that time.
The third unresolved item (289/87-09-19) will track completion of plant
modifications / procedure changes that NRR required the licensee to complete
prior to the startup from the 7R outage. The licensee is required to
modify the diesel fuel oil system to allow certain testing of the fuel oil
transfer pumps, specifically flow measurement. Additionally, the testing
of six check valves in the system, DF-V-23A/B, DF-V-7A/B, B/A, A/B, and
B/B will need to be proceduralized prior to startup from 7R.
Finally, the licensee will be required prior to 7R startup to modify sys-
tems to allow for testing of SH-P-3A/B control building chill water pumps
and SE-P-2A/B, screenhouse ventilation equipment pumps.
i
87
12.0 Exit Interview !
The inspectors discussed the inspection scope and findings with the
licensee management at a final exit meeting conducted April 24, 1987.
Interim exit meetings occurred on: March 12, 1987, in the training area;
March 17, 1987, in the surveillance area; March 26, 1987, in the fire
protection and startup testing areas. Licensee personnel in attendance at
the final exit interview are noted below.
l
J. Colitz, Plant Engineering Director, TMI-1 ;
D. Hassler, Licensing Engineer, TMI-1 i
H. Hukill, Director, TMI-1 l
C. Incorvati, TMI-1 Audit Manager i
G. Kuehn, Manager, Radiological Control, TMI-1
8. Mehler, Manager, Radwaste Operations 1, TMI-1 i
A. Palmer, Deputy Manager, Radiological Field Operations l
L. Ritter, Administration, Plant Operations j
M. Ross, Plant Operations Director, TMI-1 ;
R. Shaw, Radiol >gical Engineering Manager, TMI-1 l
C. Shorts, Manager, Technical runctions, TMI-1
H. Teichman, Radcon/ Chemistry Nonito Lead, TMI-1
R. Toole, Operations and Maintenant.a Jirector, TMI-1
No proprietary information was discussed at the exit meetings. The in-
spection results, as discussed at the meeting, are summarized in the cover
page of this inspection report.
Unresolved Items are matters about which more information is required in
order to ascertain whether they are acceptable, violations, or deviations.
Unresolved items discussed during the exit meeting are addressed in
Sections 4, 7, 8, 9, 10 and 11.
To facilitate the licensee's discussion at the exit interview on a par-
ticularly long reporting inspection period with a good number of inspec-
tors, the inspector provided the licensee with a list of open/ closed pre-
vious inspection findings. In no way did this change the status of those
findings which are reflected in the report (paragraphs 9 and 10).
. - - _ - _ - - _
INSPECTION REPORT NO. 50-289/87-09
ATTACHMENT 1
ACTIVITIES REVIEWED
Portions of the following documents / records and related activities were re-
viewed:
General
--
Tag Out Log Audit (3/25/87)
--
Shift Manning Verification (3/26/87)
--
Shift Turnovers (various dates)
--
Observed check of Administrative Procedure (AP) 1067, " Independent Verif-
ication Program: (3/22/87)
Operations
--
Containment Closecut Complete (3/19/87)
--
Rod Control Operations and Testing (3/20/87)
--
Observed OP 1106-3, "Feedwater System" "1A" feedwater pump trip logic
testing (3/20/87)
--
OP 1106-2, Revision 92, Plant Heatup to 525 F, selected step verifications
(3/21/87) '
--
Toured reactor and auxiliary buildings for final review prior to critical-
ity (3/22/87)
--
Seal injection, makeup, and letdown correlation (3/22/87)
--
Commenced boiling down Once-Through Steam Generator's (OTSG's) to 18
inches per startup procedure (3/22-23/87)
--
Reviewed (draft) Technical Change Notice (TCN) for OP 1102-2, which
changes the 18 inch control setpoint to 30 inch steam generator level
(3/23/87)
--
Tuning of proportional and integral controller for ICS "A" OTSG control in
relay room (3/24/87)
_
l
Attachment 1 2
Surveillance Tests
--
SP 1303-11.10, Revision 17, " Engineered Safeguards System Emergency Se- l
quence and Power Transfer Test" ("B" Train) (3/19/87)
--
SP 1303.5.2, "ESAS Load Sequencing Test" (3/20/87)
--
SP 1303-8.1, Revision 13, Reactor Coolant System Hydrostatic Test (3/20/87)
--
Test of Power Operated Relief Valve (PORV) (3/20/87)
--
SP 1303-11.1, Rod Drop Verification (3/21/87) l
I
--
RCS Leak Rate Test (3/21/87
--
SP 1303-11.1, Control Rod Drop Time Test (3/22/87)
--
SP 1303-11.39 (Draft), HSPS - Emergency Feedwater Pump Automatic Start (
(3/22/87) '
--
SP 1303-11.25, RB Local Leakage Test - Access Hatch Door Seals (3/24/87)
--
Heat Balance and Nuclear Instrumentation Alignment (3/26/87)
Maintenance
--
Steam Generator Level Troubleshooting and Calibration (selected activities )
3/23-26/87) {
l
--
RCS Flow Transmitter Leak Repairs (3/22/87) !
--
RCS Pressure Indicator Failed High Repair (3/22/87)
Startup Testing
--
Test Procedure (TP) 332/3, HSPS Functional Test (3/20/87) l
--
TP 664/3, PZR Heat Loss Test (3/21/87)
l
--
TP 683/1, " Remote Shutdown Outside Control Room" (Procedure Review
3/21/87 and Test (3/22/87)
i
i
l
T
ATTACHMENT 2
RCS LEAK RATE RESULTS (3/21-31/87)
Unidentified
Date Time Leakage + Losses Leak Rate
3/21/87 1451 0.1642 0.0717
3/22/87 0003 0.3999 0.3419
1550 0.1993 0.1285
3/23/87 0117 0.4421 0.1992
1133 0.0743 0.0521
3/24/87 0013 0.3150 0.2865
2341 0.0665 0.0631
3/25/87 0616 0.2832 0.2450
3/26/87 1103 0.1538 0.1190
2132 0.2927 0.2642 '
- 3/27/87 0714 0.1719 0.1713
2006 0.2103 0.1912
3/28/87 0434 0.2349 0.2315
1017 0.4173 0.4232
1947 0.2637 0.2747
- 3/29/87 0758 0.2395 0.2408
1653 0.2021 0.2237
3/30/87 0110 0.1686 0.1385
0700 0.1302 0.1188
2028 0.1760 0.1785
3/31/87 0130 0.1335 0.0971
1043 0.0842 0.0307
1822 0.1229 0.0630
"NRC independently verified
i
.. ..
......, . .
.
. . . , , - .
,. .. . 4