IR 05000373/1993040
| ML20059K648 | |
| Person / Time | |
|---|---|
| Site: | LaSalle  |
| Issue date: | 01/24/1994 |
| From: | Burgess B, Yin I NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML20059K643 | List: |
| References | |
| 50-373-93-40, 50-374-93-40, NUDOCS 9402020196 | |
| Download: ML20059K648 (9) | |
Text
~
.
.l U. 5. NUCLEAR REGULATORY COMMISSION-l
REGION III
Reports No. 50-373/93040; 50-374/93040(DRS)
Docket Nos. 50-373; 50-374 Licenses Nos. NPF-ll; NPF-18 i
!
Licensee:
Commonwealth Edison Company
-
Executive Towers West III 1400 Opus Place - Suite 300 Dowers Grove, IL 60515 Facility Name:
LaSalle County Station, Units 1 and 2 Inspection At:
LaSalle Site, Marseilles, Illinois Inspection Conducted:
November 18, 1993 through January 5, 1994 M4f L
[f Inspector:
.): T. Yin,.;f Date '
/
,
7, v
,
ApprovedByG4b..,gjs / cec ~
/
VhV
.
B. L Burgess,4 hief Date'
,
Operational Programs Section Inspution Summarv inspection on November 18. 1993 throuqh January 5. 1994 (Reports No. 50 -
373/93040: No. 50-374/93040 (DRS))
Areas Inspected:
Routine, announced inspection to assess licensee control and -
implementation of plant design changes, and corrective actions (IP 37700).
The inspection focused on a Main Steam piping snubber reduction design change, and licensee corrective actions for safety relief valve pressure set point
'
drift and weeping problems.
Rctsul t s: One apparent violation was identified.
The violation concerned a failure to meet a Technical Specification surveillance test.equirement for two of the safety relief valves.
Also discussed are identified concerns
,
regarding main steam safety relief valve leakage and lift setpoint failure problems and the f ailure potential of a main steam bypass line..
P r
,
i.
9402020196 940124 7 PDR ADOCK 05000373 G-PDR v
.-
-
... _
..
-...
... _.
. _.
.
_ _
.
_ _.
._.
.._.__. _
_
_
.-
__ _ _
. -.
_
.
-
..
t DETAILS
,
1.0 Persons Contacted
,
Commonwealth Edison Company (CECO)
+
W. Murphy, Site Vice President
.
M. Reed, Technical Services Superintendent
'
E. L. Martin, QV Director
'i R. M. Regan, System Engineers Supervisor
!
J. R. Kodrick, Maintenance Group Supervisor a
J. Lockwood, Regulatory Assurance Supervisor U. S. Nuclear Reaulatory Commission (NRC)
+
l B. L. Burgess, Chief, Operational Programs Section i
D. E. Hills, Senior Residence Inspector The above individuals and other licensee individuals attended the exit
,
'
interview on January 5, 1994.
Other individuals were contacted during the course of this inspection.
2.0 latroduction
,
The objective of the inspection was to assess the licensee's engineering
'
and technical support activities.
The effectiveness of the engineering organization.in the performance of routine and reactive site activities l
was evaluated, including the identification and resolution of technical
'
_
issues and problems.
The inspection focused on Main Steam (MS) Safety Relief Valves (SRVs), and MS snubber reduction system modifications. ~
,
,
3.0 Discrepancy Records (DRs)
L The inspector reviewed the 1992 and 1993 DR log, and observed:that there-had been 28 DRs written on SRVs and Relief Valves from February 1992 to
October 1993.
Most of the problems were failures to meet test criteria.
'
All 18 SRVs on Unit 2 MS. lines had been tested-during the last refueling outage (L2R04, January to April 1992).
Six SRVs failed LaSallel
Technical Specification (15) 3.4.2 lift setting ~ requirements.
The'
i licensee did not report the SRV. failures because they considered ~the failures to have occurred at Wyle Laboratory, where the SRVs were
{
tested, and because all of the SRV failures were within the. primary i
coolant system design pressure.
Not reporting SRV lift-set failures is-t consistent with current NRR guidance.
l t
4.0 SRV Set Pressure Drift The inspector reviewed the entire history of Unit 2 SRV set pressure
.
test records, and identified a total of 13 failures. -The test i
frequency, SRV selection, and testing were based on ASME Section XI,..
Sub-sectinn IUV-3510-3514 requirements. The tests.were performed at.the
.5
,
.,
-.
~. - -
._
,
-
.
.
..
-
-
..
-
.
--
.
.
.
-
~..
~
!
Wyle Laboratory using Procedure No. 1015.
The following table lists the i
Unit Two failures by the outage in which they were tested:
'
l OUTAGE OUTAGE DATE FAILURES-
)
L2R01 January to' June 1987 five failures L2R04 January-to' April 1992 six failures L2R05 November to December 1993 two failures.
.-!
-i The following paragraphs address potential contributors to SRV setpoint drift that the licensee had not thoroughly investigated prior to the-inspection.
In response, the inspector performed a systematic evaluation of the 13 failures, and observed the following phenomena:
a.
Two failures occurred at location A, three at M, and two at N.
Location A is the last SRV on MS line D.
Location M is the first
'
SRV on MS line A.
Location N was the first'SRV on MS line C.
These locations were the nearest to the. individual line change of.
direction, and could potentially experience highest line vibration.
b.
Among the 13' failed SRVs, nine also failed the leak rate test.
There appeared to be a correlation between the SRV leakage and set pressure drift. This correlation involved the most likely cause for SRV leakage: the steam cutting of the valve disc.and seat during repetitive cycling either during system testing or during a primary system overpressure transient..During this cycling,' the forces associated with the rapid opening of the SRV and~ subsequent'
i slamming of the disc on the valve seat during closure could alter
~
the set pressure.
- ,
c.
The SRV safety pressure settings were. divided into five groups.
The failures were distributed as'follows:
__
',
SRV LociJions Pressure Settino Number of Failures
'
J,N,R,V 1205 psi
B,H,L,M 1195 psi
"
'A,F,G,P 1185 psi
C,D,E,K 1175 psi
.
.
S, U '
1150. psi
- 1
'
Even though the 1195 psi and 1185_ psi groups.~ experienced twice as d
many failures as the other three~ groups, no scientific explanation
-
was apparent for this phenomenon.
d.
SRV serial numbers were' also investigated.
Since the'SRVs were_
l exchanged between Units 1 and 2, and some.of the SRVs.were placed
.;
as spares, the evaluation had become very complex and.beyond the
'
scope of'the inspection.
However, th licensee completed the.
evaluation, and concluded three facts.
One, some serial numbers
.j failed. repeatedly.
Two, of those SRVs tested, the SRVs that-were
,
,
f
-
.
.
.
__
_
_. _.
,_
.
..
.
-.
-
.
.
.
-
.
-
j
'
i; c:
i L
H t.
placed as spares and received limited maintenance appeared to
,
perform better in. service.
Three, there was a correlation between SRVs that leaked and -set pressure test failures.
i'
e.
The licensee's evaluation for Unit 1, based on:the inspector's
.
suggested evaluation pattern, identified-that:the Unit 1 SRVs had_
i only four pressure set point test failures. However, during the
)
l:
process of reviewing the data, the licensee found that SRVs IB21-
'
F013B and 1821-F013J failed to conform to the-ASME_Section XI,
Table IWV-3S10-1, test schedule (five years plus) requirement.
i These two-SRVs, presently installed in an operating Unit, had not-
~
been tested since their first test in 1986. -TS 3.4.2 requires'
b that the safety valve function of 17-of.18 reactor coolant' system l1 safety / relief valves shall be operable with the.specified code-safety valve function setting in operational modes 1,~2, and 3..
.
Given that the two valves had not-been tested in accordance with
- i the ASME code requirement, both SRVs were declared inoperable _ on'
~ !
December 4, 1993, in accordance with TS 4.0.3..TS 4.0.3 stated, in part, that a failure to perform a surveillance requirement-l within the specified time interval shall constitute an equipment H
failure (or inoperability).
The licensee filed an Notice of
,
l-Enforcement Discretion from TS 4.0.3 for LaSalle Unit 1 on
-
l December 6, 1993.
The enforcement discretionary request was l
verbally approved by NRC on December 6, '1993. The -failure to test
'
the two SRVs within the time interval resulted in the inoperability of both valves.
Returning Unit 1 to power operation frem the last refueling with two SRVs inoperable is an apparent
~
!
violaticr. of TS 3.4.2.
During investigation regarding root cause
. 3
L of tt u apparent violation, the inspector identified that a-formal test P ogram did not exist for either Unit 1 or 2lthat clearly identihed those valves requiring ASME testing during:each refueling outage.
This lack of. a formal program-resulted in the responsib'.e engineer not identifying the need to test the above
!
two SRV valves.
Given that the lack of a formal program to test
'
the SRVs was true for both units, the apparent violation will be assessed against both Unit I and Unit 2 (373/93040-01; 374/93040-01(DRS)).
l The utility is a participating member of the NUMARC effort in evaluating SRV lift setpoint drift and testing failures.
The utility has indicated that any initiatives that evolve from this study, in addition to-initiatives'that may come from other' utilities-investigations, will be~.
evaluated for inclusion into a _ corrective action program to improve the performance of the Safety Relief Valves at LaSalle.
r L
5.0 SRV Seat Leakaae The inspector reviewed the history of Unit 2 SRV seat leakage problems l
and obtained the following data:
!
a.
The number of leaking SRVs found during tests conducted at the Wyle Laboratory were considered to be substantial.
The leakage
I t.'
,
,.
-
-
..
- - -,
.
-
_.,
.
.
-
-
.
!
s acceptance criterion of I lbm/hr., or 35 ml/5 min. was developed by Crosby, the SRV. manufacturer.
L2R01 January to June 1987 15 leakers i
L2R02 October 1988 to February 1989 4 leakers
!
L2R03 March to June 1990 6 leakers L2R04 January to April 1992 6 leakers L2R05 November to. December 1993 8 leakers Total: 3Y leakers
b.
Not all 18 SRVs were tested during each refueling outage.
The a
SRVs sent to.Wyle during L2R02, L2R03, and L2R05 were. based on a.
high temperature reading of 200*F or above at the SRV discharge
-piping. This temperature limit was not considered to be credible based on the fact that during L2R04, all 18 SRVs were tested.-
- !
SRVs A, E, J, M, P, and V exceeded'the vendors test leakage acceptance criterion. However, the SRV temperature measurements obtained in September 1991 showed only A, D, and E had discharge
. i line temperatures that exceeded 200 F.
.
c.
Similar to the SRV pressure set point drift deficiencies, there
_
were a number of specific (same serial number) SRVs having t
repetitive seat leakage problems.
'
d.
With many. continuously leaking SRVs in the MS system, the amount
'
of steam discharged into the suppression pool during nonnal plant -
>
operation was substantial.
To maintain the suppression pool water temperature within.the 95 F TS limit, the large capacity,. low head-Residual Heat Removal (RHR) pumps and heat exchangers were
frequently used in the suppression. pool cooling mode.
For.
example, RHR was utilized in the suppression pool' cooling mode for
-,
a total of 15 times during the Summer of 1991 (7/4/91 to 8/29/91)
'y on an average of every two to six days; Each run lasted from less than an hour to approximately ten hours.
Also, RHR was _used for suppression pool cooling on five occasions in the Winter of 1991-
(1/1/91 to 2/30/91) on an average of everyEfour to 16 days.- Each
'
run lasted from'less than one hour to more than two hours.
This kind of use of RHR pumps and associated heat removal. systems was not in accordance with plant design', and could cause unnecessary
,
wear and tear on safety related mechanical and electrical
..
"
components.
It could also threaten RHR system avemulity during'
adverse plant conditions.
e.
The licensee performed 'an evaluation' after receiving NRC
,
Information Notice (IN) 87-10, " Potential for Water Hammer During'.'
- i Restart of Residual Heat Removal Pumps".- The IN stated that water
".
hammer could occur during the'use of RHR in the suppression pool-
-
cooling mode coincidental with a Loss Of Offsite Power-(LOOP), and-receipt of Safety Injection (SI) signal.
The licensee's
!
evaluation considered the above event combination to be unlikely.
~
'
The licensee further believed that the RHR pump's 30 second' coast
,
,
v
,.
.,
.%..
-
,,
r w-
,
e c
-
v
-,r
-
-
-
..
-
.
..
. -_.
.
-
u
.
l'
'
L down time could. maintain sufficient pump discharge pressure (water column). However, recent system records showed a-pressure drop /
.
water column loss in less than ten seconds.
In view of the high
!
L.
usage rate of RHR pool cooling, the licensee revised four
'
.
operating procedures (LOP-LP-04, LOP-RH-13, LOP-RH-16, and LOS-RH-l Ql). The procedures required the operator to. place the Low Pressure Coolant Injection (LPCI) pump (RHR), a subsystem of the Emergency Core Cool.ing System, in manual start in lieu of the
,
original auto start design. During suppression pool cooling and :
l after a LOOP, if an SI signal (i.e. a'large break Loss of Coolant Accident) was received, the operator was required by procedure to evaluate the potential for water hammer before manually switching i
on LPCI. However, placing the RHR system in manual would reduce
,
the redundancy of LPCI and potentially delay low pressure coolant j
injection capability under the abme scenario and circumstances.
The inspector was concerned that placing RHR (LPCI) in manual j
during suppression pool cooling, and the feasibility of shifting RHR from suppression pool cooling to.LPCI injection based on an operator decision regarding the. potential for a water hammer is a
-
decision that the operator should not be making and could adversely impact RHR injection capability.
f.
The licensee reviewed Unit I leaking SRV situations using.the inspector's approach, and found Unit 1 conditions similar to Unit-2.
We are concerned with the potential adverse impact of excessive _ use of a j
train of RHR in the suppression pool cooling mode and the. fact that the
'
utility has not addressed the SRV leakage ~ problem..This concern _is based on the amount of time RHR is unavailable for automatic response to a design basis event and the fact that current procedures' require'the operator to determine the potential. water hammer impact on the RHR system.
This decision places the operator in a~ very difficult position in that there is very little time and information available to the
,
operator in which to evaluate RHR water hammer-potential. We are also i
concerned with the number of SRVs that _ appear to continuously _ leak.
During the exit the utility _ committed to providing a response to this concern by providing data comparing LaSalle SRV leakage problems with.
industry averages, y
6.0 Snubber Damaae The inspector reviewed Modification M01-2-90-008.
This modification removed 45 snubbers from the Unit 2 Main Steam System (MS) and planned
'!
to replace 41 snubbers on the large bore (12" to 28") MS piping outside
containment, 2MS-09, a non-safety related system.
The modification was
_j requested on June 29, 1990, due to snubber deterioration found during maintenance activities.
The cause of the snubber deterioration was-attributed to vibration and stop valve and control valve operation.
The modification was evaluated and approved by CEC 0 Nuclear Engineering
E Department on July 1, 1991.
The actual work was completed on February-
.
27, 1992.
'
i
- -,
-
.
.-
.
.
.
.
. __
.
.
_
_
_. _ _
__
._.
_. _.
-
.
!
..
During a MS piping walkdown on September 24, 1991, snubber MS33-2839S,
[
located on a MS. bypass line, was found buckled and broken at~ the extension piece.
Even though the cause of the snubber. failure was believed to be due to a severe hydraulic transient that occurred when the reactor scrammed from. full power with-full bypass valve operation,.
no licensee measures were taken to identify the cause of-the transient,
)
and the magnitude and direction of the forces that destroyed the snubber.
Instead, Temporary System Change (TSC) 2-302-91 was written on September 27, 1991, stating that the loss.of Snubber MS33-2839S will' not'
affect the seismic and High Energy Line Break.(HELB) analyses delineated _in the Updated Final Safety Analysis. Report (UFSAR).- The TSC
-
did not prohibit snubber removal and replacement activities and-ignored
,
the apparent severe hydraulic transient; a transient.which had not been designed or analyzed for this portion of the MS system.
At the end of the snubber reduction effort, 45 snubbers were removed from 2MS-09.
Among these snubbers, 24 failed the manual stroke test.
Among the 24 affected snubbers, five failed the 5% rated load drag force test.
The licensee analyzed this adverse condition, and declared that_
'
the condition met the UFSAR requirements.
Once again, the licensee failed to identify the root cause of the system transient and the magnitude and direction of the_ forces that had damaged the snubbers.
To determine the snubber design basis, all four snubbers on the MS Bypass Valve 2B21BPV-3 discharge line were reviewed. The following data was obtained in the design records:
,
Snubber Number Condition Desian Load Valve Thrust 2MS33-28395 Failed-2701 lbf.
2221 lbf.
Extension
broken
2MS33-2840S Working 4839 lbf.
1283 lbf.
2MS33-2844S Working 4858 lbf.
1420 lbf.
l
'
2MS33-2845S Failed test 4968 lbf.
2001 lbf.
All four snubbers were Pacific Scientific Company mechanical. PSA-3
,
units, rated at 6000 lbf.
The snubber design loads took.into consideration valve opening thrusts, system dead weights, thermal forces, and loads from operating and design basis earthquakes. _To fail'.
,
a PSA-3 unit under normal plant operation events, the 2MS-09 syster must have gone through on~e or more un-analyzed severe transient events.
!
7.0 Licensee Snubber Inspections
,
The licensee had different sets of walkdown and inspection criteria.for l
the 41 replaced snubbers.
In lieu of the manual stroke tests conducted for the.45 removed snubbers, an observation of the cold'and hotipipe-
_l
-
snubber positions was taken.
Nine snubbers were found..that deviated
.
from-design expectations.
The licensee. hand stroked each of the nine
..
.
-
.
.
-
.
units, and found one failure. At the end of the modification, 45 snubbers were replaced, four were replaced with hydraulic units, and one was replaced by another mechanical unit.
The licensee issued another Modification 'in May 1992. -The second Modification was completed in December 1993, and. 35 mechanical units
-
were replaced by hydraulic units.
In the process, the licensee also manually stroke tested all the units before replacement.
Three'. failed'
both the stroke and the 5% rated load drag force tests.. This illustrated that, at the' time when' broken and failed snubbers were found in September 24, 1991, the licensee failed to determine the full scope.
of the severe hydraulic event that had occurred on 2MS-09, 8.0
[15 Safety Concerns The inspector reviewed the lvensee engineering; evaluation packages, and'
concurred with the utility' conclusions from a regulatory standpoint,
,
including:
The use of ABB Impell SUPERPIFt computer program.
-
The application of ASME Code Case N-411-1 for higher damping
.
factors.
The licensee position that the UFSAR piping' analyses were not
-
affected.
,
However, the licensee failed to evaluate the problem from_ the standpoint:
of the effect on_ plant personnel safety and area contamination.
Normally, to resolve any potential severe hydraulic transient that-could cause snubber and/or piping failure, identification of the root cause is necessary.
Root cause determination would then allow corrective action to prevent a severe hydraulic transient from occurring. A re-design of
a system or the changing of certain steps in the procedure are examples'.
."
If these measures were not feasible, other corrective actions to address
'
the problem, including an increase in the~ number and/or the size of the piping restraints to resist dynamic. loadings, could be taken.
However,
<
not only were none of the' above mentionediactions taken for 2MS-09,.a large percentage of the,available restraints (snubbers) had been~ _
'
removed, allowing the system to move unrestrained.
Should another transient similar to the one(s) that broke Snubber 2MS33-2839S have recurred, there will not be any sn~ubbers (allLfour on the valve
!
discharge line had been removed) to absorb system loads and energy.
To
,
address this concern, the utility was requested by the inspecto'r to
'
seriously consider monitoring the 2MS-09 line for displacement.
If
,
identified displacements during full bypass line operation are within
.
design, no further' action is required.
9.0-Licensee Followun Actions The 1icensee conducted a failure evaluation for. Snubber IMS33-2839S during the NRC inspection, more than two years after the damaged snubber.
,
.
,
h
.
--
. -.-
-
-
- --
.
.
-
.
- __
- -
-
. -
a
.
,
,
was identified.
The licensee concluded that the snubber failure was not i
solely due to the September 24, 1991 turbire trip event.
The snubber accumulated internal damage and-component f ailures due. to thermal movement of the piping during a number of plant heatups acd cooldowns.
. t This damage was possible because the internal _ failures allowed the snubber to lock at any position, and subsequent pipe temperature changes
- 7 exposed the snubber to stress loads that varied in direction and magnitude. The hypothesis that only hydraulic transient loads caused the snubber to fail was not credible given'the inspector's observation-
of the snubber damage and the MS bypass line system configuration.
Since the cause of the damage was attributed to thermal movement, and
~
the fact that the snubber reduction program had been completed, no
' l further action was'necessary.
i During a management meeting at the site on January 5,'the inspector stated that the snubber failure and the extension piece failure were.
caused by more that just thermal loadings.
On the other hand, the'NRC staff did not know the exact cause of the failure.
Based on the lack of conclusive information and the identification of other snubber failures on the system, the NRC asked the licensee to consider monitoring Valve No. 3 discharge line for three transients, unless problems were observed on the first, or second transient. The NRC further stated that the licensee should seriously consider providing three dimensional instrumentation at.the closest pipe elbows to the two failed snubbers.
The inspector observed heavy rusting on the 2B21BPV-3 valve discharge i
nozzle and connecting piping close to the nozzle.
The rusty condition
'
was unique in that the other four valves and discharge nozzles showed no rust.
The inspector questioned whether a crack existed; and if the rusting was due to continuous surface wetting during plant. operation.
Based on the inspector's recommendation, the licensee removed the-
. i insulation in the region, and performed a Magnetic-Particle Non-
destructive. Examination (MT) for the areas of concern.
The MT fou'nd no
!
indications.
10.0 Exit Interview The inspector met with the licensee representatives (denoted in
. Paragraph 1) at the conclusion of the inspection. The.. inspector-
,
summarized the purpose and findings of the inspection. The license
representatives acknowledged this information. The inspector also discussed the likely informational content of the inspection report with regard to documents or processes reviewed during the. inspection. The -
licensee representatives did not identify any such documents and processes'as proprietary.
L
!
F w
up-
'-
-e
?
" '-
e-c,+=
-y
-
e-
- P