ML20059M722
ML20059M722 | |
Person / Time | |
---|---|
Site: | Diablo Canyon ![]() |
Issue date: | 08/17/1993 |
From: | AFFILIATION NOT ASSIGNED |
To: | |
References | |
OLA-2-I-MFP-014, OLA-2-I-MFP-14, NUDOCS 9311190265 | |
Download: ML20059M722 (36) | |
Text
- lff" V. M tInff f y - .
6D -2 75 323 OLA A gjp135 Degag Mr & \
h 2 - NM~ NCR DC1-93-EM-N010, D gly 28, g/f7pa 1n3 j
MANAGEMENT
SUMMARY
i l February 5, 1993: Unit 1 was ramped to 46% power and'an<
l unusual event declared when a ground alarm was'ieceived'dn -
l Circulating Wa,tpQggdCWP) 1-1. While the ground alarm was l in, smoke was reported in the Unit 1 12kV switchgear room.
The California Department of Forestry -(CDF) was called to l assist in fighting the potential fire. The smoke lasted less j than 10 minutes, there were no perscnnel injuries, and no l equipment was damaged. The CWP 1-1, phase-B, cable was found to have failed. Note: The CWP is not a tech spec item.
l During the power reduction,.the digital rod position indications (DRPI) did not' follow the control rod indicators as control rods were inserted into the core. Control Bank "D" was declared inocerable.
An Immediate (less than one-hour) emergency report was made to the NRC in accordance with 10 CFR 50.72 (a) (1) (i) . ,
A twenty-four hour Unusual Event summary. report was made to San Luis Obispo County in accordance with the requirements of ;
NUREG-0654, Revision 1, Appendix 1.
The TRG reconvened on 3/19/93, and agreed that the 3/12/93 similar event on CWP 1-2 would be included in the scope of this NCR..
March 12, 1993: Unit'1 was ramped to 50%' power'when a ground alarm was received on Circulating Water Pump (CWP) 1-2.
While the ground alarm was in, smoke was reported in the Unit 1 -t2kM_ s' itchgear room. The smoke lasted less than 10: __
minutes and there were no personnel injuries. No _quipment in the room was damaged. CWP 1-2, phase-C (of circuit "B"), I cable was determined to have failed.
This draft, dated July 28, 1993, contains minutes from the !
July 21, 1993, TRG meeting. The TRG will reconvene on j September 14 to review the IPRT report and determine final !
root cause and corrective actions.
Root Cause: TBD Contributory Cause: TBD Corrective Actions to Prevent Recurrence: TBD l I
93ncrwp\93EMN010.KWR Page 1 of 36 9311190265 930817 e
$DR ADOCK0500g5 d3
i t ,. 1 F
~
NCR1DCl-93-EM-N010,..D3 y'[
July 28', 1993
.e i
i NCR DC1-93-EM-N010 CWP 1-1 and CWP 1-2 CABLE FAILURES I. Plant Conditions. :
Unit 1 was in Mode 11 (Power '. Operation) ~at 100%. power at the time of occurrence for both events.:
II. D_escriotion of Event A. Summary:
Event 1;. February. 5,31993:- Unit 11 was'rampedfto_
46% power and an unusual event:declaredj.Nhen.'a ground alarm was received on CirculatingfWater Pump ,
(CWP) 1-1. -While the ground alarm lwas'in,ysmoke' .
'i was reported.in the Unit i 12kV switchgear room.,
The Cal'ifornia ' Department. of Forestry.L(CDF) was ~ f called to assist in-fighting theLpotential1 fire. <
The smoke lasted less-than 10 minutes and.there- 'i were no. personnel injuries.'No equipmentzin'the room was damaged.. CWP11-1,' phase-B,fcable.was' determined to have. failed.
A one-hour, emergency' report wasjmade'to the.NRC in-accordance with 10 ' CFR = 50.72 (a) (1)l(i) ~.. .,
>t
~
' I A twenty-four hour Unusual Event summary report lwas hg 1
{
made to San Luis Obispo. County lin'accordance.with '
dd j the requirements of INUREG-0654,:. R e v i s i o n . 1~,.- ,
{i; m i Appendix 1.
.@G t ;
11 g
I hs>>
i 6I During the power reduction,.the digital rod: ..
position indications .(DRPI) . did not. follow;the j {
N control rod indicators as' control'rodsawere; y, ec= y ;
- g i ,; inserted into the core'. Control Bank "D" wam j iy
- 5# ;,xa1 _
declared inoperable resulting.in entry intosa limiting condition of o'peration'in'accordance with i .y {j . Tech Spec 3.1'.3.1. ;
4 '
N : b i t
% ' y Event 2; March 12, 1993: Units 1'was rampedLto 50%- J 2 power and.when a' ground' alarm:wasfreceived on
%' i ! G L *1 Circulating Water Pump (CWP) 1-2. .. While:the ground, , ,
i ~T alarm was in, smoke was reported;in..the)Unitil?l2kV
,; , ' I r 'E % switchgear room. The smoke lasted less thane 10' ei ;m ,p_ minutes and there were no= personnel injuries.'No-.
5g z ' i :. : s {a equipment in the room was damaged. CWP.1'2,-phase t 5s E#;# C, cable was determined to have L failed.
93ncrwp\93EMN010.KWR Page 2 of 36 y
.)
l
,_ .. a
l i0' NCR DC1-93-EM-N010,-D3 l l July'28, 1993 l
l B.
Background:
l Event 1; February 5, 1993: Emergency Procedure G-1 l
- " Accident Classification-and Emergency Plan l
Activation", requires a NUE be declared when assistance of an of f-site agency (i.e. CDF)' is-requested.
NUREG-0654, Revision 1, Appendix 1, requires a written summary of the event be submitted to San Luis Obispo County government by the end of the first work day following the declaration of any emergency class.
Tech Spec 3.1.3.1 states; "All. full-length shutdown and control rods shall be OPERABLE and' positioned l
within plus or minus 12 steps (indicated. posit.on) of their group demand position." Control Bank "D" was declared inoperable. Tech Spec 3.1.3.1 requires restoration of Bank "D" to operable status within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> or place the Unit in at-least hot shutdown mode within the following 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.
Cable Construction The 12kV cables are rated for up to 15kV
_ applications. The cables are a shielded single-conductor cable design, constructed in concentric layers. There are four cable layers of interest:
(1) the insulation layer, which is approxinately 220 mils of black EPR; (2) the tinned copper shield tape, which acts to equalize tha.electricet stress; e~
(3) the wax coated synthetic binder tape, which holds the copper shield against the cable, thereby allowing the outer jacket to be extruded over the cable during manufacture; and (4) the outside layer, which is a neoprene jacket. The function of the outer neoprene jacket is to protect'the cable during installation (impact and abrasion resistance) and to.act as a physical barrier between the outside environment and the shield once the cable is installed. The cable jackets are not bonded to the cable shield or insulation. layers.
Duct Bank Construction The 12kV motor-driven' circulating water pump (CWP) feeders consist of six separate single-conductor cables (two for each phase), routed as two separate 93ncrwp\93EMN010.KWR Page '3 of 36
NCR DC1-93-EM-N010, D3 ,
July 28, 1993 three phase circuits in two single side-by-side conduits (two circuits).
The subject cables are routed in two' separate sets of duct bank conduits, one for each unit, between the turbine building and the intake structure.
Concrete vaults are located at various intervals to serve as pull boxes for the circuits. These duct bank conduits are directly beried in sand and are covered for their entire length by a six-inch thick concrete cap. The duct bank conduits include 12kV, 4kV, and 480V power cables, 120V ac control cables, 125V de control cables, and instrument cables.
The pull boxes immediately outside of the turbine building have drains, which are routed to common sump manholes for Units 1 and 2. These manholes are equipped with automatic submersible Class II sump pumps.
The Unit 1 and Unit 2 trenches are similar, except that the Unit i duct bank rises to cross over the circulating water discharge tunnel and then slopes downhill towards the intake structure. This design makes the Unit 1 section of cable conduits near-the turbine building susceptible to submergence if the
_ pull box sump pumps are not functional and if the water within the pull boxes rises above the conduit openings.
C_able Testing In accordance with applicable industry standards, DCPP performs direct-current, high-potential testing (hi-pot testing) as a maintenance activity each refueling outage. Normal practice for hi-pot testing of motors is to hi-pot the motors from the switchgear end of the circuit, through associated cables and motor terminations.- The 12kV motors, due to hi-pot test equipment limitations, are'hi-pot tested locally at the motor with the cables disconnected. The cables are hi-pot tested separately from the motor. The maximum voltage level used for the in-service 12kV cables is 30 kV dc.
C. Event
Description:
Event 1 93ncrwp\93EMN010.KWR Page 4 of 36
^
t l '
i i
jw ' NCR DC1-93-EM-N010, D3 July 28, 1993 l
On February 5, 1993, Unit 1 was ramped down from j 100 to 46 percent power due to a ground current l alarm for CWP 1-1. While the ground alarm annunciated, smoke was reported in the 12kV i
l switchgear room due to ground resistor 1 bank heating of accumulated dust, as is expected during. ground-fault conditions. An Unusual Event was declared at l
2156 PST due to a precautionary assistance request.
to an offsite agency. An immediate emergency-report was made to report the declaration of an Unusual Event in accordance'with 10 CFR-50.72 (a) (1) (i) .
Investigation determined that the ground was located on one cable between the first pull box outside the turbine building and the next pull box at the discharge structure elevation. When the l
cables were removed, the neoprene jacket was found to be separating from all three cables of the circuit for a distance of approximately 200 feet.
The faulted portion and a similar length of the-other two cables for this circuit were removed, the water was pumped out of the pull box,t a mandrel was passed through the conduit several times to remove standing water, and new cables were installed.
When the new cable sections were spliced in, no
_ visual degradation was found at the splice locations. Cable samples were sent to Okonite, the PG&E TES laboratory, and Altran Materials for comprehensive examination and testing. Interim progress reports indicate that the outer neoprene jacket had been chemically degraded and the copper shield shows evidence of corrosion. However, electrical properties of the cable are acceptable, no evidence of manufacturing defects has been identified, and no indication of Jul installation problem or abnormal operating conditions can be found. The root cause investigation for this event is in progress.
~
93ncrwp\93EMN010.KWR Page 5 of 36 *
, s p-NCR DC1- 9 3 - EM-N010', D3) dN '
1 July >28, 1993_ <
! )
, while removing the: failed CWP 1-1 cable from7the-conduit, water was introduced into the~ conduit:from
, the discharge structure; elevation. pull box in order? -
to lubricate the" cables for removal. - No water came' out the' pull. box at'the turbine building end.of the:
conduit run. Investigation determined that.the- -
six-inch' acrylonitrile-butadiene styrene'(ABS) -
conduit was^ broken,fand-that-_this>damageLhad-
- resulted from the i d tial attempts-toJremove:the' ,
cable. '
Conditions prior to' event =, STP M-16N had just'been=
completed-(test'switchfplacedfin_" reset"). ,
I
~ TIME DESCRIPTION-2045.52 12kV STUP OR AUX X'FMR 1-1 GRD OC CWP 1-2.FDR GRD and,CWP 1-1 FDR GRD 2045.56 CWP.1-21FDR GRD out 2047.36 SMOKE DETECTED ~ (alarm) .
2048 SFM orders.CO to1 ramp unitrat_100 MW/ min. ,
2050.08 LTB ACTUATION jT-
.2051.22 BEARING PEDESTAL CURRENT LOOP ALARM ,
2051 SEM orders ramp. rate increased;to 200
~
MW/ min 2051.54 ROD POS'IND URGENT, ROD POSSIND ROD-BOTTOM, ROD POS-IND RODS AT BOTTOM. ,
2052.42- At approx '60% turbine load, SEM orders CWP 1-1 shut ldown. 'CWP 1-l' e FDR GRD cleared.
2053.35 ROD POS IND NON-URGENT-Between 2052 and 2054, SEM authorizes DRPI - '
placed11n " Data ~AiOnly". . Control; Bank':D position-indication returns, but Generale Warning and Rod' Bottom lights on all'of DRPI come on. CO pulltrods OUT~and notices DRPI ,
showing rods ~ going l.IN. When;tried again,'no R rod motion was.noted'on DRPI although" rod-step i
\
93ncrwp\93EMN010.KWR Page 6 of 36 -)
y i
4.
+ rr e -p..,4 er r-n , .m -- , .-. e p , N' ----A..
,..n, ,<,s , + . - , . - - -,c.. - a
- NCR DC1-93-EM-N010, D3 July 28, 1993 counters operated properly. A few minutes later, DRPI was placed in A + B, and all DRPI indications returned to normal, no alarms.
2103 CDF called out to assist with potential fire in the 12kV Switchgear room.
Smoke is reported and IFO requesting CDF backup.
2106 Unusual Event (NUE) was declared based on CDF response.
2110 SLO County Sheriff notified of the UE.
2114 California Office of Emergency Services notified of the UE.
2114 Fire brigade reports 12kV SWGR room is being ventilated and is clearing.
" Fire" is declared out.
2156 Immediate (within . one-hour) emergency report made to the USNRC Operations Center in accordance with 10'CFR i 50.72 (a) (1) (i) .
2232 The NUE was terminated.
Unit 1 was ramped to 46% power and an unusual event declared when a ground alarm was received on Circulating Water Pump q(CWP) 1-1. While the ground alarm was in,-smoke was. reported in the -u-Unit 1 12kV switchgear room. CWP 1, phase'-B,.
cable was determined to have failed-and is the subject of this nonconformance. Note: ~CWP is not a tech spec _ item.
The California Department of Forestry. (CDP) was called to assist in fighting the potential fire. The smoke lasted less than 10 minutes and there were no personnel-injuries.- it was determined that the electrical ground in CWP 1-1 motor caused the load bank in the 12kV cable _
i spreading room to heat up,-burning paint, dirt, ;
and dust. There was no actual flame; only:
smoke in the room caused the' alarm. No equipment in the room was damaged.
93ncrwp\93EMN010.KWR Page 7 of 36 i
NCR DC1-93-EM-N010, D3 ', '
July 28, 1993 l During the power reduction, the digital rod l position indications (DRPI) did not follow the control rod indicators as control ~ rods'were inserted into the core. ' Control Bank "D" was declared inoperable. Tech Spec 3.1.3.1 requires restoration of Bank "D" to operable status within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> or to place Unit 1 in at least hot shutdown mode within the'following 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.
Event 2 On March 12, 1993, intermittent ground current alarms.
annunciated for Unit 1 CWP 1-2. The Unit was. ramped to less than 50 percent power and the pump was secured.
Investigations determined that one cable was shorted to ground between the first pull box'outside the turbine building and the next pull' box at the discharge structure elevation. When the cables were removed,-theLneoprene jacket was found to be. separating from all~three cables of the circuit for a distance of approximately 200-feet.
There was evidence (pull box water marks) .that the cables had previously been submerged in water.
The faulted portion and a similar length of the other tws cables for this circuit were removed,.a' mandrel was passed
_through the conduit several times to remove standing water, and new cables were installed. When the new cable sections were spliced in, no visual degradation was found at the splice locations. Cable samples were sent to Okonite, the PG&E TES laboratory, and Altran Materials for examination and testing. . Interim progress reports indicate that the outer neoprene jacket has been chemically degraded and the copper shield shows evidence of corrosion. However, electrical properties of the cable are acceptable, no evidence.of manufacturing defects has been identified, and no indication of-installation or.
abnormal operating conditions can be found. The root cause investigation for this event is-in progress.
]
TJMS p_ESCRIPTION 1812.18 12kV STUP OR AUX XFMR'1-1 GRD OC CWP 1-2 FDR GND ALARM CWP 1-1 FDR GND ALARM 1812.19 CWP 1-1 FDR GND ALARM cleared 1815.33 Smoke alarm in 12kV cable spreading room 93ncrwp\93EMN010.KWR Page 8 of 36 I
.. .=. . . . . _ . . . - . .
=
NCR DC1-93-EM-N010, D3 July 28, 1993 1816.24 Commenced: Unit i rampdown 1823.52 CWP'l-2 FDR GND ALARM cleared. .CWP 1-1Lselected !
i . .
1824.02- 12kV STUP OR AUX XFMR 1-1~GRD OC cleared' D. Inoperable Structures, Components, or SystemsLthat Contributed to the Event: ,
i SM-1 and SM-2 are the pull' box drain systems and associated sump pumps for Units 1 and 2, .respectively,"for.the pull 1 boxes. -
immediately outside ofi the turbine building.;- Investigation i determined that' water had. accumulated.in'the-pull boxes asia I result of the pull box drain. systems.and associated sump-pumps not being functional for a-period of time: preceding.the cable- i failure events. [
l E. Dates and Approximate Times for Major Occurrences:
- 1. February'05, 1993;. 2045.PST: :First event-date-
- 2. February 05, 1993; 2045 PST: Discoveryfdate.
- 3. February 05, 1993; 2106 PST: Declaration of1NUE.
l
- 4. February 05, 1993; 2156 PST: AnLimmediate (within
_ , one-hour) ,, non-emergency- ,
report was.made to the'NRC.
in accordance: withe 10 CFR 50.72 (a) (1) (i) .
,5. February 08, 1993; Twenty-four-hour! unusual event report made to. San'Luis Otispo Countylin-
-accordance with the requirements of ;
NUREG-0654,-Revision-1,. Appendix;1.
- 6. March 12, 1993; 1812 PST: Second event'date. .
F. Other Systems or-Secondary Functions'Affected:
i Event 1, February 5, 1993:- During the' power' reduction, the. l digital rod position indications -(DRPI) ' did- noti follow ' the control rod indicators as control rods were inserted into the:
core. Control Bank "D" was declared inoperable. Tech Spec >
3.1.3.1 requires ~ restoration of Bank'"D" to-operable status -
'within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> or to place Unit 1 in at least hot' shutdown mode within the following 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.
93ncrwp\93EMN010.KWR .Page 9' of-36 t
)
-, - - , . - , , ~ -
i NCR DCl-93-EM-N010, D3
- July 28, 1993 l
Event 1, February 5, 1993: During troubleshooting activities, the problem was determined to be with_the multiplex select signal path from the logic cabinet to the "IBD" and "2BD" power cabinets. Two potential logic cards were suspect.
After replacement of both logic cards, both "1BD" and "2BD" power cabinets properly indicated " Group B" selected. The l
bank overlap unit was reset to zero and incremented to'144 in
! accordance with GRPI step counter. The 144 steps was determined by flux map " sults.and verified by DRPI indication. Operations verified proper rod motion by performing STP R-1A and by the effect of rod motion-(i.e.
Tavg, Delta-I, NI's, etc.).
l Event 1,' February 5, 1993: Based'upon the evidence from the PPC traces and testing _of the suspected " Bad" cards, a malfunction that prevented Control Bank D rod motion did notf l
exist. The PPC traces indicated that Control Bank D-did in fact move when GPRI and DRPI showed rod movement. DRPI was indicating properly only above 138' steps. This may notLhave been readily apparent to operators since it was at the end of a rapid load reduction l complicated by the DRPI and P-A malfunctions. One would have to detect very subtle, but distinct changes in'NI power due to rod motion, while NI power 1 is changing due to the normal ramp down transient' response.
Event 1, February 5, 1993: Investigation indicates that these are. random, independent failures and are not a result of the CWP 1-1 Ground event. The root cause analysis and corrective actions to prevent recurrence associated with the DRPI system problems are being addressed in Quality Evaluations 10404 and 10405, and will not be dispositioned in_this NCR.
G. Method of Discovery: )
1 The events were immediately apparent to plant operators due to I alarms and indications received in the control room. i i
H. Operator Actions:
Event 1, February 5, 1993: The' Unit was ramped to 50% reactor power and CWP 1-1 was secured.
Event 1, February 5, 1993: The red control system was not !
fully operational following the ramp to 50% reactor power.
Tech Spec 3.1.3.1, for inoperable but tripable rods,.was entered. 1 Event 2, March 12, 1993: The Unit was ramped to-50% actor power and CWP 1-2 was secured.
93ncrwp\93EMN010.KWR Page 10 of 36
l
- NCR DC1-93-EM-N010, D3 July 28, 1993 I. Safety System Responses:
None.
III. Cause of the Event A. Immediate Cause:
The immediate casue was a conduction path to ground, which resulted in sufficient current to flow through the ground detection' circuitry.
B. Determination of Cause:
See attached root cause analysis (Reference __)
Extens:*r root cause investigation has been conducted since the firsc cable failure in 1989. The most probable root cause for the 12kV cables is chemical degradation of the neoprene jacket, resulting in subsequent corrosion of the copper shield and-failure of the EPR insulation due to excessive electrical-stress. All six cables for both CWP circuits were severly degraded for a distance of approximately 200 feet.
The cables demonstrate good voltage withstand test capability and_show no evidence of significant voids or impurities in the EPR insulation material. The following potential root cause categories have been investigated:
- 1. Manufacturing Defects Based on evidence evaluated to date, PG&b elieves that the failures in the 12kV_ cables are not caused by-manufacturing defects. Extensive cable dissection, wafer slicing and staining, and examination have not identified any significant voids in the cable insulation that would result in cable failure. No impurities or contaminants
- have been detected in the cable. insulation. Potential that point defects, which would be isolated incidents, existed in the cable insulation has not been~ ruled out. j No basis to assume imminent failure due to point defects -
based on outage frequency hi-pot testing.
Based on' voltage breakdown testing-(in excess of 80kVAC),
no common mode-failure mechanism has been identified for-the 12kV cables.
l 93ncrwp\93EMN010.KWR Page 11 of 36 l I
I-NCR DC1-93-EM-N010,'D3 ,
July 28,1993
- 2. Installation Related Problem Cable pulling tensions for thel12kV-cables routed between the turbine building and the intake structure were evaluated. The evaluation concluded that the vendor specified pulling tension ~and sidewall pressure limits were not exceeded.
Based on video probe inspections and the ease of cable replacement, the duct bank is intact, except for the conduit section damaged during CWP 1-1 cable removal, which is no longer used.
There is no visual evidence of damage to the cable shield, insulation, or conductor to indicate. installation damage.
Installation activities have been eliminated as a potential root cause for the:12kV cable failures.
- 3. Aging Related Problem The mechanical properties of.the jacket and insulation for the chemically degraded portions of the cable;are poor'~ .
The electrical properties of the 12kV cables have been
_ described as " good".
- 4. Design The cable design and application has been reviewed and determined to be adequate for the assumed-environment.
The assumed environment was expected to be dry with occasional short-term submergence. However, based onlthe documented extended submergence conditions ~ experienced at DCPP, a neoprene' jacket may not be the best' selection.
The cable was sized based on ampacity requirements for the CWP motors. Therefore, the 12kV motor feeder cables are operating at the upper end of their temperature rating (approximately 8092, worst case, compared to the design rating of 9092) . However, per Okonite,' the copper shield binding tape is designed to melt at approximately 8592, which provides indication of misapplication of the cable.
The DCPP 12kV cables show no evidence of binding tape melting. Therefore, the cables are operating.within their design rating.
93ncrwp\93EMN010.KWR Page 12 of 36
NCR DC1-93-EM-N010, D3 July 28, 1993 l i
The duct bank design has been reviewed. Based on cable pulling tension calculations, the pull boxes have appropriate spacing for cable installation. The ABS plastic conduits are acceptable for underground installations.
The pull boxes immediately outside of.the turbine _ building have drains, which are routed to common sump manholes for Units 1 & 2. These manholes are equipped with automatic submersible Class II sump pumps. The Start-up Feeder cable pull box, the pull boxes associated with the circuits to the intake structure, and the diesel-fuel oil piping trench, all drain into this common sump.
The Unit 1 and Unit 2 trenches are similar, except that the Unit 1 duct bank rises to cross over the circulating water ditaharge tunnel and then slopes downhill towards the intake structure. This design makes the Unit 1 section of cable conduits near the turbine building susceptible to submergence if the pull box sump pumps are not functional and if the water within the pull boxes rises above the conduit openings. . Video probe inspections show that the 12kV circuit conduits have low spots with standing water.
Pull box inspections between the turbine building and the intake structure for both Unit 1 and Unit 2 indicate that
~' pull box flooding is only a concern for the Unit 1 section between the turbine building ~and the discharge structure elevation. All other pull boxes show the water line has only reached the bottom of the first level of conduits for the Unit 2 section between the turbine building and the-
"~
discharge structure elsvation and have not reached any conduit elevations for the remainder of the pull boxes between the discharge structure elevation and the intake structure (2" drain conduit at bottom of pull boxes).
Based on the concrete duct and pull box _ design, there are no low spots for the 12kV Unit 2 startup power circuits routed along the turbine building between Unit 1 (incoming 230kV power location).and Unit 2. These 12kV circuits have also been exposed to water, but show no' evidence of-degradation. This is postulated to'be a result of the circuits being energized, but not loaded, and therefore, similar to the 4kV cables, are operating substantially below their conductor design temperature of 90*C.
93ncrwp\93EMN010.KWR Page 13 of 36
NCR DC1-93-EM-N010, D3 -
July 28, 1993
(
- 5. Maintenance Hi-pot testing does not appear to be a contributor. DCPP hi-pot test levels are below Okonite recommended levels for the DCPP cables and does not electrically overstress the cable. DCPP will continue outage frequency hi-pot tests (better to force failure during outage than have'an in-service failure).
Both Unit 1 & 2 common pull box sump systems were not adequately maintained, and had been out of service for a number of years (i.e. since 1987). Both Unit 1 and Unit 2 sump systems are now back in' service.
- 6. Environment Extensive cable dissection anu analysis has'been performed. Based on the documented water in the pull boxes, specific analysis for water related degradation was performed. There was no evidenc'e.of water tracking in the EPR insulation. In addition, there is'no evidence.of foreign chemicals in either the neoprene jacket nor in the EPR insulation. The ' copper shield shows no evidence of corrosion. Therefore, there is no evidence that. water is a contributor. The removed 4kV cable jacket is stronger
_than uninstalled 1972 vintage 12kV cable jacket that was in the warehouse. The 4kV cable has better-crosslinking than the 12kV cables and therefore better resistance to moisture and chemicals.
1 Moisture Intrusion: The cable oesign basis is for wet j and/or dry conditions. The cable can be submerged for prolonged periods. Inquiries to other nuclear power plants did not identify any trends for medium voltage EPR insulated cable failures in similarl applications. The Okonite black EPR cable is~widely used in electric utility distribution systems, including PG&E, and has-been_ highly reliable in similar applications.
Chemical Attack: Laboratory chemical analysis results have established that'the 12kV neoprene jacket was attacked by a chloride and/or fatty _ acid. High ambient temperatures accelerate the chemical related jacket-degradation. PG&E has evaluated the cable operating temperature and estimated that, in-the worst case, the-12kV CWP cables are operating with insulation-temperatures-in the range of 80 to 85'C. The hypothesized method of chemical migration and degradation is by way of water that 93ncrwp\93EMN010.KWR Page 14 of 36
l l i
l I
l
- l. NCR DCl-93-EM-N010, D3 l l July 28, 1993 l l
i intruded into the conduits through the cable pull boxes that are located immediately outside of the turbine building; the water intrusion resulted from the inoperable sump pumps and associated pull box drainage system.
Electrical transients (i.e. switching) can stress cables during normal operations. The long circuit distance coupled with the high resistance grounding design, could cause voltage transients to be induced due to switching.
This is considered to be a remote possibility; however, investigative action #13 was assigned to investigate voltage transient measurement techniques.
l Chemical ~ analysis has found no evidence of diesel fuel oil in the jacket or insulation.
The Unit 2 CWP 2-1 cable replacement shows no avidence of chemical attack.
C. Root Cause:
PG&E has concluded that the most probable failure mechanism l for the 12kV cable insulation failures is long-term chemical degradation of the neoprene jacket, probably due to chloride.
and/or a fatty acid, followed by corrosion of the copper shie_ld. When the shield deteriorates, uneven electrical stresses occur, ultimately resulting in a cable ground.
PG&E believes that the 12kV cable failures occurred-over an extended period of time (greater than a year). -This conclusion is based on the relatively mild pFe (approximately .m, 8.5) of the liquid found beneath the neoprene-jacket and the copper shield binding tape showing no evidence of high cable operating temperatures (the binding tape melts at l approximately 85*C).
D. Contributory Cause:
Analysis indicates'that there is substantial moisture present
~
in the ethylene-propylene-rubber (EPR) insulation (2% by l
weight, versus expected value of approx .7% by weight).
Moisture is therefore postulated as a contributory cause since the chloride needs to be in a moist environment-for the type of chemical attack experienced by the 12kV cable to occur.
l 93ncrvp\93EMN010.KWR Page 15 of 36
+
NCR DC1-93-EM-N010, D3 '
1
' July 28, 1993
- IV. Analysis of the Event A. Safety Analysisi The 12kV power and~ associated control circuits:potentially. '
affected by these cable failures are' associated with'the CWPs.
The motor-driven CWPs are part of the saltwater system which' removes energy f rom the turbine e:;haust steam (SG) entering the main condenser by providing cooling water to the condenser. Each unit is provided with two nonsafety-related-12kV motor-driven pumps located in.the intake structure. At 100 percent unit power level, both pumps mustrun to support ,
full load operation.
The 12kV system has high-resistance grounding, which allows continued operation for a limited time in the event of a single-line-to-ground fault. . Operators have~ received simulator training on ground fault incidents. The normal
~
plant procedure, as demonstrated in the past failure-incidents, is to declare the associated component inoperable, and then troubleshoot and repair the faulted circuit. The.
ability to operate for a limited time with a ground on the CWPs ensures that time is available to bring.the plant to a stable condition where the affected pump can be removed-from service and the circuit repaired without challenging any safe _ty systems.
The previously analyzed condition 2 event-de9eribed in the plant Final Safety Analysis Update (FSAR) for.the loss of offsite power and turbine trip bounds'this event; Results of the " Complete Loss of Forced Reactor Coolant Flow" analysis show that for a loss of all AC power, no adverse. conditions occur in the reactor core. The departure from nucleate boiling ratio is maintained above 1.30. The reactor coolant system is not overpressurized and no water relief will occur through the pressurizer relief or safety valves. .Thus, no cladding damage results and, consequently, there11s no release of fission products to the reactor coolant system. Since the February 5, 1993 ground fault alarm (equivalent to loss of i power to CWP 1-1) event is bounded by this previously analyzed condition, the health and safety of the public were not affected by this event. I B. Reportability: I
- 1. Reviewed under QAP-15.B and determined to be l non-conforming in accordance with Section 2.1.2.
i 93ncrvp\93EMN010.KWR Page 16 of 36 l
l 1
. . . . . _ _ =__ ... . . _ . _ -
l e
- NCR DC11 93-EM-N010, D3 -;
j July 28, 1993 ,
j .-
- 2. Event 1, February 5, 1993:. Reviewed under 10 CFR 50.72 (a) (1) (1) and determined to require an immediate i
(within one-hour) notification.to the'NRC due tor the l declaration of an Unusual _ Event. This does not require.a
! 30-day written followup LER. This immediate. notification was made at 2156 PST on February ~5, 1993.
NOTE: 10CFR50. 72 (a) (1) (i) , immediateinotification j requirement is applicable since an Unusual Event w==
l declared, thereby meeting the following criteria: "(a) !
( General Requirements. (1)' Each nuclear power reactor ,
licensee licensed under 50'.21(b)'orL50.22 oflthis part l
shall notify the NRC Operations Center via the Emergency _
Notification System of: (i) The declarati'on of anyjof'the Emergency Classes specified;in the' licensee's' approved
- Emergency Plan. (3)-The. licensee shall notify.the NRC; ;
immediately after notification of tFi appropriate StateLof' local agencies and not laterathan'one hour after the time the licensee declares one.of the Emergency Classes ~."
Event 1, February 5, 1993: A twenty-four;hourtUnusual Event summary report was made to San Luis Obispo Countyfin accordance with the requirements of1NUREG-0654, Revisioni 1, Appendix 1. ,
Event 2, March 12, 1993: Reviewed under-10 CFR 50.72 and
._10 CFR 50.73 per NUREG 1022 and determined to.not be reportable.
However, pursuant to Item 19;of Supplement.~1 to NUREG--
1022, PG&E submitted a voluntary: Licensee' Event Report-No.
1-93-005-00 regarding:the' failure ofEcertain medium ,
voltage 4kV'and 12kV cables. -;
l
- 3. Reviewed under 10'CFR.Part 21 and determined 1 that this-problem will not require a'10 CFR 21Lreport,'since it'is ,
being evaluatedLunder 10 CFR'50.72- j L 4. Event 1,' February 5, 1993: -This! problem hasibeen reported:
L via an INPO Nuclear Network entry..
- 5. Reviewed under 10 CFR'50.9 and: determined to'be:not- !'
reportable since these' events do not1have-a significant-
-implication'for public health.and~ safety or common. defense
~
and security.
- 6. Reviewed.under'the criteria of AP'C-29 requiring:the issue c and approval of an OE and determined that an OE is not' required.
- j. 93ncrwp\93EMN010.KWR Page. '17 'of '.3 6' j L
1 l ._
NCR DC1-93-EM-N010, D3 July 28, 1993 V. Corrective Actions A. Immediate Corrective Actions:
Event 1, February 5, 1993: 3 phases of circuit D05V00A', from BP015 to BP021A, for CWP 1-1 were replaced.
Event 1, February 5, 1993: ERP-#93-02 was formed to investigate this event Event 2, March 12, 1993: all cables for all phases of circuits E06V00A & B, from BP017 to BP039C,'for CWP 1-2 were replaced, after examining the conduit with a boroscope to determine if there was potential damage as found previously on CWP 1-1 conduit. In addition, the other cables'that were not
~
associated with the February 5, .1993 ' fault, were replaced'as a '
precaution.
B. Investigative. Actions:
- 1. Event 1, February 5, 1993: ERP 93-2, "CWP 1-1 Ground",
was convened to investigate this event. ;
- 2. Event 1, February 5, 1993: Provide cable samples to TES, Altran Inc., and Okonite for testing (chemical,
__ electrical, etc).
1 RESPONSIBILITY: R. Hanson l DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE'#01 STATUS: RETURN
- 3. Event 1, February 5, 1993: NECS/GC to provide design and asbuilt details, (architectural and electrical) showing original duct bank details and modifications since installation (i.e. I&C Building construction). First priority is details from turbine building to discharge structure, due by 02/18/93. Second priorityfis details from discharge structure to intake, due by 02/25/93.
RESPONSIBILITY: C. Hazari DEPARTMENT: NECS Electrical Engineering Tracking AR: A0293956, AE #02 STATUS: RETURN
- 4. Event 1, February 5, 1993: NECS/GC to provide design and asbuilt details, (architectural and electrical) showing original duct bank details and modifications since 93ncrwp\93EMN010.KWR Page 18 of 36
NCR DC1-93-EM-N010, D3 July 28, 1993 installation (i.e. I&C Building construction). First i
priority is details from turbine building to discharge structure, due by 02/18/93. Second priority is details from-discharge structure to intake, due by 02/25/93.
RESPONSIBILITY: G. Villareal '
DEPARTMENT: NECS Civil Engineering Tracking AR: A0293956, AE #03 STATUS: RETURN
- 5. Event 1, February 5, 1993: Investigate technologies to inspect conduit and/or determine duct bank routing.
Requirement is for access to 6"~O.D. ABS conduit with approximately 500 foot range. Potential source: Foster l
Miller Company.
l RESPONSIBILITY: R. Hanson i DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #04-STATUS: RETURN NOTE: TES NDE group has 300ft boroscope available.
- 6. Event 1, February 5, 1993: Provide, to Altran Inc., a sample of the pulling compound utilized by the Oregon contractors while attempting to remove faulted cable.
Also provide samples of other cable pulling compounds that may be present to assist in the chemical analysis.
RESPONSIBILITY: R. Hanson DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #05 STATUS: RETURN c .m i
l
- 7. Event 1, February 5, 1993: Provide samplesfof original, ;
uninstalled cable, for chemical analysis benchmark to TES, I Okonite, and Altran. j RESPONSIBILITY: R. Hanson DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #06 STATUS: RETURN ;
- 8. Event 1, February 5, 1993: Contact the Hazardous Mats ;
i group and determine the history of chemical spills in the-between the areas of the turbine building cable vault and discharge structure pull _ boxes.
RESPONSIBILIIT: R. Hanson DEPARTMENT: Electrical Maintenance 93ncrwp\93EMN010.KWR Page 19 of 36
NCR DC1-93-EM-N010,-D3 .
July 28, 1993 Tracking AR: A0293956, AE #07 STATUS: RETURN
- 9. Determine which CWP was aligned to for automatic _ restart.
RESPONSIBILITY: R. Hanson DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #08 STATUS: RETURN NOTE: CWP 1-2 was selected for restart.
- 10. Event 1, February 5, 1993: Determine whether the'CWP l cables have been determinated, partially pulled back,'then i repulled and reterminated due.to past emergent work (i.e.
- intake structure civil mods, Raychem cable replacement, l etc.). Coordinate with electrical maintenance since this l
has been partially completed due to ASW cable.
investigation.
RESPONSIBILITY: C. Hazari ECD: 02/18/93 DEPARTMENT: NECS Electrical Engineering Tracking AR: A0293956, AE #09 STATUS: RETURN NOTE: No record of re-pulling was found.
- 11. Event 1, February 5, 1993: ' Determine ~whether the DRPI l failure is related to this event. Include' potential affects of ground potential rise due to ehe 12kV ground j fault current.
RESPONSIBILITY: M. Nowlen DEPARTMENT: DCPP I&C' Maintenance Engineering.
Tracking AR: A0293956, JUS #10 STATUS: RETURN NOTE:~No other indications that the DRPI failure is
! related to the ground fault. No'other instruments l noted with spurious operation.
- 12. Event 1, February 5,'1993: Determine l regularity and scope of ground resistance maintenance.: . Include-history of recent maintenance and' ground resistor resistance measurements.
RESPONSIBILITY: R. Hanson DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #11 STATUS: RETURN 93ncrwp\93EMN010.KWR Page 20 of 36 1
.- NCR-DC1-93-EM-N010, D3
[ July 28,11993 NOTE: PM's to clean, inspect','.and megger every.18 !
months. Does not presently.make. resistance checks.
- 13. Event 1, February 3, 1993: Investigate effectiveness of l DCPP cable monitoring and testing program'.
RESPONSIBILITY: R. Hanson -
DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AEL#12 l STATUS: RETURN-
- 14. Event 1, February 5, 1993: . Develop? action plan.to inspect ,
failed and/or unfailed conduits to determineLtheir condition.,
RESPONSIBILITY: R. Hanson DEPARTMENO Electrical Maintenance Tracking'AR: A0293956, AE #13 l l STATUS: RETURN '
l
- 15. Event 1 and Event 2: Obtain samples, for! chemical . .
analysis, from withinLthe cable vaults associated with the Unit 1 CWP pump motor feeder cables,..locatedLoutside the.
I turbine building'12kV switchgear cablefspreading room.
i RESPONSIBILITY: R. Hanson ECD: 04/30/94 DEPARTMENT: PGEM
' Tracking AR: A0293956, AE #18 STATUS: ASSIGNED
- 16. Event 1 and Event 2: Obtain samples,7forLchemical .
. analysis, from within.the' cable vaults' associated with the -!
Unit 2 CWP pump motor foeder cables, located outside'the turbine building 12kV switchgear cable spreading room.
RESPONSIBILITY: R. Hanson DEPARTMENT: PGEM:
Tracking AR: A0293956, AE #19; l STATUS: COMPLETE
- 17. Determine'the actual' discharge water'. temperature!at'the-condenser.
RESPONSIBILITY: R. Hanson-DEPARTMENT: 'PGEM' Tracking AR: A0293956, AE #15 l STATUS: RETURN 9?ncrwp\93EMN010.KWR 'Page 21 of'36 j w s-W9 e w 4 4 w w m m W "+
NCR DC1-93-EM-N010, D3
- July 28, 1993
- 18. Obtain Unit 1 conduit temperatures in a spare conduit between the cable vaults outside the turbine building 12kV cable spreading room and the discharge structure elevation.
RESPONSIBILITY: R. Hanson DEPARTMENT: PGEM Tracking AR: A0293956, AE #16 l STATUS: RETURN
- 19. Investigate the feasibility of providing seals for the conduits (turbine building end only) routed between the outside pull boxes in the diesel generator area and the discharge structure. The purpose of the conduit seals would be to prevent potential chemical. spills from entering the conduits. All conduits in the bottom five feet of the boxes need to be addressed.
RESPONSIBILITY: C. Hazari ECD: 10/15/93 DEPARTMENT: NES-EE Tracking AR: A0293956, AE # 22 STATUS: ASSIGNED
- 20. Inspect the spare CWP 1-1 potentially broken conduit during 1R6.
_ RESPONSIBILITY: R. Hanson ECD: 05/01/94 DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #23 STATUS: ASIGND
- 21. Investigate methods for sealing tne pull boxas at the turbine building (outside in the DG area) anc how this information should be documented for future pull box seal replacment.
RESPONSIBILITY: R. Hanson ECD. 10/15/93 DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #24 STATUS: ASIGND
- 22. Engineering to investigate options available for cable replacement for the 12kV Standby /Startup feeder circuits, routed along the turbine building from Unit 1 side to the Unit 2 side. These feeder circuits consist of one 1500 MCM cable for each phase. The preferred replacement option would utilize 500 MCM cables.
93ncrwp\93EMN010.KWR Page 22 of 36
l l
- NCR DC1-93-EM-N010, D3 I July 28, 1993 I
RESPONSIBILITY: C. Hazari ECD: 10/15/93 DEPARTMENT: NES-EE ;
Tracking AR: A0293956, AE #-25 l STATUS: ASSIGNED C. Corrective Actions to Prevent Recurrence: I
- 1. Event 1 and Event 2: Clean out the cable vaults and pull- ,
l boxes for the Unit 1 CWP pump motor feeder circuits to the 1 intake structure, repair the associated sump pumps as necessary, and repair the cable vault and pull box lids to j help prevent moisture intrusion.
RESPONSIBILITY: R. Hanson ECD: 04/30/94 I DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #20 f
Outage-Related? Yes OUTAGE: 1R6 OE Related? No NRC Commitment? No CMD Commitment? No
- 2. Event 1 and Event 2: Clean out the cable vaults and_ pull boxes for the Unit 2 CWP pump motor feeder circuits to the intake structure, repair the associated sump pumps _as necessary, and repair the cable vault and pull box lids to help prevent moisture intrusion.
l ' RESPONSIBILITY: R. Hanson COMPLETE DEPARTMENT: Electrical Maintenance Tracking AR: A0293956, AE #21 Outage Related? Yes OUTAGE: 2R5 OE Related? No ,,,, ya NRC Commitment? No CMD Commitment? No
- 3. Replace ASW 1-1 4kV cables between the' pull box:in the 12kV Switchgear room and the discharge structure elevation pull box. Send samples of the cable out for electrical and chemical analysis.
l RESPONSIBILITY: R. Hanson RETURN DEPARTMENT: Electrical Maintenance _
Tracking AR: A0293956, AE #17 Outage Related? Yes OUTAGE: 2R5 OE Related? No NRC Commitment? No CMD Commitment? No 93ncrwp\93EMN010.KWR Page 23 of 36
i l l
I l NCR DC1-93-EM-N010, D3
- i l July 28, 1993 NOTE: NCR DC2-92-EN-N054 corrective actions establish a i
preventative maintenance program for the sump pumps'and drains associated with both the Unit 1 and Unit 2 cable vaults located outside the turbine building 12kV Switchgear cable i spreading rooms.
1 D. Prudent Actions (not required for NCR closure)
Since a megger test (1.e. insulation resistance measurement) ;
l isolates the component from ground, the resistor banks- l l resistance is not evaluated as part of the PM program. ;
Therefore, revise the resistor bank PM's to include resistor- l bank resistance measurement and evaluation. '
RESPONSIBILITY: ECD: l l Tracking AR: A0 l
VI. Additional Information ;
A. Failed Components:
None.
B. Previous Similar Events:
NCR DC2-89-EM-N104 On October 29, 1989,. ground current alarms associated with j the Unit 2 Auxiliary Saltwater (ASW)- Pump 2-2 annunciated '
twice and immediately cleared both times. The pump was removed from service. A ground was found on one cable that runs between the turbine building pull box and the first pull box at the discharge structure. Cables in this pull box were found submerged in water. The faulted portion and a similar length of the other two cables for this circuit were removed, the water was pumped out of the pull box, a mandrel was passed through_the conduit several times to remove standing water, and new cables were installed. Visual examination found-no obvious physical-defects in the removed cables. Cable samples were sent to-Okonite and to the PG&E Technical and Ecological Services (TES) laboratory for exau.Ination and testing. The cable testing determined that the cable met or exceeded the original mechanical and electrical stress limits set forth in the original purchase specification, except for some minor loss of mechanical strength of the jacket material.
The testing laboratories determined the failure to be an isolated event.
93ncrwp\93EMN010.KwR Page 24 of 36 1
NCR DC1-93-EM-N010,LD3.
July 28, 1993 2
QE Q0009710 On May 3, 1992, intermittent ground current. alarms;were received associated with Unit 1 nonsafety-related 4kV Bus D. The 4kV Bus D was removed from service. The-ground ~
was determined to be located.on-one cable.between the. pull. ,
box located immediately outside~of the intake structure and the Bus 14D transformer in the intake structure The pull box located outside of theJintake structure was found to have water inside, which was pumped'out to facilitate l cable replacement. Approximately 40 feet /of the single faulted cable was_ replaced. Visual examination.found no obvious. physical. defects in the removed cable. Cable samples were sent.to Okonite and to the.PG&E TES laboratory for examination and testing. The examination and testing determined that the physical properties of the insulation were normal.and that t the jacket properties displayed some loss of elongation which-was' considered.
normal for neoprene installed for almost 18 years. The electrical properties were normal for 18-year'old, cable.
The testing laboratories could identify no' definitive.
reason for the failure.
NCR DC1-92-EM-N054
_On October 31, 1992, ASW Pump 1-2 on Unit.:1 was removed from service for refueling outage maintenance. ks partfof the procedure to return.the pump--to service,. a motor hi-pot test was conducted. During the hi-pot test, the cable.
insulation developed a ground fault'at:approximately 6kV, Investigation determined that_the fault;was located on one cable between the first pull box outsida the turbine
~
building and the next_ pull box at the discharge structure elevation. Cables-in thisipull box were also found.
submerged in water. The faulted portion and a: similar length of the other two cables for this circuitr were removed, the water was pumped;out of the* pull box, a' mandrel was passed through-the conduit.several; times to -
remove standing water, and new cablesLwere installed. -
Other than at the fault point, visualLexamination revealed' no obvious physical defects in the. removed cablesi Cable samples were sentLto: Okonite, the PG&E TESJ1aboratory,cand.
Cable Technologies Laboratory.(CTL) for examination'and-testing. In February 1993, an additional sample'was sent to Altran Materials for chemical analysis. ' Interim.
progress reportsfindicate that theielectrical. properties of the cable are acceptable, no firm evidence of manufacturing defects has been identified, and no 93ncrwp\93EMN010.KWR Page 25 .of 36
- :c
- g
NCR DC1-93-EM-N010, D3 *
-July 28, 1993 indication of installation problems or abnormal' operating conditions can be found. The root cause investigation for this event is in progress.
SREF 88-119 INPO Significant Event Report entered'12-SEP-88. DCPP Unit 2 - Ground faults on a.12 kilovolt system. An electrical ground fault at a reactor coolant pump motor terminal connection led to multipleLeontrol. room.
annunciator alarms and successive electrical. failures.
Before the exact location of the fault could be determined, a fire developed in aLatartup transformer's.
grounding resistor bank, and the reactor was manually scrammed.
C. Operating Experience Review:
- 1. NPRDS:
None.
- 2. NRC Information Notices, Bulletins, Generic Letters:
IE Information Notice No. 86-49: Age / Environment Induced Electrical Cable Failures.
On November 21, 1985, San Onofre Unit 1; experienced a loss of-offsite power when a. transformer was tripped by its differential relays because of a. failed cable to the class 1E 4160-V bus. The_most'likely cause of the cable failure was determined to be temperature-induced accelerated aging and degradation of;the cable insulation. The notice also, stated'"Another important facet of the periodic maintenance and testing program for cable circuits is.the'walkdown: inspection.to identify actual or potential environmental conditions (heat, water, chemicals, etc.) in the immediate vicinity of the cables that could adversely affect cable conditions.
PG&E's response to this notice was that DCPP's class 1E cable outside of the spreading areas and-switchgear rooms'is run inside rigid-iron conduit and is properly ;
routed away from high temperature piping and equipment.
93ncrwp\93EMN010.KWR Page 26 of 36
NCR DC1-93-EM-N010, D3 July 28, 1993 IE Information Notice No. 92-81: Environmental Qualification of Control Cables.
Based on tests conducted by Sandia National l
Laboratories for the NRC, certain electrical cables with bonded Hypalon jackets may not meet. environmental qualification requirements. Sandia conducted tests on cables manufactured by three different companies to determine the minimum insulation thickness necessary to perform its intended function should the insulation be damaged during installation, maintenance, or other activities. During LOCA testing, all 10 of the Okonite-Okolon cable. samples failed. The test specimens were single-conductor, 600 volt, 12 AWG control cables insulated with ethylene-propylene rubber (EPR) with bonded Hypalon jacket.
Information Notice 92-81 was issued to address NRC concern with environmental qualification of Okonite cables with bonded Hypalon jackets that have not been qualified for service conditions exceeding 50*C for 40 years. According to the NRC, qualification testing that does not use the actual jacket service configuration may not be representative of actual cable performance. The Staff cautions that the integrity of cables could be affected if the cables are used inside containment, used in continuous power circuits, routed with power cables, or routed close to hot piping.
This information notice is not applicable to the 4kV cable failure since the 4kV cable is does not require
~ ~
environmental qualification, does not use a Hypalon jacket, and the neoprene jacket is not bonded to the cable shield / insulation.
None.
D. Trend Code:
Responsible department __, and cause code __.
E. Corrective Action Tracking:
- 1. The tracking action request is A0293956.
93ncrwp\93EMN010.KWR Page 27 of 36
l NCR DC1-93-EM-N010, D3 July 28, 1993 -
l
- 2. Are the corrective actions outage related? YES-F. Footnotes and Special Comments:
Following the March 12 event, PG&E conservatively replaced all ,
Unit i non-failed medium voltage circuits between the pull l boxes outside the turbine building and the next pull box at- i the discharge structure elevation (ASW Pump 1-1 4kV cables and ;
the second 3-phase 12kV circuits for'CWPs 1-1 and-1-2). 'Also, )
one complete circuit, between the turbine building switchgear and the motor terminations at the intake structure, of-the Unit 2 12kV CWP 2-1 motor feeder circuits was replaced. The ,
neoprene jacket on the Unit 1.12kV CWP cables showed evidence i of chemical degradation, similar to'that on the.previously-replaced Unit 1 12kV CWP' circuits. Visual examination of the ASW 1-1 4kV cables and the CWP 2-1 12kV cables revealed no defects. I l
G.
References:
i i
I
- 2. Initiating Action Request A0293569. -l
- 3. Unusual Event declaration summary letter, dated February 8, 1993, to SLO County Administrator.
I
- 4. Event Response Plan (ERP) 93-02.
- 5. NCR Tracking Action Request A0293956. l
- 6. Report Tracking Action Request A0293573. !
- 7. Rod Control System problems associated-with this event; Action Requests A0293570, A0293576 and A0293578.
- 8. SREF 88-119.
- 1. On February 10, 1993 the TRG convened and considered the following:
93ncrwp\93E!CIO10.KWR Page 28 of 36
l .'
l NCR-DC1-93;EM-N010, D3- .
I July 28,.1993 l
i Investigative Actions-1-14 were discussed and assigned.; j
' Discussion ensued'regarding.the similarity'between'this failure and the recent 4kV ASW Pump 1-2 cable failure. ;It. j was agreed.that the preliminaryifailure: modes;arei different since the 4kV cables did'not exhibit' neoprene jacket degradation like:the 12kVECWP, feeder. 'Both,the 4kV l & 12kV circuits are medium? voltage, shielded 1 cables, with ,
EPR insulation and neoprene;outerijackets.- Therefore, the-TRG. concurred that.the existing safety analysis-forEDCl- i 92-EM-N054 isLnot:affected cy1this event. ,
TRG discussed the requirements:for: cable' samples to;TES, -!
Okonite, and Altran.Inc..: .Altran requested that7three 1- ,
i foot pieces be; forwarded for' analysis. JOne' piece from;the ~ a j failure vicinity,.one' piece from:the " medium". damage. area, .;
and one piece from the~ low. damage area.. Alsofreguestedi l -
! that some oF the liquid,J if available,'from the' conduit be '
! sent for analysis. A'10Lfoot1section oftthe' cable willi also be sent.to Altran for electricalitesting Additional ,
cable' samples will be'provided.to.Okonite and TES.
l The TRG discussed the issues 5related:tofground_resistorb
, smoking in the 12kV; cable spreading room. -Initial- ;
indication is1that-these resistors ~are cleaned every-18 1
~
months _ (outage f requency) , but_ due to inaccessibility, not..
! much other maintenance'is: performed ;
t
~The CWP cables are 500MCM, 15kV, EPR' insulate'd:with a neoprene jacket. -TheLfaulted cable.was-notidegradedrat . 1 the pull box ends, then both ends' exhibited approximately :
80 feet of " mushy"-jacket" material,cand then there was-approximately 200 feet ofTeable in;theicenter portion of.
- the run that.had no jacketEat all. JTheffault1was in the " ,
section of the cable that did:notThavefany' jacket'present.. !
There was no detectable! hydrocarbon odor. -Theicopper foil .
shielding was extremely corroded-inLthe' areas where.the' ,
i jacket was missing. The mushyjappearance,of.theijacket- -
l preliminarily indicatesia?long-term chemicalLattack. 'l A small amount of liquids (i.e.; drops)?waslobberved between-
~
l
- the jacket'andithe. insulation. fA-litmus; paper; test;of liquid: indicated a'pH.of between 5 and'7. Note: .the'
~
Li liquid may have been introduced asipartfofJthe cable: .
' removal process.
.l l
l Resistance measurementsJof the'~ semi-conductorJshield !
material indicated: 10k-ohm /cm;forLportioniof'cableiwith ~l good jacket,-approx- 300/400k-ohm /cm'for.the portioncof:the' ,
-1 i
93ncrwp\93EMN010.KWR 'Page' ~29. of736. 1 i
i ji
. t l
NCR DC1-93-EM-N010, D3 July 28, 1993 -
O cable with bad jacket. Expected values for new cable would be approximately ik-ohm /cm.
The CWP 1-1 cables were last hi-pot tested cn1 11/5/92 (Reference W.O. R0085212)
During the cable removal process, the cable moved for approximately 75 feet, then jammed and would not move, even with approximately 8000 pounds pulling tension.
Normal installation imaits are 2000 pounds pulling tension maximum, with' normal values measured in the 500 to 1000-pound range. While trying to loosen the cables.for. final removal, approximately 750 gallons of water was pumped-into the conduit, and no water was observed coming out the other end. This indicates a potential conduit break between the Turbine building cable vault'and the discharge structure elevation.: The CWP conduits are the l'owest conduits in the duct bank.
The conduit duct bank construction, i.e. is there a concrete cap, did the I&C building construction potentially interfere with the conduit bank, etc. will be investigated prior to the next TRG.
Based on this failure, and the recent 4kV' cable failure, an Integrated Problem Response Team (IPRT) was formed to in accordance with procedure OM7.ID7 to help investigate The medium voltage cable failures and.make recommendations.
THE TRG WILL RECONVENE ON WEDNESDAY, 2/24/93'TO DISCUSS INVESTIGATIVE ACTION RESULTS ROOT CAUSE.
- 2. On February 24, 1993 the TRG convened.and considered the -
following:
Hazardous Materials responded that there have been DFO spills in the arca of the cable vaults. -There have been other spills as well. A complete list with dates, etc. is being developed. Past experience indicates.that the 2-1/2" DFO stainless steel pipe routed along the turbine building was attacked by microbes, that live within:the. I fuel oil, in the early 1980's. Ate. welds from the inside, resulting in leaks. This occurred near the roll-up doors east of the DFO tanks, not in vicinity of the cable vaults. ~
l 93ncrwp\93EMN010.KWR Page 30 of 36
)
NCR DC1-93-EM-N010, D3 July 28, 1993 l
Preliminary chemical test results indicate no hydrocarbons (DFO) present. The jacket, shield, and water sample were all tested. Lots of halogens (i.e. chlorine)'have been detected. Degradation found'at the. tape overlap point of the copper shield. Green color on the copper shield is '
galvanic corrosion, red color could be from microbe attack ~
or from overheating during the ground fault. l Preliminary electrical' testing shows termination. failure at 80kV-AC (i.e. electrical properties acceptable).
The degraded condition of the shield indicates the following hypothesis; Since the shield acts to equalize the potential across.the insulation,. shield degradation could introduce electrical stress-(voltage differential) on the insulation at the point where shield no longer protects the cable, which can result in insulat i on failure. This is why termination stress cones are used.
Investigative action #14 assigned to-examine other duct bank conduits to determine potential for ground water intrusion through broken conduits.
Electrical stresses will concentrate at moisture intrusion portions of the cable. Industry practice shows need.to minimize moisture exposure and potential intrusion.
'The shield is grounded at only'the switchgear end of the cable. If shield currents are measured, theEtheoretical current would only be capacitive charging current. With the 3-phases are measured together, the total would be zero. Shield degradation would provide a hi-resistance ""
ground path, and leakage current .:2asuremenEs would be higher. Shield leakage currents have been measured as indicated below: (NOTE: All measurements are in amperes)
FDR FOR FDR FOR CWP 1-1 CWP 1-2 BUS 14D BUS 14E A1 0.36 A1 0.33 A 0.16 A 0.30 A2 0.48 A2 0.62-
- B1 0.39 B1 0.34 B- 9.30 B 0.17 B2 0.77 B2 0.49 C1 0.56 C1 0.54 .C 0.26 .C 0.30
- C2 1.11 C2 0.43 VECTOR SUM: 0.85 0.77- 0.06 0.05
- CWP 1-1 15kV cable failure'for this NCR.
- This value may indicate potential shield degradation 93ncrvp\93EMN010.KWR Page 31 of 36
NCR DC1-93-EM-N010, D3 '
-July 28, 1993 THE TRG WILL RECONVENE ON THURSDAY, 3/18/93 TO' DISCUSS INVESTIGATIVE ACTION RESULTS AND TO CONTINUE THE ROOT-CAUSE DETERMINATION PROCESS.
- 3. On March 19, 1993 the TRG convened and considered the following:
The TRG agreed that the CWP 1-2 cable failure on March 12, 1993 should be included in the Ecope of this NCR.
Investigative Actions 15-18 and Corrective Actions 1-3 were assigned.
Duct bank construction - (i.e. grade slopes, etc. ) were discussed. Refer to Reference 9, " Figures related to the CWP'l-1 and'l-2 Cable Failures".
No water has been observed. leaching from the conduits into the pull boxes, which provides further indication that the water / chemical intrusion isLform the cable vaults / pull-boxes, and not from a deformation of the conduits.
Preliminary CWP 1-2 Cable Failure Investigative Action Results (measurements taken at actual time of failure):
- 1. Sulfates and constituents of' sea water found in.
the liquid sample obtained from under the CWP 1-2 cable jacket. pH measured at 8.5
- 2. Insulation temperature measured at 80 F.
Water from.within the conduit measured at 9 3'F .
- 3. Boroscope indicates. standing water in the conduits.
- 4. Hardness measurements were obtained on some of the Raychem Flamtrol cable within the.CWP 1-2 cable vault. The hardness, measured at three different locations on the same cable, was approximately 95 durameter "A" hardness scale, which is consistent with "as manufactured" specifications.
THE TRG WILL RECONVENE ON,:4/8/93 TO DISCUSS INVESTIGATIVE ACTION RESULTS AND TO CONTINUE THE ROOT CAUSE DETERMINATION PROCESS.
- 4. On May 14, 1993, the TRG reconvened and. considered the following; The most probable root cause for the 12kV cables is chemical degradation of the neoprene jacket, resulting in subsequent corrosion of the copper shield and failure of 93ncrwp\93EMN010.KWR Page 32 of 36
NCR DC1-93-EM-N010, D3 July 28, 1993 the EPR insulation due to excessive electrical stress.
All six cables for both CWP circuits were=severly degraded for a distance of approximately 200 feet.
The cables demonstrate good voltage withstand test capability and show no evidence of significant voids or impurities in the EPR insulation material. The following potential root cause categories have been investigated (Refer to determination of cause section) :
A. Manufacturing Defects B. Installation Related Problem C. Aging Related Problem D. Design E. Maintenance F. Environment Recommended corrective actions will be forthcoming upon completion of the IPRT investigation and report submittal.
The TRG will reconvene in July to review'the IPRT recommendations and determine corrective. actions.
- 7. On July 21, 1993, the TRG reconvened and considered the following; A general overview of the failure history and
~ investigation was provided to bring the new TES individuals up to speed.
The preliminary root cause of the 12kV cable failures is chemical degradation of the outer neoprene jacket, leading '
~
to shield deterioration. '
The TES draft report of electrical and mechanical testing was discussed, in conjunction with the involvement of the TES contracted outside laboratories and Okonite. It is expected that TES will consolidate all TES and-Rose conclusions into the TES report.
Okonite partial discharge testing-on_the dry 4kV cables performed up to approximately 11kV with no indications of partial discharge. Okonite will perform " salt water" cable testing. This is an ongoing effort and will be part of the long term cable aging program (i.e. will not keep NCR open).
93ncrvp\93EMN010.KWR Page 33 of 36
NCR DC1-93-EM-N010, D3 . :
July 28, 1993 j The Altran report and Okonite 12kV; cable testing will be stand alone and will not be completely covered in the TES final report.
No evidence of abnormal manufacturing process (i.e.
excessive voids) or installation damage was detected.
The installed cables meet.the original production test requirements. Fatty acid esters and chlorides have been detected in the 12kV cable jacket and insulation.
The' chemical'compvaition resembles that of a' light detergent or' cleaner type compound (i.e. ivory liquid or borax). Altran has concluded that the cable degradation was experienced over'a long time period and may have been.exasperbated by high ambient / operating _ temperatures.
The Unit 2 pull boxes and drains between the turbine-building and the intake structure have been cleaned. Unit 1 boxes and drains will be cleaned.1R6.
The inspections of the boxes after significant rain will continue. The TRG and/or IPRT needs to determine whether these rain inspections need to be long term or only until all boxes and drains are cleaned and associated sumps'are l verfified as operating consistently.
-The moisture under the CWP cable jackets was a concentrated sea water type mixture. The pull box samples were a diluted sea water type mixture.
The TRG concurs that it would be prudent to replace one ASW 4kV circuit, from switchgear to. motor. terminals, for Unit 1 during 1R6. This circuit would~then be examined and tested, and used'to determine the scope of future actions.
Using a lead sheathed cable for moisture (and chemical) protection would create new reliability problems for the circuits (i.e. splicing), and'will not be considered further. A polyethylene' jacket may be utilized to provide increased moisture / chemical intrusion resistance.
Consideration is being given to installing a new medium voltage circuit in the failure area"and leaving the-existing circuit in place as a sample.
Discussion ensued regarding methods to control' chemical intrusion into the pull boxes _and associated raceways.
Consideration will be given to raising / sealing the pull 93ncrwp\93EMN010.KWR Page 34 of 36
NCR DC1-93-EM-N010, D3
-July 28,'1993 i box covers, installing rain gutters on the turbine i building to reduce runoff, and to pull box conduit seals. :
An investigative action was assigned to NES to investigate the feasibility of installing conduit ranis. An investigative action was assigned to electrical maintenance to investigate methods.of sealing the pull box covers.
In addition, since the Unit 2 12kV. standby /startup-feeder circuits are routed next to.the turbine building and may have been exposed to water and chemicals, an investigative action was assigned to NES to investigate options available for repairing the 1500MCM circuits if one should fail. These cables are considered as'being subjected to the same environment as the 12kV failed cables based on deeper pull boxes, drainage of the conduits back toward the pull boxes, the circuits;are'not normally lo=ded, and- '
there is no evidence of moisture in interim circuit pull boxes. The TRG recommended a prudent action to obtain a jacket sample for analysis.
To help determine if the same chemical: exposure has ever been present in other pull boxes at the turbine building, one 600V spare control circuit sample was taken from the Unit 1 ASW 1-l ' pull box (#1/0 cable) . Examination of the cable found evidence of fatty acids in the cable
~~ insulation and galvanic corrosion of the t' inned copper.at the tin / copper interface on the cable. ThisDgalvanic corrosion has erroded the tin, which does not present a functionality problem, but could. complicate attempts to i splice circuits'together. However, this galvanic corrosion could cause interference problems for instrument cables (i.e. cross talk) if the mL '_ ar shierd"has '"1 deteriorated. Since there have not been any problems during instrument loop testing, and the tested sample had exposed cable terminations in the pull box (i.e. spare circuit taped up within the box), this is considered a long term problem consideration. It is recommended that i additional control / instrument circuit' samples be obtained during 1R6.
Final TES report is expected by 8/2/93, final'Altran .
report expected week of 7/23/93,'and final IPRT report.is ;
expected to be issued by 8/16/93.
Overall NCR completion date of 7/15/94 was established.
l TRG will reconvene on September 14, 1993.
93ncrwp\93EMN010 KWR Page 35 of 36
e.
NCR'DC1-93-EM-N010, D3 '-
July 28', 1993 '
I. Remarks:
None. ]
.i l
-. }
'i I
93ncrvp\93EMN010.KWR .Page 36 of 36
. - - . . . .