ML20046D380
| ML20046D380 | |
| Person / Time | |
|---|---|
| Site: | Quad Cities  |
| Issue date: | 08/13/1993 |
| From: | Guzman J, Paul Prescott, Zelig C NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML20046D378 | List: |
| References | |
| 50-254-93-17, 50-265-93-17, NUDOCS 9308190070 | |
| Download: ML20046D380 (22) | |
See also: IR 05000254/1993017
Text
,
. - _.
- _ _ _ _ _ , __-- - _ .
,
I
4
'
L.
.
.
.
O. S. NUCLEAR REGULATORY COMMISSION
'
REGION III
l
Reports No. 50-254/93017(DRS); No. 50-265/93017(DRS).
Docket Nos. 50-254; 50-265
Licenses No. DPR-29;.No. DPR-30
l
Licensee: Commonwealth Edison Company
Executive Towers West III
,
1
1400 Opus Place
Suite 300
Downers Grove, IL 60515
I
l
Facility Name:
Quad Cities Nuclear Power Station - Units 1 and 2
Inspection At:
Quad Cities Site, Cordova,-IL 61241
'i
Inspection Conducted: June 9 through August 12, 1993
Inspectors:
J. Guzman, DRS
L
P. Prescott, DRP
j
C. Zelig, DRS
Approved By:
f
// h
8/['J
J
C. Vandefniet
-
Da't e '
Team Leader
- '
1
Approved By: r.~[ldrD _-
kr3hil
,
'
'3
/-(_
a
!
s
G. C. Wright, Chief
Date'
i
Engineering Branch
j
i
Inspection Summary
"
Inspection on June 9 throuah Auaust 12. 1993 (Reports No. 50-254/93017(DRS):
No. 50-265/93017(DRS))
Areas Inspected: Special inspection conducted in response to the Unit-1 HPCI
exhaust rupture disc actuation event at the Quad Cities Nuclear Power Station
on June'9, 1993. The review included validation of the sequence of events, 'a
-
,
I
review of operational performance and maintenance history for the HPCI system,
and evaluation of the following areas; (1) root cause for the disc failure,
-(2) response to the. event, (3) effects of the . steam release, .(4) corrective
.. actions program, (5) testing controls, (6) system engineering, (7) preliminary
investigation team findings, and (8) licensee's event classification and
reporting of the event.
I
Results: The rupture discs on the Unit 1 HPCI system turbine exhaust were
-l
breached. The resultant steam release contaminated and burned five
i
individuals. The turbine drains on both HPCI and Unit 2 RCIC systems were
found to be malfunctioning. Condensate in the Unit 1 HPCI turbine casing.
,i
,
resulted in excessive wear to the thrust bearing.
The fire doors separating
the Units 1 and 2 HPCI rooms were blown off their hinges
Both sets of
9308190070 930813
[
ADDCK 05000254
O
V
.
R
-
.
-.-
-
. . . - . - . .
.
. ..;
_ __________________ __ _ ______
'
.
.
secondary containment doors leading from the Unit 1 HPCI room were blown open.
The root cause of the event was a gaseous overpressure condition, resulting
from a cohesive mass of water expelled from the turbine.
Rupture disc age and
service degradation possibly contributed to the failure.
Test controls prior
to and during the running of the Unit 1 HPCI surveillance were inadequate.
Neither the rupture discs or the HPCI drain level switches were in the
licensee's preventive maintenance program.
Evaluations completed by
contractors and the licensee identified problems with the adequate review and
correction of deficiencies regarding the HPCI drains' level switches, however,
prompt corrective actions were not taken. System engineering lacked
understanding of integrated plant and system operations. The licensee's
investigation team issued a preliminary report with several recommendations
which was reviewed by the inspectors, however, detailed corrective actions
were not yet specified. Significant operational safety parameters were not
approached, however, the event called into question the operability of both
Units' HPCI and RCIC systems and posed a threat to the accident' mitigation
capabilities of Unit 1.
The overall HPCI and RCIC system materiel condition
and housekeeping for both units was considered poor.
Two apparent violations were identified. Section 2.4.1 describes three
examples of a failure to follow procedures and inadequate procedures.
Section
2.4.2 describes problems associated with corrective actions program.
G
v..
. . . . .
..
_ _ _ _ _ _ _ _
l-
.
.
.
.
TABLE OF CONTENTS
Pace
1.
Persons Contacted. . . . . . . . . . . . . . . . . . . . . . . . .
I
2.
Inspection Results . . . . . . . . . ..
l'
..............
l.
Description of Occurrence. . . . . . . . . . . . . . . . . .
I
!
2.
Sequence of Events . . . . . . . . . . . . . . . . . . . . . 2
~
l
3.
Equipment Analysis . . . . . . . . . . . . . . . . . . . . . 3
!
1.
HPCI Steam Isolation Valves. . . . . . . . . . . . . .
5
L
2.
HPCI Exhaust Check Valve (2301-45) . . . . . . . . . .
5
',
L
3.
HPCI Exhaust Closable Check Valve (2301-74). . . . . . 6
4.
Pressure Switches PS-2368A & B . . . . . . . . . . . . 6
i
5.
Pressure Switch PS-2355 . . . . . . . . . . . . . . .
6
6.
HPCI Turbine Thrust Bearing. . . . . . . . . . . . . . 7
7.
Level Switch LS-2365, HPCI Turbine Inlet Drain Line
Level. . . . . . . . . . . . . . . . . . . . .
. . .
7
'
8.
Level Switch LS-2369, HPC1. Turbine and Exhaust Drains
>
Level. . . . . . . . . . . . . . . . . . . . .. . . . .
7
9.
HPCI Room Secondary Containment Doors. . . . . . . . .
8
'
10.
HPCI Room Fire Doors . . . . . . . . . . . . . . . . .
9
4.
Evaluation of Human Performance. . . . . . . . . . . . . . .
9
1.
HPCI System Operations . . . . . . . . . . . . . . . .
9
2.
Failure to Recognize Significance of Rupture Disc and
-
i
level Switch Problems. . . . . . . . . . . . . . .
. 11
!
3.
System Engineering . . . . . . . . . . . . . . . . .
13
4.
Effectiveness of Licensee Response and Event
Evaluation . . . . . . . . . .
14
..........
5.
Materiel Condition and Housekeeping of HPCI and RCIC
Rooms. . . . . . . . . . . . . . . . . . . . . . . .
16
l
.,5.
Quality Assurance Deficiencies . . . . . . . . .
. . . .
16
6.
Radiological Consequences. . . . . . . . . . . . .
16
i
...
7.
" Rupture" Mechanism. . . . . . . . . . . . .
16
....
..
8.
Conclusions.
17
........
..
..........
l
1.
Opposite Unit and RCIC Applicability . . . . . . . .
17
2.
Recommended Actions. . . . . . . .
17
.....
..
3.
Test Cont rol s. . . . . . . . . . . . . . . . .
. .
17
4.
Ineffective Corrective Action Program. . . . . . . .
18
1
y
5.
System Engineering . . . . . . . . . . . . . . . . . .:18
3.
Exit Interview . . . . . . . . . . . . . . . . . . .
19
......
4.
Simplified HPCI System Drawing . . . . . . . . . . . . . . .
.
20
l
-i
,
l~
E
j
,
,
'
.
.
'
.
DETAILS
(
l.0
Principal Persons Contacted
i
!
Commonwealth Edison Company (CECO)
M. Wallace, Vice President, Chief Nuclear Officer
L. O. DelGeorge, Vice President, Nuclear Oversight an1 regulatory Services -
- R. Pleniewicz, Site Vice President
- G. Tietz, Executive Assistant
- R. Bax, Station Manager
'
- R. Moravec, Manager, Site Engineering and Construction
J. Burkhard, Superintendent, Quality Verification
D. Craddicu, Superintendent, Maintenance
-H. Hentschel, Manager, Operations
,
J. Leider, Superintendent, Technical Service
- A. Chernick, Supervisor, Performance Improvement
o
'
- L. Moerke, Supervisor, Engineering Design
>
A. Scott, Supervisor, System Engineering
R. Walsh, Supervisor, Station Support Engineering
U. S. Nuclear Reaulatory Commission (NRC)
,
,
J. Martin, Regional Administrator, Region III
J
E. Greenman, Director, Division of Reactor Projects (DRP)
- T. Martin, Acting Director, Division of Reactor Safety (DRS)
- M. Ring, Chief, DRS Operations Branch
H. Clayton, Chief, DRP Branch 1
l
- P. Hiland, Chief, DRP Section IB
- T. Taylor, Senior Resident Inspector, Quad Cities (DRP)
1
1
- Denotes those present during the exit conference call on August 12, 1993.
2.0
Inspection Results
l
l
2.1
Description of Occurrence
On June 9,1993, while preforming the Unit 1 High Pressure Coolant Injection
(HPCI) system quarterly surveillance test, the turbine exhaust rupture disc
burst making the HPCI system inoperable and injuring five people.
The
pressure surge, associated with the rupture disc burst, blew the fire doors
separating the Unit 1 and 2 HPCI rooms from their hinges into the Unit 2 HPCI
room. The surge also caused two sets of secondary containment doors leading
from the Unit 1 HPCI room to be blown open, damaging the doors' latching
mechanisms.
The steam release was terminated when the inner and outer HPCI.
steam supply isolation valves (2301-4 & 5) closed in response to the room high
temperature (155'F) isolation signal; operators also inserted a manual turbine
trip.
!
All five injured personnel were transported to local medical facilities where
l-
they were decontaminated and treated for steam burns.
Four were treated and
l'
,o
g
3
1
_
i
..
,
.
rel. eased; one, suffering first and second degree burns to 20% of her body, was
hospitalized for 10 days.
2.2
Seouence of Events
-The sequence of events in the control room occurred in a very rapid fashion.
The initial alarm associated with the event was the:HPCI turbine rupture disc-
high pressure alarm at 08:50:56. This alarm instantaneously cleared and two
seconds later the HPCI thrust bearing wear active face alarmed actuated.
'
These alarms were followed two seconds later by the HPCI room high radiation
and four seconds later by the HPCI room high temperature alarm.
Several
intermittent 125 and 250V DC alarms were also received during the event.
After the initiation of HPCI system isolation and turbine trip, 20 and 35
,
seconds after the burst, the high radiation and temperature alarms cleared.
A more detailed sequence of events, taken from the plant computer alarm
,
printer has been included below.
!
9 June 1993
07:25:00
Unit 1 HPC.I Placed on Turning Gear
'
07:55:00
Hydrogen addition taken off to reduce HPCI room
>
radiation levels,
h
08:20:00
1/2A Standby Gas Treatment System (SBGTS)
,
started for DOP and freon tests.
08:30:00
Personnel Start Arriving in HPCI Room
08:50:00
Loop A torus cooling placed in service.
,
08:50:00
Unit 1 NSO pushed BLOCK MOTOR SPEED CHANGER
pushbutton per QCOS 2300-5 Step I.23.b.
08:50:56
HPCI Turbine Rupture Disc High Pressure
08:50:56
HPCI Turbine Rupture Disc High Pressure - OK
08:50:58
HPCI Thrust Bearing Wear Active Face
08:51:00
Rx. Bldg. Hi Rad: HPCI Cubicle Sta 16
08:51:02
125 V Battery Ground: Negative (alm)
08:51:02
Area Hi Temp Stm Lk Detect: HPCI Sys TE-2374C
08:51:03
250 V DC Battery Ground: Positive (alm)
08:51:03
250'V DC Battery Ground: Positive (cir)
08:51:07
HPCI Pump Area Hi Temp: Ch C
08:51:12
HPCI Pump Area Hi Temp: Ch 0
'
08:51:12
HPCI Pump Area Hi Temp: Ch B
08:51:12
HPCI Grp4 PCI Vivs DC.Div Isol
08:51:16
125 V Battery Ground: Negative (c1r)
08:51:16
HPCI Pump Area Hi Temp: Ch A
08:51:16
HPCI Grp4 PCI Vlvs AC Div Isol
08:51:16
HPCI Grp4 PCI Vlvs Not Open: Valve 2301-5
08:51:16
HPCI Grp4 PCI Vivs Not Open: Valve 2301-4
'
08:51:22
Area Hi Temp Stm Lk Detect: HPCI Sys TE-2374
e
!
NOTE: N50 tripped turbine when he noticed isolation valves
i
closing.
08:51:31
HPCI Turbine Tripped
08:51:33
HPCI Thrust Bearing Wear Active Face - OK
'
08:52:32
Rx Bldg Hi Rad: HPCI Cubicle Sta 16 - OK
08:53:31
HPCI Pump Area Hi Temp: Ch B - OK
2
i
i
i
.- - N
.
_
.
f
-
08:54:03
HPCI Pump Area Hi Temp: Ch D - OK
08:54:05
Area Hi Temp Stm Lk Detect: HPCI Sys 2374 - OK
J
08:54:23
HPCI Pump Area Hi Temp: Ch C - OK
{
08:54:37
HPCI Pump Area Hi Temp: Ch A - OK
l
NOTE: Received emergency call from an Auxiliary Operator in Core
j
j
Spray room stating four other injuries.
Informed control
room that fire doors between HPCI rooms and interlock doors
,
damaged.
i
09:00:
Ambulances, Security, and Rad Protection
l
t
notified,1/2 B SBGTS started.
NOTE: Several Unit 2125 and 250 V DC alarms were received and.
j
cleared between 08:54 and 09:34.
-
i
09:04
QGA 300 entered and exited for conditions
cleared on isolation
'
09:15
Mechanical Haintenance dispatched to assist in
>
closing interlock doors. Shift Engineer (SE)
referred to Tech Specs for secondary containment
requirements.
09:18
Unusual Event (UE) declared for removal of
injured
1
09:21
SE held control room brief
09:24-
NARS notification made
09:25
Electrical Maintenance (EM) dispatched to HPCI
i
room to string lights
09:34
HPCI interlock doors closed and secure
4
!
09:35
3rd. ambulance called
1
09:50
Emergency Notification System (ENS) I and 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />
notifications made
i
10:01
Transferred command and control of General Site
Emergency Plan (GSEP) UE to Station Director
10:10
Units 1 & 2 HPCI and RCIC quarantined
10:24
Authorized EM access to HPCI area to restore
lighting and to assess damage
11:21
All injured personnel decontaminhtc?
11:40
Authorized Industrial Hygienist to sample HPCI
area for asbestos
l
11:59
UE terminated
12:24
ENS called to notify. NRC of GSEP termination
2.3
Eouipment Analysis
Through the course of this inspection several components in the Quad Cities
HPCI systems were analyzed, inspected and tested to determine what effect they
had on the event.
In the following report sections the HPCI system and
rupture disc configuration and rupture disc design are described providing
-reference to the. system (see attached drawings in Section 4).
Evaluations of
each significant system component which possibly contributed to the event is
then discussed.
Included in these evaluations are discussions on the
3
l
l
!
,
a
.
.
r
se.condary containment and fire doors which were forced open by the event's
.
pressure surge.
l
The HPCI system provides high pressure emergency core cooling for small line
breaks. These breaks could result in the reactor vessel water level dropping
to a level where the core is not adequately cooled, but with little or no
decrease in reactor vessel pressure.
!
,
,
The HPCI pump is driven by a two-stage General Electric turbine.
Steam is
supplied from the "B" main steam header, upstream from the main steam line
inner isolation valve and flow element. The 10" supply line has an inner
isolation motor-operated valve (MOV) (2301-4), outer isolation MOV (2301-5),
turbine steam supply MOV (2301-3), turbine stop valve (2317), and turbine
control valve (2321). After passing through the turbine, the' steam is
exhausted through a 24" line past a spring-loaded, dual plate check valve
(2301-45) and a closable check valve (2301-74) into the suppression chamber.
,.
The turbine exhaust line has a 16" dual rupture disc arrangement to protect
'
against an overpressure condition. The rupture discs are specified as having
a 150 psig burst pressure (manufacturer's range, 130-150 psig at 366 degrees).
The rupture disc is a solid, 16" diameter, .0085" thick, Type 316-stainless
steel, conventional domed disc. The rupture disc is attached to an
.
identically shaped vacuum support.
The vacuum support is a Type 347 stainless
!
steel, 0.040'.' thick dome, made from six pie-shaped sections.
The pie pieces
have tabs that prevent the vacuum support and the rupture disc from collapsing
l
if there is a vacuum in the exhaust piping. An air space is provided between
,
the vacuum support and the rupture disc, allowing steam pressure to act evenly
'
on the disc.
The arrangement on the HPCI exhaust employs two identical discs in series,
separated by a 21/2" spacer to accommodate the leak detection pressure
switch. The vendor indicated the dual disc configuration is intended to
ensure that bursting the bottom rupture disc will result in an upper disc
burst due to impact of bottom disc pie pieces on the upper disc. However, if
the lower disc were to get a small leak, the pressure switch would alarm,
informing operators of a degraded lower disc while the upper disc managed to
contain the pressure.
A pressure switch (PS-2355 a Barksdale model D2H) is used to sense the
pressure between the rupture discs.
If the pressure exceeds the 10 psig
setpoint, an alarm is actuated on the main control board indicating leakage
past-the inner disc.
This pressure switch has no control function associated
with it.
Additional overpressure protection is provided by two independent Barksdale
pressure switches (PS-2368A model OlT and PS-2368B model D2H) each with a
setpoint of 100 psig.
These switches are located between the rupture disk tap
off and the first check valve (2301-45).
Exceeding either pressure switch
setpoint will actuate a HPCI turbine trip.
HPCI turbine area high temperature switches (TS-2370, 2371, 2372, 2374, A,
B,.
C, & 0), have a setpoint of 155'F and are used to detect steam leakage in the
4
!
L
.
HPCI rooms. Activating any two channels of any two trains will initiate a
HPCI steam supply isolation; automatically closing the inner and outer
isolation valves (2301-4 & 5).
The HPCI system steam lines are designed to be maintained free of condensate
by two sets of drains. The inlet drain taps off upstream of the turbine steam
supply MOV (2301-3). The drains themselves consist of a standard thermostatic
steam trap arrangement with an air-operated bypass valve (2301-31). The
bypass valve is controlled by a level switch (LS-2365) which monitors the
level in an upstream condensing pot. The inlet drains are normally emptied by
vacuum dragging to the main condenser. The other drain system collects
condensate from the turbine casing, exhaust line, and steam inlet piping
downstream of the turbine steam supply MOV (2301-3) in a drain pot.
Level
switch LS-2369 monitors the drain pot's level and operates a solenoid-operated
valve (2301-32) sending the condensate to the HPCI gland seal condenser.
Alternately, the condensate can pass through a standard thermostatic steam
trap into the torus.
2.3.1
HPCI Steam Isolation Valves
The HPCI system is isolated by a Primary Containment Isolation System (PCIS)
Group IV signal.
The HPCI isolation signals are generated by three signals:
(a) HPCI turbine area high temperature; 155'F; (b) HPCI steam line high flow;
300% of rated flow for 3 seconds; and (c) reactor vessel low pressure; 100
psig. The logic for this isolation is one-out-of-two taken twice.
During the burst event, the isolation was actuated by high HPCI turbine area
temperature. HPCI turbine area high temperature alarms were received
approximately 15 seconds after the rupture discs burst.
The HPCI steam
isolation valves indicated "not open" approximately 4 seconds after receipt of
the isolation signal.
'
1
The HPCI steam isolation valves (2301-4, 2301-5) are designed to close within
50 seconds after receipt of an isolation signal.
HPCI isolation valve timing
is based on preventing core uncovery following a pipe break outside primary
containment and containing released fission products following pipe breaks
,
inside the primary containment. HPCI isolation valves 2301-4 and 5 are tested
each refueling outage using Quad Cities Operational Surveillance (QCOS) 2300-
4, HPCI System Cold Shutdown Valve Test.
This test was run twice in December
,
1992, and demonstrated closing of 2301-4 in 45.5 seconds and 2301-5 in 35.8
l
seconds.
Based on the information above, it appears valves 2301-4 and 5
'
operated according to design.
i
2.3.2
HPCI Exhaust Check Valve (2301-45)
i
HPCI exhaust line configuration was analyzed for possible problems that may
'
have contributed to the event. Review of maintenance records showed that the
HPCI exhaust check valve had recurring seating problems and a high f ailure
,
rate on leak rate tests (LRT).
It did not, however, indicate problems
i
associated with the mechanical operations of the valve.
)
5
k
._.
!
..
)
-
'
l
On June 11, 1993, HPCI Steam Exhaust Check Valve CK-1-2301-45 and 74
,
'
Functional Test was performed to demonstrate the valves' operability. The
test closed the downstream stop check valve (2301-74) and pressurized the
space between it and the upstream exhaust check valve (2301-45). This
. demonstrated that the exhaust check valve (2301-45) performed as intended by
_i
holding the required pressure. However, it did not demonstrate valve
i
operation in the open direction._
'
On June 12, 1993, the licensee opened and inspected the valve and found' that
it was operable in the exhaust configuration; however, its seating surface was
damaged. The inspection determined that the valve did not cause a flow
blockagi and was, therefore, not a contributor to the event.
Nonetheless, the
i
licensee replaced the valve.
I
2.3.3 HPCI Exhaust Closable Check Valve (2301-741
A maintenance records review of closable check valve 2301-74 (downstream of
exhaust check valve 2301-45) indicated the valve was relatively problem free.
A review of recent LRTs showed the valve was reliable.
The initial portion of the June 11, 1993, functional test, described in
paragraph 2.3.4, was to pressurize the line between the exhaust closable check
i
valve (2301-74) and exhaust check valve (2301-45) with air. When air was
l
l
applied, at approximately 2 psig, the pressure would not hold and the valve
(
l
made an audible disc movement. This proved the valve operable in the exhaust
'
mode and that the valve was not a contributor to the event.
2.3.4 Pressure Switches PS-2368A & B
'l
i
Neither high exhaust pressure switch PS-2368A or B (setpoint 100 psig) alarmed
during the event. On June 11 and 12, 1993, the licensee performed calibration
and valve alignment checks on the switches and their associated isolation
valves.
Both instruments were found correctly aligned, PS-2368A was set at
97.4 psig and PS-2368B was set at 95.0 psig.
This demonstrated that both
_
instruments were in calibration and should have alarmed if there had been an
overpressure condition.
Instrument response testing was conducted by technicians from the Instrument
l
Maintenance Department.
The testing results showed that while the switches
'
were in place with snubbers (pulsation dampeners) attached, they responded -in
approximately I second. The switches _were removed from the rack and taken to
I
the instrument shop for further testing. Bench testing showed that, with the
j
snubbers attached, PS-2368A responded in 1.2 seconds and PS-2368B responded in
1.5 seconds.
This disparity is probably attributed to the switches being
3
different models. With the snubbers removed the switches had a response time
of 20 to 30 milli-seconds. The total response times could have allowed an
- )
overpressure condition to exist with the rupture disc relieving the pressure
before the switches had time to respond.
2.3.5
Pressure Switch PS-2355
This pressure switch was the first alarm received during the event, indicating
6
W
_.
t
i
that there was a problem with the inner rupture disc.
The pressure switch has
+
a setpoint of 10 psig, it was found set at 10.I psig and properly aligned.
!
1
2.3.6
HPCI Turbine Thrust Bearina
A HPCI Thrust Bearing Wear Active Face alarm was received two seconds after
the rupture disc burst. The licensee performed a " bump test" to determine if
!
damage had been done to the bearing. This testing indicated excessive bearing
wear had occurred and tne licensee initiated removal.
Once the bearing was
removed, it was clear that damage had occurred and that the bearing needed to
!
be replaced. Visual inspection showed that the wear was very recent. The
j
turbine vendor representatives stated that they had seen this type of damage
^
before and that it was caused by water in the turbine.
2.3.7
Level Switch LS-2365. HPCI Turbine Inlet Drain Line Level
This level switch controls the air-operated HPCI turbine inlet drain bypass
valve (2301-31). The licensee stated that they had experienced recurring
!
problems with the alarm associated with this switch.
Several work requests
had been issued to troubleshoot and correct these problems, however, a
permanent solution had not been achieved.
Inspection after the event indicated that the return spring on the level
.
!
switch had failed which caused the switch to remain in the deenergized
position, thereby keeping the 2301-31 valve shut. With level switch (2365)
and the bypass valve (2301-31) inoperable, the licensee postulated that water
could have backed up into the turbine inlet if the inlet steam trap were
malfunctioning. Condensate could then be entrained in the steam flow when the
turbine stop and control valves were opened.
It could then have passed
j
i
through the turbine and out the exhaust, bursting the rupture discs.
!
The steam trap would have had to malfunction and not drain condensate from the
j
inlet line for this to happen. When instrument technicians inspected the
'
level switch, they attempted to test the float by adding water to the
condensing chamber
The technicians stated that the water they poured into
the condensing pot was sucked out the chamber bottom.
This indicated that the
steam drain, still on the line, was functioning properly drawing condensate
out of the line and to the main condenser. Technicians closed the isolation
valves to the pot and determined that the float was working.
A new level
switch was installed and the instrument returned to service.
2.3.8 Level Switch LS-2369. HPCI Turbine and Exhaust Drains Level
This level switch was inspected and the switch's float portion was not
functioning, due to a bent non-ferrous tube preventing movement of the
magnetic slug which activated the switch.
This prevented the solenoid-
operated drain to gland seal condenser valve (2301-32) from operating,
thereby, preventing the turbine drains from going to the gland seal condenser.
The only remaining path for turbine ' drains was through a steam trap and to the
torus.
This drain path has an approximately 50 foot rise the condensate must
be pushed through to get to the torus. This can only be accomplished when the
turbine is on the line and sufficient pressure is available to drive the
7
L
_
_
___
_
_
.
condensate up the line.
Therefore, prior to pressurizing the drain system by
opening the turbine throttle valve and driving condensate into the torus.
Condensate could have backed up into the turbine and caused damage to the
thrust bearing when steam was admitted to the turbine.
2.3.9 HPCI Room Secondary Containment Doors
Secondary Containment Doors open out from the Unit 1 HPCI room and consist of.
two sets of double doors. One door in each set is latched by a slide bar into
the floor, the other doors latch to them via a normal door latch. Both sets
of doors were blown open when the rupture disc burst. The inside doors were
accessible and were inspected shortly after the event. The opening force
damaged the HPCI room fire alarm bell and some piping insulation. The bar
latch on the bottom of one of the inside doors was forced out of its socket,
chipping the concrete.
The secondary containment is designed to minimize any ground level release of
radioactive material which might result from a serious accident. The design
basis accidents for the secondary containment are a LOCA and a refueling
accident. The Standby Gas Treatment System (SBGTS) removes radioactive
particulate matter and radioactive halogens with the efficiency required to
provide sufficient margin between expected offsite doses and 10 CFR 100
guidelines for the postulated LOCA or refueling accident.
The SBGTS is
designed to automatically start a single train or start both trains
simultaneously on various plant protective system initiation signals.
During the event, secondary containment doors were forced open by the pressure
surge and secondary containment was breached.
Release of radioactivity,
however, was not a concern since one train of SBGTS was in operation prior to
the burst, due to an ongoing SBGTS test, and the second train was manually
,
actuated subsequent to the event. Additionally, the potential radiological
consequences of a HPCI rupture disc release were compared with radiological
consequences analyzed in the UFSAR for a main steam line break outside
containment. Based on relative volumes released and on isolation times for
both systems. it is reasonable to assume that releases from a HPCI rupture
disc burst would not exceed 10 CFR 100 limits.
Secondary containment is designed to mitigate the consequences of a postulated
LOCA and the refueling accident.
It is not designed as a pressure boundary
against high energy line breaks outside the drywell.
For this reason, the
HPCI Room secondary containment doors are not rated to handle the postulated
accident pressures which are discussed in the High Energy Line Break (HELB)
analyses or the postulated pressures discussed in the Environmental
Qualification (EQ) analysis (6.3 psig in the HPCI Rooms).
Additionally, the HPCI room secondary containment doors do not create a path
for steam to travel to compartments which may not have been environmentally
qualified. Although the secondary containment doors are designed as dust
tight doors, they were not purchased or designed with pressure retention as a
primary design basis-.
Both doors were ordered to the requirements of S&L
Drawing B-657 in accordance with the Steel Door Institute (SDI) Manual SDI-
100.
8
%.
-
..
..
_ - _ _ _
.
.
__
,
.
~
J
The inner and outer secondary containment doors were manually closed by a
,
station shift supervisor at approximately 0930, re-establishing secondary
containment.
They were repaired and considered operable on June 10, 1993.
2.3.10
HPCI Room Fire Doors
The fire doors between Units 1 and 2 HPCI rooms were blown off their hinges
and into the Unit 2 HPCI room. One door impacted the Unit 2_RCIC test return
'
line to the contaminated condensate storage tank (CCST) which passes through
,
the Unit 2 HPCI. room. This caused the piping hanger, at the point of contact,
'
to move approximately two inches down the pipe; this appeared to be the' extent -
of damage due to the door's failure. The other door. landed on the floor-
causing no apparent damage.
Immediately after the event, a firewatch was
'
stationed in the HPCI rooms to compensate for the failed doors.
'
The licensee evaluated the effect of these doors acting as missiles and their
impact on Unit 2 HPCI operability. The evaluation concluded that the doors,
acting as projectiles, would not affect the licensee's ability to achieve safe
,
shutdown. When questioned about this, the engineering staff stated that the
.
HPCI system is not necessary for the affected unit to achieve a safe shutdown.
Their analysis evaluated Unit 1 HPCI room pressurization resulting from the
rupture disk burst but not from a postulated HELB in either room.
The Quad Cities HELB and EQ analyses discuss a postulated pressurization of
,
the HPCI rooms of approximately 6.3 psig (21 psia).
Based on the distance
'
that the fire doors were projected during the relatively lower rupture disc
pressurization, the inspectors asked if HELB pressurization had been
considered for replacement HPCI room fire doors.
The inspection team's
concern was that the doors acting as missiles during a HELB would
'
unnecessarily endanger the opposite unit's HPCI system.
The licensee has not
'
evaluated the effects of a HElB pressure surge on the fire doors. Another
missile hazard concerning the fire doors is the safety of personnel who may
_3
seek shelter behind the fire doors during future HPCI system testing.
lhese
concerns with the evaluations that were conducted regarding the fire doors
J
were discussed with the licensee.
2.4
Evaluation of Human Performance
l
Evaluations of human performance before, during, and after assessing the event-
identified several weaknesses, specifically in the areas of procedural
adequacy and system engineering. These topics are discussed in detail in the
following report sections. The licensee's efforts to assess and investigate
the event were also evaluated.
2.4.1 HPCI System Operations
The event occurred during the performance of Quad Cities Operational
Surveillance (QCOS) 2300-5, " Quarterly HPCI Pump Operability Test."
In
addition to quarterly testing, monthly testing is accomplished using QC05
2300-1, " Periodic HPCI Operability Test." Each HPCI system is operated at
least 16 times during the calendar year to support monthly and quarterly
surveillance testing. The procedures were reviewed and several individuals
9
W
.
.
l
involved with the test were interviewed to establish what occurred prior to,
l
!
during and following the Unit I rupture disc burst event.
'
When the test was being performed, four individuals involved with the testing
were in the Unit 1 HPCI room to perform local actions necessary to complete
QCOS 2300-5. ' A fifth individual, a radiation protection technician, entered
the Unit 1 HPCI room performing routine rounds. Through interviews it was
i
determined that a pre-job briefing had not been conducted.
The only official
j
j
instructions addressing potential personnel hazards were contained in the
precautions to procedure, QCOS 2300-5: " Ensure personnel near the HPCI turbine
are aware of the impending test and the potential for steam leaking around the
turbine." The monthly test, QCOS 2300-1, was also reviewed and the same
precaution was found in that procedure, however, no further personnel safety
i
precautions were found. Additionally, interviews indicated that no
announcement was made over the plant paging system prior to starting the HPCI
turbine.
Even though the personnel involved with the test were aware of the
impending test, because no announcement was made, they were surprised by the
turbine start. The uninvolved radiation technician was not made aware of the
impending test or the potential for steam leakage. The failure to ensure
personnel near the HPCI turbine were aware of the impending turbine start is a
failure to follow the precaution specified in the approved plant procedure and
i
is an example of an apparent violation of 10 CFR 50, Appendix B, Criterion V.
'
Through interviews it was determined that past operational practice resulted
in the HPCI room being evacuated during initial turbine startup. Recently,
i
this practice has not been consistently followed, in that one individual
l'
stated he had gotten used to the test and no longer waited outside the room.
Quad Cities Administrative Procedure (QAP) 300-2, Conduct of Shift Operations,
Step C.13.j states, " Briefings shall be conducted by cognizant personnel for
i
individuals involved in an evolution that is to be performed.
The detail of
the briefing is dependent on the degree of complexity, routineness, logistics,
i
or number of people involved." Step C.14.i states "A shift briefing shall be
1
conducted by the Shift Engineer, or his designee, for all Shift Operators,
i
after assuming the shift.
Evolutions involving many individuals, especially
l
from two or more departments or disciplines, may require large formal
I
briefings or preplanning sessions.
If the evolution is complex and involves
close coordination, the briefing session shall be coordinated by the Operating
,
Engineer or his designee and should include:
'
(1)
A review of the appropriate section of the procedure by key
parties.
(2)
Examination of each individual's specific involvement and
responsibility.
'
(3)
Discussion of expected results or performance; review of
limitations, hold points, emergency actions to be taken if
contingencies arise.
i
(4)
Assurance that everyone understands the interface and
communications required. "
10
1
k
'
1
The failure to follow approved procedures by not conducting a pre-test
briefing in accordance with QAP 300-2 is another example of an apparent
i
violation of 10 CFR 50, Appendix B, Criterion V.
Continuous communications were not maintained (and were not routinely
maintained) between the control room and the HPCI room while rolling the
turbine.
Personnel interviewed stated that intermittent communications were
common and were generally used to report completion of actions.
Procedural
!
review determined that no guidance was provided for handling communications
between the HPCI room and the main control room during testing.
If continuous
j
,
l
communications had been maintained, the control room operator may have been
able to reduce the steam released into the room by tripping the turbine rather
i
than waiting for the isolation valves to close.
l
QCOS 2300-5, Step 17, required the opening of the HPCI turbine steam supply
valve (2301-3), thereby pressurizing the turbine inlet line up to the turbine
stop valve (2301-17). Step 22 directed air-operated drain valves 2301-64 and
2301-65 to be opened for approximately 10 seconds.
These drain valves were
opened to remove moisture from the steam inlet line between the two main line
,
valves (2301-3 and 2301-17). When the valves were opened during this HPCI
-
test, the operator did not see steam coming from the sump after the required
,
10 seconds as nornally expected. He stated that he was on his way to inform
J
the control room of the lack of observable steam when the rupture occurred.
The use of a time limit (10 seconds) instead of the normal expected outcome
'
(steam issuing from the drain to the sump) to a procedural action step is an
example of an inadequacy in the test procedure.
Tnis inadequacy represents a
procedure inappropriate to the circumstances and is a further example of an
,
apparent violation of 10 CFR 50, Appendix B, Criterion V.
l
2.4.2 Failure to Recognize Sionificance of Rupture Disc and Level Switch
Problems
Rupture discs on Units 1 and 2 HPCI turbines were original equipment.
s
Maintenance records were reviewed; no record of corrective or preventive
maintenance (PM) being performed was found.
The rupture disc vendor manual
{
was also not available on-site after the event.
.i
After being contacted by the licensee and the inspection team, the rupture
l
disc vendor, Black, Sivalls & Bryson, Inc., supplied technical literature,
I
l
original invoice copies, and rupture disc drawings.
In a letter dated June
L
11, 1993, the vendor stated that the rupture discs' serviceable life is one
i
year under normal operating conditions.
Severe conditions such as elevated
j
temperature, pulsating pressure and corrosion may reduce the expected life.
1
The vendor further stated that the warranty covers a period of 12 months and
they did not recommend using the discs that had been in storage for the past
20 years.
HPCI steam drain level switches were also not included in the licensee's PM
1
' program.
Records indicated the level switches, associated valves, and traps
i
had experienced several corrective maintenance actions over the years.
During
l
interviews, personnel from operations, maintenance and engineering staffs
11
l-
'
j
.
j
_- _ _ _ _ _ - _ _ _ _ _ _ -
'
,
.
!
indicated that recurring alarm problems existed with the HPCI steam inlet
drain level high alarm on the main control board.
A review of the maintenance
history showed that since April 1986 through September 1992, eight separate
work requests were written concerning this problem.
DATE
WORK REO .i DESCRIPTION / DISPOSITION
04/17/86
Q49134
Alarm cleared and immediately came back
Reattached bracket on switch ran new wire
02/24/88
Q62996
Alarm comes up occasionally and resets
Canceled
05/16/88
Q63347
Alarm periodically comes up, alarm is valid
Repaired drain trap
08/21/88
Q67475
Alarm comes every 5 minutes
Inspected trap, everything OK
11/12/90
Q88199
Alarm kept coming up, reset
Canceled, duplicate of Q88174
05/01/91
Q88174
Alarm is coming up at least once per hour
Repaired trap
06/05/92
099386
HPCI trap sticks to cause high level alarm
Test failed, inspected trap, everything OK
09/18/92
Q01500
HPCI drain high level still alarming
Canceled, WR# 002610 written with expanded
scope.
A General Electric Service Information Letter (SIL) No. 531, HPCI and RCIC
Magnetrol Level Switches, was issued on february 7,1991. This SIL
recommended the installation of an improved switch in high temperature
applications.
This directly applied to the HPCI and RCIC system drain level
switches.
The licensee had received the SIL and the system engineer had a
copy, however, no action had been taken regarding the recommendations.
The Quad Cities Station High Pressure Coolant Injection (HPCI) System
Maintenance Basis Document, a contracted reliability-centered maintenance
(RCM) study, was performed and a final report issued on December 31, 1991.
The rupture discs were not included as a part of the analysis of individual
components.
In addition, fault analysis completed during the study did not
include a rupture disc burst scenario.
The RCH study, however, identified the Torus High Level, HPCI Turbine Inlet
Drain Pot High Level, and HPCI Turbine Exhaust Drain Pot High Level switches
as not being in the facility's preventative maintenance (PM) program.
Their
inclusion was recommended as part of the ECCS test instrumentation because of
the alarm and control functions preformed by all three switches.
Each switch
received a category I ranking in the cost benefit analysis performed on all
140 recommended RCM tasks.
RCM Task Implementation Sheets were filled out in
l
June 1992 recommending an 18 month PM frequency.
The sheets were reviewed by
I
the PM Coordinator and Maintenance Staff Superintendent in June and by the
HPCI system engineer in September and October 1992.
The system engineer
recommended that both drain level switches PM frequency be increased +o every
third refueling outage because the switches were not part of the ECCS test
instrumentation.
This was the only evaluation that was completed and appears
to be counter to the RCH recommendation. All three sheets were forwarded to
12
,
_ _ _ _ _ _ _ _ _ . _
r
.
i
.
!
Electrical Maintenance on December 4, 1992. No further action had been taken
on implementing the recommendations.
A Quad Cities Vulnerability Assessment was performed on the HPCI system and a
report issued in November, 1992.
This assessment was to " determine if any
!
vulnerabilities existed which could impact the system's ability to mitigate a
core damage sequence." During this assessment twenty SIL's were reviewed,
,
i
however, SIL 531 was not one of them.
The following documents were reviewed;
'
UFSAR, Design Basis Document, station procedures, P&ID's, Control and
!
Instrumentation Drawings, modification documents, NRC Generic Letters, General
Electric SIL's, work package documents, Instrumentation Maintenance Department
l
instrument data files, and Corporate input documents; however, the HPCI RCM
i
study was not included. Electronic databases, including the Nuclear Tracking
System, were used extensively during the assessment, however, none of the RCM
implementation tasks had been entered into a formal tracking system.
The
assessment's final report did not identify any problems with the HPCI rupture
l
discs, torus high level, or HPCI steam drain level switches.
t
An audit of license compliance was conducted by the Site Quality Verification
(SQV) organization in January, 1993. This audit noted that the RCM study
!
identified a failure to test the torus high level switch.
It stated that this
failure to test was a failure to meet Technical Specifications and was the
result of an inadequate HPCI RCM study review. This failure was cited by the
'
NRC as a violation of Technical Specifications in March,1993. Though the
'
specific finding was corrected, a comprehensive review of the RCM was not
initiated and the HPCI drain level switches, which were listed as the same
category recommendation and on the same page, were not identified.
l
lhrough documentstion and discussions with the Dresden HPCI system engineer,
the team verified that a PM schedule had been instituted for the Dresden HPCI
'
rupture discs and drain level switches. The discs have been inspected every
I
outage and the Unit 3 discs were changed during the last outage due to
mechanical damage (Dresden and Quad Cities HPCI systems are very similar).
,
'
Both system engineers indicated that communication between system engineers at
different sites was frequent and beneficial but that the rupture discs and
'
level switches had simply been overlooked.
The interview with the Quad Cities HPCI system engineer revealed that a
'
rupture disc inspection had been added as a PM item approximately two weeks
l
prior to the event.
This was a result of the licensee's engineering review of
and upgrading to ANSI /ASME OM-1-1981 for the 10 year inservice inspection
program revision. Code section 1.3.4.2 requires the replacement of
,
nonreclosing pressure relief devices every five years.
This review also
determined that an engineering evaluation might be needed for the rupture
discs as a component still in service beyond its service life.
j
A review of past industry HPCI system failures found several documented
i
rupture disc and drain level switch problems which had been apparently
i
overlooked by the licensee.
In addition, the licensee performed a status reW ew of all RCM tasks in
January 1993 noting that recommendations for the level switches were still
13
i
!
i
-
_.
-
_
_
~
.
4
-incomplete. To date, none of the 140 RCM tasks'were entered into the-
,
- licensee's formal tracking system.
The failure to initiate prompt corrective
action regarding the HPCI drain level switches is an apparent violation of 10
CFR-50, Appendix B, Criterion XVI.
.
2.4.3 System Enaineerina
The licensee's system engineering program was evaluated to determine its.
effectiveness in identifying and correcting problems associated with plant
systems. Several system engineers were interviewed by the inspectors and
,
their system notebooks and procedures governing the program were reviewed.
All engineers expressed satisfaction with their positions and liked the
'!
challenges associated with their jobs. This is an improvement over previous
L
inspections where dissatisfaction was a common concern. The engineers lack a
j
wide experience base which limited their integrated system knowledge. This
4
has been helped, to some degree, by a four or five week plant systems course
which all the engineers had attended.
!
'!
Through interviews, the inspectors determined that system engineers had good
i
general knowledge of their systems; however, they showed weaknesses.when
questioned about the system outside of their discipline specialty. The
engineers also had difficulty identifying the interface between systems and
!
prioriti.ing problems with their systems.
i
System nt ' books and system engineer inspection logs were reviewed and showed
improveme; t over previous findings; however, minor deficiencies remain. . One
concern was the replacing of system engineers with " cognizant" engineers for-
,
several small systems. Cognizant engineers' duties are more relaxed than
i
system engineers and the . system assigned to this status do not receive the
t
attention other systems do. The use of " cognizant" engineers was not.
-i
described in the Quad Cities Technical Procedure (QTP) 010-5, Station System
i
Engineer Program, which establishes the. requirements for program
implementation. Cognizant systems were established using Technical Staff
.;
Policy Letter (TS) 4, System Engineer Group Responsibilities, 'a lower level
document, describing the more relaxed requirements for these systems
Of
.
particular concern was the inclusion of the Standby Liquid Control system, a
safety-related system with associated technical specifications in the
'
cognizant system program.
All concerns regarding the system engineering program were discussed with
[
management who expressed interest in improving the current program.
,
2.4.4 Effectiveness of Licensee Response and Event Evaluation
r
. An initial HPCI area inspection was performed by the licensee immediately
i
after the event. The re-establishment of secondary containment was completed
'
and tha area was quarantined. An initial radiological survey was conducted
and the area was found to be slightly above normal operational contamination-
!
levels.
The resident inspectors were present in the control room to observe
operations personnel recovery from the event.
The event was correctly
classified and proper emergency notification were made.
Facility. management
14
!
I
-
.
-
1
'
J
.
.
,
oversight was present immediately after event initiation and throughout the -
i
recovery phase. Some confusion about the number of injured and which
!
hospitals could handle which people was evident; this did not delay the time-
it took the injured to receive proper treatment.
,
t
To evaluate the event the licensee assembled a multi-disciplined engineering
and technical team from the corporate: and Dresden station. offices, including
the Dresden HPCI system engineer. The team arrived onsite the day of the
event; developed a charter outlining maintenance activities, including. removal
of the rupture disk, internal inspection of the downstream check and stop
i
check valves; and evaluation of related instrument calibrations.
,
In the licensee's preliminary report summary, the licensee stated their
charter was restrictive and limited their investigation to determining the
event's root cause only.
It further stated that the investigation was not an
" attempt to evaluate personnel safety, emergency response capability, or any
of the other peripheral aspects of the event." A review of the original
charter, however, showed that initially, the team started to evaluate a
broader scope than stated in the report.
Further, the recommendations,
included in Appendix H of the report, show an evaluation of procedures,
personnel safety, and other concerns of a broader scope than stated in the
summary.
The evaluations completed in these broad topics, however, did not
!
appear to receive the same thorough review as those dealir.g with the mechanics
of the failure.
One finding in the preliminary report, the failure to pre-position the HPCI
test return valve (2301-10), demonstrates the insufficient review given some
portions of the investigation. The report stated that normally this valve is
placed in a preset position prior to rolling the turbine. This allows
approximately 10 minutes for the turbine to warm up and remove any condensate
from the lines. While the report stated the valve was not pre-positioned
during the test, thereby, allowing water in the turbine to be carried into the
exhaust line, operators' written staterents and operator interviews indicated
the valve had been pre-positioned.
It was also determined that pre-
,
positioning the 2301-10 valve only takes about one minute normally, and five
minutes at the most. This was supporte: by valve testing in 1992 when the
valves went from full close to open in approximately 66 seconds. Though this
was only one example and does not appea to be commonplace, it was a -
significant part of the event evaluaticr. that was incorrectly interpreted.
The restrictive charter did not ensure a thorough investigation was conducted
on all aspects of the event.
Furthermcre, it placed the burden of
identification and evaluation of issues associated with the event on the plant
staff, in addition to their normal work assignments, instead of an
independent, singularly focused team.
The licensee issued a signed preliminar report on June 17, 1993, pending-the
'
completion of several maintenance and testing actions planned for the Units 1
and 2 HPCI systems. The inspectors dis:ussed the list of recommendations
contained in the report with the licensse and were told that they intend to
treat the recommendations as proposed corrective actions until the final
report is issued.
15
i
-
.
i
2.4.5 Materiel Condition and Housekeepina of HPCI and RCIC Rooms
The event expectedly impacted the cleanliness and materiel condition of the
.!
HPCI rooms, however, all inspection team members noted that these conditions
- did not appreciably improve after restoration work was completed.
In fact,.
the trash, dirt, and debris found in the HPCI and RCIC rooms, particularly
under and around the turbires and pumps had been allowed to accumulate over an
!
extended period of time.
h addition to the housekeeping concerns and
i
standing water in areas of the rooms several equipment deficiencies were noted
that indicate a general poor state of materiel condition.
These deficiencies
1
included a drain valve missing both nuts on the packing gland follower, an
j
unsupported length of tubing on Unit 2 HPCI, cabling supported by duct tape,
r
and a leaking relief valve on the Unit 2 RCIC. The condition of these rooms
was shown to and discussed with the licensee who concurred with the findings
,
1
of the team.
i
i
2.5
Quality Assurance Deficiencies
I
The Site Quality Verification (SQV) group performed an audit, in January 1993,
l
of the licensee's compliance with Technical Specifications,10 CFR, and other
!
requirements and design basis documents, regarding the HPCI system. The
audit, in part, evaluated plans for improving availability / reliability of the
HPCI sy-tem. The evaluation questioned the HPCI system operability with steam
drain failures. Although the responses were correct for the specific question
stated, neither SQV nor the system engineers connected the ongoing _ drain float
switch problem with the potential for turbine damage.
SQV did identify the failure to implement Torus High Level Switch testing
during this same audit. The failure's cause was attributed to an inadequate
RCM study review. The audit finding resulted in action being taken regarding
the switch testing, however, no action was taken to perform a thorough review
of the RCH study.
'
2.6
Radioloaical Conseauences
l
The event's radiological consequences were minor with respect to exposure to
the public and the environment. Although secondary containment was breached,
s
no contamination was spread beyond the rooms immediately surrounding the Unit
1 HPCI room.
Radioactivity release was minimized by the turbine building
volume, relatively low volume of steam released, low steam radioactivity
content, and one train of the SBGTS being in operation prior to the burst.
Consequences to facility staff, in general, were the same as to the public.
j
However, five individuals were contaminated and required medical treatment.
Had the five individuals not been in the HPCI room at the time of the burst,
but standing outside the room, the inspectors believe the injuries would have
been much less depending on where the personnel were standing.
2.7
" Rupture" Mechanism
The normal HPCI turbine exhaust pressure is between 25-35 psig, the HPC] high
exhaust pressure turbine trip setpoint is 100 psig, and the HPCI rupture disc
16
_
.
-
.
.
.
b
setpoint'is 150 psig.
It is known that: water was present in the HPCI turbine
[
casing; the high exhaust pressure alarms were not activated; the disc burst
,
from a gaseous overpressure condition (not necessarily 150 psig); and that the
l
normal exhaust path was available.
The actual rupture mechanism, however, is
t
not clear.
It is postulated that the water in the turbine formed a slug upon
exiting the turbine, compressed the air / gases between the exhaust line and the
inner disc, and burst the discs.
This could have caused an overpressure
'
condition (150 psig) which the high exhaust pressure switches did not see due
l
to their relatively long response times. However, it is also possible that
the disc burst at less than 150 psig due to disc age and/or service
j
degradation. Further insights into the burst pressure could be demonstrated
!
by performing a burst test on the Unit 2 HPCI inner rupture disc, given its
j
similar operational and maintenance history. The licensee is evaluating the
possibility of conducting such a test, but has not determined how or where
such a test could be performed.
2.8
Conclusions
!
After completing the special inspection charter, the team was able to reach
[
specific conclusions relating to the event and system operations and
!
administration which are discussed in the following paragraphs.
j
2.8.1 Opposite Unit and RCIC Aonlicability
,
Unit 1 HPCI turbine rupture disc failures were directly applicable to the Unit
-
2 HPCI system.
Further, the event is also applicable to the Units 1 and 2
i
RCIC systems because all four turbines have the same type rupture disks from
'
the same vendor. These discs were all constructed from the same material and
in the same manner, however, the RCIC discs are only eight iilches in diameter.
'
All RCIC and HPCI rupture discs have been in service for over 20 years.
The licensee is currently planning to replace all rupture discs on Units 1 & 2
HPCI and RCIC. The discs will be placed in quarantine until an evaluation of
each one can be completed.
T'
iicensee also intends to perform a burst test
on the Unit 2 HPCI rupture discs in the near future.
i
,
The level switches (LS) 2365 and 2369 on the Unit 1 HPCI steam inlet and
.
exhaust drain systems were found to be malfunctioning. The licensee inspected
!
the Unit 2 level switches and discovered that they were also malfunctioning.
!
The Units 1 and 2 RCIC turbine inlet drains have a similarly designed system
!
and were inspected. The Unit 1 RCIC drain was operating properly, however,
the Unit 2 RCIC drain was not.
The root cause determinations for the Unit 2
HPCI and RCIC steam drain failures will be identified at a later date by the
l
licensee's staff.
2.8.2 Recommended Actions
The short-term recommended actions to be taken by the licensee, including
replacement of all HPCI and RCIC rupture discs, and testing and repair of all
steam drain level switches should sufficiently correct the pre-event material
condition, regardless of the actual rupture mechanism.
!
l
17
)
1
!
.-.
.
t
.-
!
The longer-term recommended actions have not been defined because a final
i
report of the_ event has not been prepared by the licensee.
Recommendations
j
have been made by the licensee's investigation team covering procedure
I
changes, preventive maintenance inclusions, inter-station information sharing
1
improvements, and other items. The licensee stated the recommendations would
1
be treated as corrective actions.
l
2.8.3 Test Controls
,
Test controls exhibited at the facility during the event were poor. There was
l
no pre-job briefing; non-involved personnel were in the HPCI room; inadequate
i
procedural precautions; paging system announcement was .not made prior to
starting the pump; and continuous communication was not used.
Implementation
of any one of the above safety practices may have prevented, or at least
minimized, injury to the personnel. The problem with test control at the
facility needs to be addressed in a thorough plant-wide manner.
1
2.8.4 Ineffective Corrective Action Proaram
i
Sufficient information was available well prior to the event regarding the
!
possibility of HPCI steam drain level switch failures. The drains' purpose
and function were not well understood nor was the potential impact on HPCI
system operability. Additionally, when the RCH study was completed the review
1
of proposed recommendations was not adequately reviewed. When SQV' identified
i
the failure to perform and adequate review, the specific failure was addressed
and corrected, however, an adequate RCM study review was not_ accomplished.
Additionally, none of the 140 RCM Implementation tasks has been entered into a
formal deficiency tracking system. This finding indicates a breakdown in the
l
control of the licensee's corrective action program on the part of facility
t
management by demonstrating an ineffectiveness to address and correct self-
!
identified deficiencies.
I
2.8.5 System Enaineerina
!
!
System engineer's while interested and enthusiastic,_ lacked detailed system
operations knowledge.
This was identified through interviews with the
,
engineers who, while generally familiar with their systems, showed weak
>
.
integrated system knowledge, cross-discipline knowledge, and problem
prioritization skills.
-
further, system engineering remaint an entry-level position with no clear
!
career progression program available to the engineers. Those interviewed
stated that as much as they liked the job, to get ahead you needed to go to
!
operations or maintenance.
There was no practice or incentive to return to
the system engineering program once you left.
Finally, system ownership remained a weak point, although the engineers knew
which systems belong to them and were familiar with work in progress.
It was
i
commonly found that when corrective actions or problems were assigned to other
'
work groups, the system engineers lost track of them until the issue was
returned to them. This is specifically what happened with the recommendation
to include the HPCI steam drains in the licensee's PM program.
18
1
c
- -
.
. . . . .
.
- _.
_
.
- _.
_
l
'
i
-
I
L
3.0
Exit Interview
l
!
- Members of the team and regional management met with licensee corporate and
j
facility representatives on July 9,1993, and summarized the purpose and
findings of the inspection. The team discussed the results of the inspection
!
and the licensee acknowledged the inspection findings. .The team also
discussed the inspection report's likely informational content with regard to
(
documents or processes reviewed by the inspector during the inspection. The
i
licensee did not identify any such docunent/ processes as proprietary. A re-
1
exit was conducted via telephone on August 12, 1993 at which time the apparent
]
violations were discussed between regional management and the licensee staff.
l
!
t
i
1
!
l
l
i
!
!
!
!
l
I
!
I
1
,
I
i
19
1
Y
-
- :n
.
RUPTURE DISCS 2
E *Er$oEE
2301-3
,
l
p y""p 4
pq
2365g
"N
2355
,
10 NG
mnaint
N
$ ALE
'
{
)
,
,
J
i
i TURBINE
C>k}--.D <]2301-65 )g
DRAIN
h 23688
O
2301-31
lhl
)
I
l
2301-28
>N
)
2368A
i
To Wn
imrsta j
-
DRAIN
v
Condenser
2301-32<;J 2369
2301-4
2301-5.
6<]
To Gland Seal
3
i
mo
Condenser
LJ
a
/.
~
/ 2301-45
1
2301-74
v
g-
SUPPRESSION
POOL
Simplified HPCI System Drawing.
- -
- - .
.
- -.- . . .
.
.
.
.
--
- .
- . - . -
---