ML20043C234
| ML20043C234 | |
| Person / Time | |
|---|---|
| Site: | Limerick  |
| Issue date: | 05/24/1990 |
| From: | Kaufer B NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| Shared Package | |
| ML20043B529 | List: |
| References | |
| TASK-AE, TASK-T90-07, TASK-T90-7 AEOD-T90-07, AEOD-T90-7, NUDOCS 9006040261 | |
| Download: ML20043C234 (9) | |
Text
.
j W>
l AEOD TECHNICAL REVIEW REPORT l'
UNIT:'
Limerick 2-TR REPORT NO.:
AEOD/T90-06 DOCKET NO.:
50-353 DATE:
May 24, 1990
' LICENSEE:
Philadelphia Electric EVALUATOR / CONTACT: B. Kaufer NSSS/AE:
GeneraliElectric/Bechtel
SUBJECT:
REVERSED SENSING LINES CONNECTIONS EVENT DATE:
January
- 9 90
~
SUMMARY
. This study was initiated to review the safety significance of an event at i
Limerick Unit 2, LER 90-002, on January 5,1990. This event involved
~c _'
. reversed connections of sensing lines to a containment hydrogen / oxygen analyzer. Hineteer additional events involving this issue were revealed from searches on NUDOCS and SCSS databases.
NRC actions regarding this 1
. issue include an AE00 engineering evaluation report and Information Notices
-issued in 1984 and 1985 (Ref.:1 - 5). Most of the events were plant specific.
In conclusion, this review determined that the: frequency of occurrences is relatively small and the consequences of the events resulted
- in limited safety significance.- Therefore, no further AE0D action is recomended.
)
+ '
i 1.
INTRODUCTION
' '4 y:'
On January 5,:1990, the Limerick Unit 2 containment Hydrogen /0xygen (H2/02) 1 analyzer was declared inoperable due to a high concentration indication
- during nitrogen inerting of the containment.
Following troubleshooting, it was determined that the root cause of this event was due to reversed connections of the analyzer sensing lines, due to mislabeling of the analyzer ports.-
The analyzers were supplied by Comsip-Delphi (Model K).
Installation of the analyzer was accomplished by connecting the labeled inlet / outlet ports
.to the desired plant lines.- Because of the' reversed connections, the normally closed sample. point isolation valve was in the line exhausting to containment while the sample point line had only a ca). This resulted in
=
an indeterminate percentage of-sample recirculation tirough the analyzer rather than normal exhaust to the suppression pool or drywell.- The licensee determined that the aralyzer had been inoperable since its installation in 1989.
Pre-operational testing of the analyzer did not identify the improper j
configuration of the system. Thethreeotheranalyzersintheplant(two i
on each unit) were. verified to be properly labeled and connected.
This review was performed to determine if other events involving reversed sensing lines had occurred and to evaluate consequences of these events as to frequency, safety, and root cause.
9006040261 900524 PDR ADOCK 05000353 j
P PDR
QS
,i4
-2 H
+
- 2.
DISCUSSION
, TheNuclearDocumentControlSystem(NUDOCS)andtheSequenceCodingand
' Search System-(SCSS) databases were used to search for similar events. The results of this search identified 20 events since 1984, including the i
Limerick event, which are outlined in the attached table. NUDOCS also found
. that four Information Notices have been issued concerning similar problems.
Additionally, AE0D Engineering Evaluation Report No. AEOD/E408, entitled, l
" Reversed Differential Pressure Instrument Sensing Lines," was issued on April 13, 1984, regarding this subject.
This review was done to determine if the continuing occurrences of reversed c
connections showed an increasing trend and safety. significance on plant performance. This review determined _that the frequency of occurrences is relatively small and the consequences of the events resulted in limited safety significance. Therefore, no further AE0D action is recommended.
Most of the events involved an error in which tubing or sensing lines to i
instruments were reversed.
The effects of reversing these lines on plant operation were varied based upon the component and system involved, its intended function, and the time 1eriod before discuvery.
Reductions in plant safety margins caused by tiese events were partially negated by the availability of redundant equipment or other safety-related systems.
The use of inadequate design documents, deficient startup procedures, and weak verification methods are considered the primary causes for these events. -Many of these events were the result of personnel using incorrect drawings to install or modify system components. Normally, this type of error would be discovered during the verification process or-the pre-o>erational testing phase of work. A review of the events clearly showed tiat, in most cases, a-m ' *es used for testing modified components did not detect the errors i 4 wr*
.m. Additionally, the quality control methods used to verify cet sci irm u Mns were ineffective.
3.
FINDINGS AND > Ni.N0h FINDINGS As detailed in the attached table, the events identified involved varying
. instruments on varying plant systems.
In general, all these events were
' of limited safety significance to the plant involved.
In all but one of the events, the sensing line errors had been undetected for some period of time.
Most of the events were caused by inadequate installation or modification-documents and inadequate pre-operational or startup procedures. Events were independent of reactor type (unlike events reviewed in the previous study in
~ hich BWRs were mostly involved).
Review of events based on the amount-of w
occurrences and dates did not detect any significant trends.
i ft
,1: ?
The errors found in 15 of the events involved instances where sensing lines to a single' component were reversed (ie., hiand lo 3ressure. legs, or.
. incorrectly connected ports on a solenoid valve). T1e remaining 5 events, related to errors made by incorrectly connecting sensing lines to a corresponding or backup component on the sene system.
Two of the events' wore AMir,itely attributed to original construction, and
- two others' were susnged en heir:g related to original construction. Of the remaining-16 events, Wee usociated with modifications while three involved maintenance bercrs.
i
.CONCLUS!DNS I
n, 9
I Based on_the review, events involving reversed tubing continue to occur about three times per year. Since 1984, the events have, unlike those previously reported, occurred to varying' components on different systems.,
--The root causes of these events remains basically the same, inadequate design documents, deficient startup procedures, and weak verification methods.
InoNier'toreducethenumberofeventsoccurring, licensees.needtoensure; adequate attention to the details of installation, modification, and maintenance work performed on these types of components. Additionally, i
proper quality control of work and better post-modification or functional i
testing will further ensure a reduction.
_In conclusion, this review determined that the frequency of occurrences is i
relatively small and the consequences of the events resulted in limited safety significance. Therefore, no further AE0D action-is recomended, i
i
s
~
f" '. >,
- l
,=:
4 g n.
n
.9 REFERENCES
.1.
.AEOD ENGINEERING EVALUATION REPORT NO. AEOD/E408, REVERSED DIFFERENTIAL PRESSURE INSTRUMENT-SENSING LINES, April 13, 1984.
2.
'IE INFORMATION NOTICE NO. 84-45, REVERSED DIFFERENTIAL PP. ESSURE INSTRUMENT SENSING LINES, June 11, 1984.
3.
IE INFORMATION NOTICE NO 85-11,-IMPROPER INSTALLATION AND
' TESTING OF DIFFERENTIAL PRESSURE TRANSMITTERS, January 11,'
1980.
4.
IE INFORMATION NOTICE NO. 85.23, INADEQUATE SURVEILLANCE AND POSTMAINTENANCE AND POSTMODIFICATION SYSTEM TESTING, March 22, 1985.
-5.
'IE INFORMATION NOTICE NO. 85-75 IMPROPERLY INSTALLED
' INSTRUMENTATION,' INADEQUATE QUALITY CONTROL AND INADEQUATE POSTMODIFICATION TESTING, August 30, 1985.
a S
4 5
3 Il
.-i-
?
1 I
l q
t I
y, '
n TABLE 1 REVERSED SENSING LINES CONNECTIONS - EVENTS 1.
HATCH 2 LER 84-003 01/15/84 EVENT:
COLD SHUTDOWN Operations lowering reactor water level w/RWCU.
PCIS GROUP 5 isoletion signai occurred.
RWCU outboard isolation valve did not close.
CAllSE:
RWCU DUMP FLOW TRANSMITTER.
INSTALLED INCORRECTLY DURING DESIGN CHANGE.
During installation of new transmitter personnel inadvertently connected sensing lines backwards.
2.
SURRY 1 LER 84-007 03/28/84 EVENT:
100% POWER Post maintenance testing on Unit ? determined that main feedwater regulating (FRG) valves would not have tripped closed on feedwater isolation signal.
When Unit I was checked same problem was found.
CAUSE:
FRGs 3-WAY S0LEN0ID VALVES.
INSTRUMENT AIR TUBING FROM THE FRGs ACTUATOR HAD BEEN CONNECTED TO THE WRONG 50LENDID PORT.
Work performed during outage without approved procedure or work request.
3.
WPPSS 2 LER 84-106 09/27/84 EVENT:
4% POWER Increasing flow in jet pump caused recorder to indicate onscale and be declared inoperab h.
CAUSE:
REACTOR VESSEL FUEL ZONE LEVEL TRANSMITTER.
INCORRECT TUBING INSTALLATI0H.
Instrument tubing had been modified base on incorrect instructions in modification package.
4.
MCGUIRE 1 LER 84-030 11/01/84 EVENT:
HOT SHUTDOWN Upper Head Isolation (UHI) valves f ailed to close automatically as required.
CAUSE:
UPPER HEAD ISOLATION ACCUMULATOR LEVEL TRANSMITTERS.
INSTRUMENT IMPULSE LINES TO THE HI PRESS AND L0 PRESS PORTS WERE CROSSED.
~
Error was made when original switches were replaced.
i S.
LASALLE 2 LER 85-029 06/10/85 EVENT:
COLD SHUTDOWN During a field walkdown it was discovered that piping to switches had been installed backwards.
Inoperable switches resulted in Unit 2 Div. 1 ECCS being inoperable from when they had been replaced.
CAUSE:
REACTOR VESSEL LEVEL SWITCHES.
PIPING TO INSTRUMENTS HAD BEEN INSTALLED IN REVERSE OF SYSTEM DESIGN.
Incorrect piping was due to field change which was initiated on design drawings but did not get transferred to the contractor installation drawings.
6.
LASALLE 2 LER 85-031 06/22/85 EVENT:
COLD SHUTDOWN Problem with switches originally noted during testing, but piping routing error of discovered later during walkdown.
CAUSE:
RHR SUCTION HIGH FLOW ISOLATION SWITCHES.
PIPING T0 INSTRUMENTS HAD BEEN INSTALLED IN REVERSE OF SYSTEM DESIGN.
Incorrect piping was due to field change which was initiated on design drawings but did not get transferred to the contractor installation drawings.
7.
LASALLE I LER 85-053 07/17/85 EVENT:
COLD SHUTDOWN Similar to event described above but discovered differently.
CAUSE:
RHR SUCTION HIGH FLOW ISOLATION SWITCHES.
PIPING TO INSTRUMENTS HAD BEEN INSTALLED IN REVERSE OF SYSTEM DESIGN.
Caused by previously rejected drawing change request which was rejected for unrelated reasons.
8.
PALO VERDE 1 LER 86-021 02/06/86 EVENT:
2% STARTUP Incorrect piping to transmitters discovered during review being performed for new modification.
CAUSE:
LOW PRESSURE TRANSMITTERS TO PNEUMATICALLY CONTROLLED 'A' TRAIN ADVs.
TRANSMITTERS WERE FOUND TO BE CONTROLLING THE BACKUP HITROGEN SUPPLY TO 'B' TRAIN.
As-built drawings used to verify were incorrect.
i 1
M,
n 9.
ROBINSON 2 LER 86-009 03/22/86 EVENT:
13.5% POWER Incorrectly installed sensing lines along with other complications caused level in 'B' SG to swell resulting in a turbine trip and subsequent reactor trip.
CAUSE:
'B' SG STEAN FLOW TRANSMITTERS.
SENSillG LINES WERE CROSSED.
Lines were inadvertently crossed when new tubing was being installed.
- 10. BRUNSWICK 2 LER 86-019 07/22/86 EVENT:
100% POWER Evaluation of the axial flux profile revealed TIP traces for path B-10 did not correlate as expect (d with traces using channels A, C, and D.
CAUSE:
TIP SYSTEM TUBING PATHWAYS.
TUBING WAS IMPROPERLY CONNECTED.
Cause attributed to misidentification of tubing which occurred following temporary removal for preventive maintenance during 1984 refuel.
- 11. HOPE CREEK LER 86-056 08/08/86 EVENT:
3% STARTUP Both reactor building to torus vacuum breakers were 2
determined to be inoperable. Due to exceeding LCO unit shutdown was connenced.
CAUSE:
DIFFERENTIAL PPESSURE TRANSMITTERS (DPT) CONTROLLING
~
INBOARD AIR-0PERATED BUTTERFLY VALVES.
HIGH SIDE AND LOW SIDE OF DPT WERE REVERSED.
=
Cause attributed to design / installation deficiencies.
- 12. CATAWBA 2 LER 86-045 10/11/86 EVENT:
COLD SHUTDOWN During), testing of turbine driven auxiliary feedwater pump (CAPT pump did not start because valve controlling steam flow did not open.
2 CAUSE:
STEAM FLOW CONTROL VALVE TO CAPT.
S0LEN0ID VALVE WHICH OPENS CONTROL VALVE WAS INCORRECTLY TUBED.
Cause attributed to incorrect installation due to incorrect design drawing.
a
=
6,m
.13; CATAWBA 1-LER 87-012 03/11/87; h
3 FVENT:
' HOT SHUTDOWN I ',
- During PORY testing it was determined that 3? valves were-unable to perform using backup nitrogen and were declared '
j inoperable.
CAUSE:
PRESSURIZER PORVs (3).
INSTRUMENT AIR SUPPLIES FOR THESE VALVES WERE INCORRECTLY CONNECItu.
Due to design oeficiencies from original installation.
Resulted in train ' A' emergency nitrogen supply to train
'B' and vice versa.
1
- 14. MCGUIRE 1 LER 87-007 03/17/87 EVENT:
100% POWER During testing of Unit 2 DG Halon system which verifies-operability of the manual-pneumatic actuator, the main bank
- l failed, due to. incorrect actuation tubing connections, j
When Unit I was checked the same problem was found.
CAUSE:
DG HALON SYSTEM MANUAL-PNEUMATIC ACTUATION.
ACTUATION TUBING TO HALON PILOT CYLINDER ON MAIN AND RESERVE BANL5 WA5 REVERSED.
j Cause attributed to deficient maintenance procedures and insufficient' vendor information.
' l i
- 15. PEACH BOTTOM 2 LER 87-008 05/29/87 j
l EVENT:
-COLD SHUTDOWN During investigation of suspected low flow through the 'B'
)
channel of the control room ventilation radiation nonitoring system, it was discovered that incorrect piping configuration existed, involving the sample lines-to i
several solenoid valves.
CAUSE:
CONTROL ROOM VENTILATION MONITORING SYSTEM SOLEN 0ID VALVES.
SAMPLE LINES FROM THE NORMAL VENTILATION DUCT AND THE i
EMERGENCY-VENTILATION DUCT TO SOLEN 0ID VALVES WERE REVERSED.
Cause attributed to an original installation error.
1 x
- 16. MCGUIRE 2 LER 87-010 07/31/87 j
EVENT:
100% POWER While replacing the pneumatic module in the upper personnel I
air lock (PAL), it was determined that tubing was i
incorrectly connected, which resulted in only the inner i
door seals being tested.
n.
t CAUSE:
PERSONNEL AIR LOCK AUTOMATIC LEAK RATE MONITOR.
INTERFACE TUBING BETWEEN PNEUMATIC MODULE AND PERMANENT TUBING WAS INCORRECTLY CONNECTED.
Cause attributed to design deficiencies including insufficient information.
17.
BEAVER VALLEY 2 LER 87-026 10/08/87 EVENT:
55% POWER During main feedwater pump testing, SG 'B' feed pump recirculation valve opened followed the SG
'A' valve.
Complications subsequently led to a reactor trip.
CJ.USE:
'B' MAIN FEED PUMP SUCTION FLOW INDICATING SWITCHES.
PIPING TO SWITCHES WAS REVERSED.
Cause attributed to personnel e, ror during walkdown following installation of switches.
18.
BRUNSWICK 1 LER 88-014 06/06/88 EVENT:
100% POWER While performing special testing HPCI was determined to be inoperable due to reversed high and low pressure leg piping.
CAUSE:
HPCI STEAM LINE HIGH FLOW INSTRUMENTS.
HIGH AND LOW PRESSURE LEG PIPING TO INSTRUMENTS WERE REVERSED.
Cause attributed to design drawing deficier.cies.
19.
FERMI 2 LER 89-030 12/08/89 EVENT:
2% STARTUP While increasing pressure Channel 'C' trip unit was noticed to be off scale when the other 3 units were back on scale.
EVENT:
REACTOR VESSEL LEVEL 2 TRIP UNIT TRANSMITTER.
HlGH AND LN PRESSURE SENSING LINES WERE INSTALLED BACKWARDS.
Cause attr cuted to error made during replacement of transmitter and inadequate verification.
20.
LIMERICK 2 LER 90-002 01/08/90 EVENT:
100% POWER Durinn nitrogen inerting, indications of 02 and H2 concentrations were higher than expected, and subsequertly declared H2/02 analyzer inoperable.
CAUSE:
H2/02 ANALYZER.
PORTS ON SAMPLE SKID FOR SAMPLE LINE AND EXHAUST TO CONTAINMENT WERE REVERSED.
Cause was due to an error in labeling of connection ports by vendor.