IR 05000416/1989015
| ML20245D671 | |
| Person / Time | |
|---|---|
| Site: | Grand Gulf  |
| Issue date: | 06/20/1989 |
| From: | Blake J, Coley J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML20245D665 | List: |
| References | |
| 50-416-89-15, NUDOCS 8906270196 | |
| Download: ML20245D671 (9) | |
Text
--.
-
-
-
-
,
_
,
%
.
o
,
g
,
UNITED STATES '
?[A PEGgq}og
.,
w,'
.
NUCLEAR HEGULATORY COMMISSION
'(
,
. $* s.
. g REGloN 18'
= - -
.'
o
' 101 MARIETTA STREET, N.W..-
i~ (
f ATLANT A, GEORGI A 30323
Ea '
j.
%, w
/
f e
,
,
,
,
.,,-
-
,
W l
,
<
a;
.
.
,
.
.
Report No.:
50-416/89-15
'
r, l
Licensee: -System Energy Resources,:Inc.
]
Jackson, MS 39205
'
Docket No.:
50-416 License'No. : ' NPF-29 Facility Name:
Grand Gulf ~
'
Inspection Conducted: 'May 30 - June 2, 1989 Inspector: k..
LA 4 - 2O-89
'
J.
'
D' ate. Signed
~
V h 2e Sf Approved by:
.
/
._
J. J.
lake, Chief Date Signed
ter als and Processes Section
'
ngi eering Branch
,,
Division of Reactor Safety.
SUMMARY Scope:
This special, unannounced inspection was in the areas of on-site followup of the Byron Jackson recirculation system pump shaft crack.
Results:
Plant management was actively involved in efforts. to' ensure quality and safety as indicated by the enhanced noise and vibration detection monitors that had'
.
been. installed at Grand Gulf.
This equipment allowed early detection of pump
' problems and subsequent safe shut down.
I The coordination of management, engineering and maintenance efforts was excellent during the recovery operations as indicated by a. total' down time of 2 weeks.
The licensee also preserved in as undisturbed state as possible all r
rotating.. parts from both loop "A" and loop "B" pumps for analysis. ~ The. Plant.-
Safety Committee' conservatively limited future operation of the recirculation system pumps to one' fuel cycle 'unless an analysis; identifies the cause of the'
failure and corrective actions can be taken or,inspbetions~ performed.'
'
'
g e
g.
+
+
i4 r
,
'
"
8906270196 6906'1
.
i j
.
,
,
!~
-
PDR ADOCK 05000416
-,
s-i Q
,
"
-
%
>
.
+
.
.
- _ _ _ _ _ _ _ _ _.
-
.
.
.
.
.
.
One weakness was noted by the inspector regarding failure of the licensee to document and report the erosion observed on the pump journal bearing inside surface (paragraph 2.d).
This information could be significant for determining the cause of the pump shaft failure.
Similar erosion has been observed in another BWR-6 plant with similar pump design and shaft cracking problems.
In the areas inspected, violation or deviations were not identified.
I
1
_ _ _ _ _ _ _ _ _ - _ _
_
.
.
.
.
&
t l
REPORT DETAILS 1.
Persons Contacted Licensee' Employees J. Brihmadesam, Metallurgist, Corporate J. Bucklew, Mechania Maintenance Supervisor
- R. Byrd, Plant Licensing
- W. Eiff, Principal Quality Engineer R. Green, System Engineer, Recirculation System
- C. Hayes, Supervisor, Quality Programs
- R. Hutchinson, Plant General Manager
- L. Molder, Operations Superintendent C. Quick, System Engineer, Feedwater System
- J. Roberts, Manager, Performance and. System Engineering
- J. Summers, Compliance Coordinator S. Thomas, Principle Engineer, Nuclear Plant Engineering
- M. Wright, Manager, Support
- S. Zinke, Superintendent,-Systems Engineering Other licensee employees contacted during this inspection included craftsmen, engineers, technicians, and administrative personnel.
NRC Resident Inspectors
- H. Christensen, Senior Resident Inspector
- J. Mathis, Resident Inspector
- Attended exit interview
2.
Onsite followup of Byron-Jackson Pump Shaft Crack at Grand Gulf Unit 1 (93702)
a.
Background Information l
On May 21, 1986, General Electric (GE) Nuclear. Energy issued Surveillance Instruction Letter (SIL) No. 003 which, in part, addressed thermal fatigue cracking in pump shaft and covers.
Byron Jackson (B-J) issued a related advisory (Technical Note 8701-80-005)
in January 1987.
Both documents centained the conclusion that although the observed cracking was undesirable,. it was considered to be relatively benign based on analysis and~ observations.
Cracks were
,
expected to exhibit virtual arrest or show very small growth for depths exceeding 0.2 inches (shafts are 5 to 8 inches in diameter with an approximate 1-1/2 inch wall thickness in the area of concern.
The January B-J technical note recommended that utilities disassemble and inspect the pumps at the first convenient opportunity.
However,
-
,
... _ _ _ - _ _ _ - - _ _. _ - _
.. - -
-.. - _ - - - - - -
I
,.
.
3
.
.
-
-
i GE SIL No. - 459 dated December 15,1987 ' reported that more recent l
findings, indicated that -the cracks may represent a threat'to pump reliability.
The SIL reported that, recent observations and measurements for'the pumps in operating installations 'showed that some. cracks had penetrated to depths ' greater than 0.3 inches in both shafts 'and.
covers. The depths were greater than previously anticipated based on:
~
,
L the expected frequency of thermal cycling. 'In the cover, the cracks j
were axial andicould be of sufficient depth to' penetrate the closed -
cooling water '(CCW) circuit if cracks.and flow passages were Lat corresponding azimuths.
This condition represented a risk of.
O contaminating a. normal clean system and possibly subjecting :it to j
higher pressures.
In the' shaft, some cracks changed orientation from
- 1 axial to circumferential.
This represented an. increased risk' of j
shaft -failure under. normal operating conditions if.the ' cracks propagated to sufficient depth.or under abnormal conditions if the
shaft is subjected to increased; dynamic loading as.could occur from
)
cavitation or a foreign object becoming lodged in the pump.
j
During the period,in which the above observations were' made, B-J performed an analysis of the crack initiation 'and growth mechanisms.
This work included ' analysis of test data from a full scale operating temperature pump shaft and cover mock-up.. The data verified previous conclusions that mixing cold seal purge flow with. hot: system water-initiates cracks.
The data also showed that previously unknown. low frequency fluctuations occur.
Shaft and cover crack growth analysis considering these low frequency fluctuations supported the field observations and indicated that thermal cracks could penetrate to depths beyond the observed depth of approximately 0.3 inches, but '
that propagation rates _ would be low.
The analysis.also indicated that cracking in pumps without seal purge ' could occur but was expected to be less severe and to' occur' at'a higher position on the shaft and cover.
As a result of the above, GE recommended that i
Boiling Water Reactors (BWR) owners consider the findings and develop
.)
contingency plans to prevent negative impact on power production and
{
to minimize effects on ' outage maintenance ' activities.
GE's j
recommendations for such contingency plans included the following.
!
(1) Review pump and system design, period of operation and l
maintenance history.
Correlate with other field experiences to
assess the potential impact on the pump or pumps in question.
For pumps with a high potential. for significant. cracking, the
'
following actions should be considered:
Install shaft probes to monitor. shaft vibration response..to j
.
protect against consequential damage' in the event crack
'
propagation changes from thermally driven' to mechanically l
driven.
j
i I
ll
,
_ _ _.. _ _. _ _ _ _. _ _ _. _
_
_
_
-
.
,
i.
.
.
Monitor CCW effluent for contamination by reactor coolant to provide early warning and to minimize contaminatica of external CCW -system and equipment in the event communication develops between these systems.
(2) Review ASME Section XI ISI plans and commitments and establish plans which consider the recent findings described in the SIL.
Schedule inspections of pumps with greater than 80,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> of pump operation to detect cracks and implement corrective actions before excessive crack growth occurs.
(3) Prepare an inspection plan and develop the following:
Methods for examining pump shafts and covers
-
Criteria for return to service
-
-
Repair methods and alternatives considering ALARA Plans for replacement of existing parts or use of improved
-
designs BWR utilities were also requested to inform GE of the results of the pump examinations.
Grand Gulf Nuclear Station had installed additional shaft vibration monitoring probes and loose parts monitoring davices.
The system installed for vibration monitoring is described in detail in document no. (6) in paragraph 2.c (Rotating Machinery Surveillance System
[RMSS] - Final Report).
The system for loose part monitoring was engineered and supplied by Babcock and Wilcox (B&W) and is described in Section 4.4.6.1 of the Grand Gulf FSAR. The licensee also had monitors on the CCW system to provide early warning of contamination of the system by reactor coolant.
However, the licensee did not examine the recirculation pumps during the cycle 3 outage in March 1989, based on an evaluation performed by GE and described below.
The Grand Gulf recirculation pumps had essentially the same design as the pumps at a foreign plant.
One of the pumps at the foreign plant had been examined in August, 1988 after approximately 30,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> of service.
The examinations revealed crack indications in the shaf t and heat exchanger.
As a result of these findings the rotating elements and heat exchangers of the foreign plants pumps were replaced.
Representative samples of the shaft and heat exchanges which were removed from the foreign plant were metallographically examined by GE under contract to System Energy Resource Inc (Grand Gulf).
The metallographical examination revealed axial and circumferential l
I l
l
__
_ _ _
_ _ - _
--
.
.
.
.
cracking with a maximum crack depth of.195 inches.
The cracks were transgranular with no environmental influence, no evidence of arrest and clearly showed fatigue striations.
One other observation noted in the examination of the rotating element was the existence of a groove 360 degrees around the journal bearing at an axial location of abcut four inches above the weld joining the journal to the impeller.
Similar appearing grooves were observed in each of the two rotating elements inspected at the foreign plant.
A representative section of a journal was also decontaminated and inspected in San Juan and Vallecitos Laboratory.
The axial position of the deepest part of the gruove examined was quite uniform around the circumference.
The maximum depth of the groove was close to.2 inch of the.7 inch journal thickness.
There was a " chevron" or
" river and tributaries" pattern above and below the deepest part of the groove.
The total width of the band having some metal removal was about 1.5 inches.
The area where metal was removed had a smooth surface with an eroded appearance.
GE concluded that the smooth surface was not consistent with the classical cavitation phenomenon which results in a pocket appearance and the mechanism which initiated the apparent erosion was not identified.
However, GE's evaluation concluded that the licensee could operate another cycle before examining the Grand Gulf B-J recirculation pumps (which had only seen 26,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> of service) based on crack growth projections.
b.
Event Chronology On May 11, 1989, at 1330, during a normal operational pump startup of B pump with the pump on fast speed and as the flow control value was being opened to increase flow, a noise apparently coming from "B" pump was first noticed on the Loose Parts Monitoring System.
Between 1330 and 1600 it was noted that the vibration amplitude had increased from 17 mils to 31 mils at the pump coupling and 5 to 11 mils at the motor.
Therefore at 1638 on May 11 the pump was placed in low speed operation.
Shaft vibration decreased to 11 mils at the pump coupling and 5 mils at the motor.
The loose parts monitor audio signal of impact was at the same frequency as the pump RPM i.e., 7.5 Hz at slow speed operation.
There were no abnormal pump or motor operating parameters other than noted above.
The mechanical seal including the seal staging continued to operate normally.
After evaluation of the vibration data and loose parts monitor audio signals the B pump was removed from service early on May 13, 1989.
On May 16, the plant entered Cold Shutdown (Mode 4) and began work to replace the "B" Recirculation Pump.
Af ter the pump and motor were uncoupled, a test run of the motor did not indicate a vibration problem with the motor.
On May 23, the
"B" Recirculation pump replacement evolution had progressed to the point where a visual inspection of the pump impeller and pump shaf t interface area could be conducted.
These
-
_ _ _ _ _ _ - _ _ _
_
,.,
-
r
+
+
3'
-
,
-
.
.
-
5 i
examinations revealed what ' appeared to be a through-wall' crack which extended approximately-. 320 ' degrees around _ the shaf t on' the sloped area of the shaf t above shaf t-to-impeller. weld. ' Diametrical.
~
~
measurements taken between the shaft 'and the journal bearing.also Lindicated that: the shaft and the impeller were_ out of alignment.
j l
. Tabs on~ locking devicesifor 6 fasteners on the hydrostatic. bearing
.
were' also' observed to-~be missing, apparently caused' from.
.
embrittlement resulting from improper heat treatment.
)
c.
Event Followup. objectives:
j l.
On May 30,.1989 the inspector' arrived at.the' Grand Gulf Facility to
]
'
'
'~
determine whether licensee actions - to identify the. cause of this d
event and to replace the cracked pump shaft were. proceeding in an effective manner.
The inspector learned thatlboth A'and B pumps had
been re-assembled using replacement parts. from the Unit 2 recirculation pumps and hot runs were inprocess.
[
The inspector therefore, started collecting information and l
interviewing cognizant licensee personnel regarding the causes of the I
event, the event chronology, functioning of. safety systems and system monitors, consistency of licensee actions with licensee requirements and approved procedures, proposed licensee action to identify the cause of the event, corrective actions taken prior.to resumption of-
facility operation, verification that plant and system performance were within the iimits of the analysis described in the Final Safety Analysis Report (FSAR), and assurance that the licensee had preserved in an undisturbed state those components that failed, or that had received damage which may' indicate the cause'of the failure during a subsequent analysis.
Documents reviewed in the process of evaluating the licensee actions are delineated below:
Document (DOC) Identification Title (1) Drawing IF-7836 Rev. D, Byron' Jackson Reactor Circulating, Pump (2) Byron Jackson Technical Manual Vertical Recirculation
"
No. 8020DV Pump (3) General Electric Surveillance Byron-Jackson Recirculation Inspection Letter, SIL No. 459 Pump Shaft and Cover Cracking (4) General Electric Report, Dated, Recirculation Pump shaft:
-
. April 25,.1989.(System Interaction Cracking Study, Invesitgation Phase)
(5) General Electric Report, Dated, Evaluation of Byron-Jackson March 1989 Recirculation Pumps at Grand j
Gulf (BWRG) With Respect. To Pump Shaft and Heat Exchanger Cracking '
Issue L
'
,.
.
..
.
.
Document (DOC) Identification Title (cont'd)
(6) Rotating Machinery Surveillance Development and Demonstration System - Final Report of Surveillance and Diagnostics of Rotating Machinery for Reducing Radiation Exposure (7) Maintenance Work Order M-93647 Pump Internals Inspection (8) Material Nonconformance Report - 0216-89 (9) Operating Plant Experience Report, Dated May 23, 1989 (10) Justification for Ccntinued Operation - Grand Gulf
"A"
. Recirculation Pump - Draft Report (Final Report Never Issued)
(11) Memo to C. R. Hutchinson, Grand Nuclear Station General Manager from D. L. Pace Nuclear Design, dated May 30, 1989, " Loose Parts Analysis of the Locking Tabs Used on the Reactor Recirculation Pump."
(12) Grand Gulf Final Safety Analysis Report paragraphs, 4.4.6.1
" Loose Parts Monitoring,15.3.3 Recirculation Pump Seizure and 15.3.4 Recirculation Pump Shaft Break.
(13) NRC Memorandum to Edward L. Jordan from Thomas M. Novak dated October 18, 1989 (AE00/E807), Memorandum forwarding a AE0D Engineering Evaluation Report for " Pump Damage Due to Low Flow Cavitation," dated August 1988 d.
Analysis of Licensee Actions Grand Gulf Management, engineering and maintenance efforts were outstanding in the recovery effort as exemplified by the rapid replacement of rotating parts in loop "A" and "B" recirculation pumps and the sheer volume of instructions, evaluations, reports, etc.
required to perform this evolution.
Enhanced safety systems such as the loose parts monitors and the rotating equipment surveillance system operated as designed and may have prevented more serious damage to the pump.
The licensee's permanent corrective actions will be determined after a detailed analysis of damaged rotating parts.
The event was within the limits described in the FSAR.
The licensee has preserved in an undisturbed state all rotating components.
Inservice testing of the pump was performed as required.
The plant safety committee at Grand Gulf conservatively limited future i
operation of the recirculation pumps to one fuel cycle unless
!
analyses of the rotating parts identified the cause of the failure and corrective actions can be taken or inspections performed to verify crack depth.
- _ _ _ _ _ _ -
_- -
..
.
-
,
-
.
....
.
.
However, discussions held, with cognizant licensee personnel revealed details that may have a pronounced effect on the resolution of the pump shaft failure.
Personnel stated that the Grand Gulf "B" loop recirculation pump also had erosion 360 degree around the journal bearing at the same location axially as the foreign reactor pumps.
This finding was not documented in the Material Nonconformance Report (0216-89). Personnel further stated that, the loads necessary to drive the crack in the shaft to failure in such a short period of time could be due to the fact that the recirculation pumps on a BWR-6 have flow control valves on the discharge piping, downstream of the pump.
These valves are nearly shut during startup and low power operation while the pumps run at full speed.
The recirculation pumps are design to deliver 44,000 gallon of water / minute at high speed.
When the pumps are operated in this configuration they vibrate noticeably (17 to 18 mils) and the recirculation coolant is just recirculating in the pump.
The inspector considered these observations to be critical since there is information available on pump damage due to low flow operations (Refer to Doc. No. (13) in paragraph c above).
In addition to the pump operating in this configuration at start up, the pump ran in this configuration during the first 2 years of low power operation.
In sur aary the inspector suspects that the erosion observed in the
"B" pump at Grand Gulf and the foreign pumps is caused by cavitation and results from the mixing of the recirculation coolant during low flow operation with the pump at high speed.
The inspector also regarded the lack of information on the erosion observed in the Grand Gulf "B" pump journal and the failure of GE to identify precisely the mechanism causing the erosion in the foreign pumps a weakness in the licensee's program to resolve technical issues effectively.
Within the areas examined, violations or deviations were not identified.
3.
Exit Interview The inspection scope and results were summarized on June 2,1989, with thote persons indicated in paragraph 1.
The inspector described the areas inspected and discussed in detail the inspection results listed below.
Although reviewed during this inspection, proprietary information is not contained in this report.
Dissenting comments were not received from the licensee.
!
I
_ _-