IR 05000317/1993023
| ML20056G978 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs  |
| Issue date: | 08/30/1993 |
| From: | Briggs L, Meyer G NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20056G977 | List: |
| References | |
| 50-317-93-23, 50-318-93-23, NUDOCS 9309080026 | |
| Download: ML20056G978 (13) | |
Text
_ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.
! -
!
l l
U.S. NUCLEAR REGULATORY COMMISSION REGION I SPECIAL INSPECTION REPORT INSPECTION REPORT NO:
50-317/93-23 and 50-318/93-23 FACILITY DOCKET NO:
50-317 and 50-318
.
FACILITY LICENSE NO:
DPR-53 and DRP-69 f
I LICENSEE:
Baltimore Gas and Electric Company
'
Post Office Box 1475 Baltimore, Maryland 21203 FACILITY:
Calvert Cliffs Nuclear Power Plant, Units I and 2 INSPECTION DATES:
July 6 through 30,1993 INSPECTORS:
Larry E. Briggs, DRS Greg S. Galletti, HFAB/NRR
!
Henry K. Lathrop, Resident Inspector 30 6 INSPECTOR:
'
- Erry E. Briggs, Lead Inspec[r L
Date PWR Section, Operations Branch Division of Reactor Safety t
f l
3 Toj D APPROVED BY:
-
8(enn W. Meyer, Chief h
Date l
PWR Section, Operations Branch Division of Reactor Safety l
t 9309080026 930831 DR ADOCK 05000317 PDR
.
I i
SPECIAL INSPECTION REPORT NOS.
q
'
!
50-317/93-23 AND 50-318/93-23
EXECUTIVE SUMMARY i
A special inspection was conducted at the Calvert Cliffs Power Plant from July 6 through July 9,1993, with further in office review from July 12 through 30,1993. The inspection
,
'
was initiated to address NRC Region I concerns that the five events occurring between June 8 and July 1,1993, represented a significant negative trend in performance and a
decline in the plant staff's ability to address and correct operational problems.
The inspection objectives were to analyze in detail the last two events to determine root j
J cause(s) using the guidance of the NRC's Human Performance Investigation Process (HPIP),
l'
NUREG/CR-5455, Volumes I and II; review and assess BG&E analysis of all five events to
,
,
determine adequacy of analysis and corrective action recommendations; and compare BG&E l
results with NRC results.
j l
~
The NRC HPIP method of analysis starts with a preliminary events and causal factors chart (E&CF) to help understand the sequence of events and direct questions. Coupled with the E&CF chart is the SORTM (stimulus, operation, response, team p rformance, and management), which asks questions for each causal factor identified in the E&CF chart in a l
flow diagram, yes-no format. The yes-no answer further directs the analysis to the six basic l
root cause categories. The second part is to collect evidence and interview those involved, including management. The remaining process involves several steps, which fmalizes the E&CF chart (Attachments 1 and 2) and analyzes data to determine root cause(s). Root
causes are broken into six basic categories with subdivisions in each category. The six basic
j categories are: procedures; training; communications; standards, policies and administrative i
controls (SPAC); human-machine interface; and immediate supervision. Two events were i
reviewed, independent of the facility review, by the NRC using the HPIP method of analysis.
i NRC review of the June 25,1993, reactor trip from approximately 3 percent power due to a
,
j low level of minus 45 inches in number 21 steam generator (SG) indicated that all six root causes were involved in the event in varying degrees. The most important three root causes
,
were:
j (1)
Training was less than adequate (LTA). Although training was given, it was not given for low power manual operations of the new feedwater bypass valve controller.
(2)
SPAC was LTA. Operations personnel and system engineers involved with initial startup testing of the new digital feedwater control system did not provide feedback to
management regarding large swings of SG level that were experienced at low power.
(3)
Communications turnover was LTA. The off-going control room supervisor (CRS)
,
provided information that was correct for previous shift operation at 0.1 percent power. However, the oncoming CRS misunderstood the information and thought it
,
W
--
. -.,, -,..
,,,,, _.
..,.,.,
,,,...,w,
,,.,.,,,,,,..,
y,..
- - -
e.
... T
applied to actual plant conditions of approximately 3 percent power. The oncoming CRS was also not aware that power had just been increased from 0.1 to 3 percent power by the previous shift.
NRC review of the July 1,1993, incorrect restoration of the drained No.12 service water heat exchanger (SRW HX) to service following maintenance identified three root causes of the event. They were:
(1)
Immediate supervision (job preparation was LTA). The CRS did not ensure that a job prebrief or discussion was conducted about the method to be used to return the SRW HX to service. The CRS also did not coordinate the evolution, thereby allowing the SRW HX inlet valve to be opened to refill the heat exchanger without prior notification of the control room.
,
(2)
Communications were LTA. Neither the CRS nor the taggers performing the valve manipulations informed other control room personnel that the SRW HX was being returned to service.
(3)
Procedure was LTA. There was no procedure for this infrequently performed operation, although there is a procedure to return the drained salt water side of the
heat exchanger to service, which is a frequently-performed operation.
The inspectors performed an in-office review of the facility analyses of all five events. The facility results of the two events that were also reviewed by the NRC closely paralleled the NRC results, with some differences in terminology. The inspectors reviewed the other three facility event analyses for content, accuracy, and recommended corrective actions. The inspectors found the facility analyses to be very good. The inspectors also took facility developed root causes and fit them to the HPIP categories to determine if there were any indicated common root causes which might indicate a programmatic breakdown. There was not a programmatic oreakdown indicated; however, one root cause was evident in three of the five events. The common root cause identified was immediate supervision, work preparation or prebriefs LTA.
In summary, the facility has performed thorough and complete analyses of the five events, root causes identified and implemented or recommended corrective actions. Facility management appeared to be routinely involved in daily plant operations and expressed concern about the recent events during interviews. Increased emphasis by management in the area of job preparation, prebriefs and the need for a questioning attitude during these activities is warranted to ensure that there is not a further negative trend in the area of plant operations. During the interviews all personnel displayed a positive, forthright attitude.
. - -
,,
,
.
.
DETAILS 1.0 SCOPE
,
The inspection was initiated to address NRC Region I concerns that the five events occurring between June 8,1993, and July 1,1993, represented a significant negative trend in operational performance. The inspectors were tasked with primary and secondary inspection objectives. The first was to review the iwo most recent events independent of the facility review. The review was to use the NRC Human Performance Investigation Process (HPIP)
to determine the root cause(s). The review consisted of personal interviews with those involved, including their supervisors and management, and on-scene observations. The second objective was to compare the root cause(s) determined by the facility against those determined by the NRC. Included in the second objective was a review of the other three events to evaluate the thoroughness, objectivity, and recomiaended corrective actions of the facility's review. The review also compared root causes of all five events to determine if any common rost cause indicated a programmatic bicakdown. A programmatic breakdown was not identified.
The following is a listing of the five events that occurred during the above noted period, the first three are discussed in NRC Report 50-317/93-16 and 50-318/93-16:
On June 8,1993, with the plant (Unit 2) in Mode 3, a reactor trip signal and a
auxiliary feedwater actuation system (AFAS) block signal were actuated while performing auxiliary feedwater (AFW) system testing. The actuation occurred due to
!
!
high differential pressure between the steam generators (SG), which resulted because steam was being drawn from only one SG with the main steam isolation valves closed.
- On June 10, 1993, with Unit 2 approaching criticality and Unit I at 100 percent power, a partial loss of offsite power caused an automatic trip of Unit 2; Unit I was manually tripped.
- On June 11,1993, Unit 1 tripped from 16 percent power due to a high level in the 12C low pressure feedwater heater when its high level dump valve was left in the closed position.
On June 25,1993, Unit 2 tripped from about 3 percent power due to low level in No.
- 21 SG.
On July 1,1993, the No.12 service water heat exchanger (SRW HX) was refilled too
rapidly and almost caused a complete loss of the Unit I service water system when the head tank level dropped from 65 inches to 10 inche ___ _ _ -- _ _____
_-_.
!
t
2.0 INDEPENDENT EVENT REVIEWS
,
2.1 Automatic Reactor Trip Due to Low Steam Generator Level l
'
On June 23,1993, an automatic reactor trip occurred from approximately 3 percent power due a low water level on No. 21 SG. Just prior to shift turnover, the reactor power was raised from approximately 0.1 percent to 3 percent. The new digital feedwater control system, installed this outage, was in automatic controlling SG level on the feedwater regulating valve bypass valves (bypass valves). Shortly after shift turnover, the control room i
operator (CRO) noted that the No. 21 SG level was experiencing a slow oscillation (about 50 minutes for a full cycle) with an increasing amplitude. The CRO and the control room
,
supervisor (CRS) discussed the oscillations and decided to put the bypass valve controller in manual to reestablish level at zero inches. At plus 25 inches level, the CRO placed the controller in manual and pushed the keypad button 4 or 5 times in the close direction. As l
the level began to slowly decrease, the CRO pushed the open keypad button several times (4
~
or 5). The level continued to slowly decrease, and the CRO gave several open commands and observed SG level. This continued for about 45 to 50 minutes until SG level reached i
minus 45 inches, at which point the reactor automatically tripped, per design.
As noted in the executive summary, all six root causes were involved in this event with varying degrees of importance. The tliree most important will be discussed.
Training Irss Than Adequate (LTA)
i l
l
'
During interviews with the per.ennel involved in the reactor trip and training department i
personnel, the inspectors learned that both classroom and simulator training had been conducted for the new digital feedwater control system. The training, however, did not practice actual hands on simulator training in manual control at low power level. The
,
previous feedwater bypass valve controller (as still exists on Unit 1) was analog with a knurled knob and had a analog 0 to 100 percent demand signal meter. The new system has a keypad push button and a small LED bar-type meter, which is difficult to read and does not indicate demand changes until the keypad is pushed quite a few (more than 10) times or the
,
l keypad is held down, which causes an ever increasing rate of change of demand signal. The operator stated that he did not use the demand indicator because it was difficult to read. The
!
operator was also not aware that each push on the keypad resulted in only one tenth of one
percent change in demand signal, which would result in virtually no change in feedwater flow and woald not be visible on the demand indicator.
The simulator had displayed SG level oscillations when the new digital feedwater control system was initially modeled, but they were tuned out prior to operator training. The
,
simulator displayed a smooth control of SG level in auto from startup to full power during l
the training.
l I
.
l l
!
I
__
Standards, Policies and Administrative Controls LTA This weakness coiic rns the lack of employee feedback to management and operations personnel concerning the response of the new digital feedwater control system at low power.
The operating crews and the engineers involved in the initial startup testing of the new system did not insure operations management and other operating crews were aware that large swings in SG level were to be expected at low power levels. Large level swings had been observed during initial testing; however, the oncoming crew, who experienced the reactor trip, had not been involved with the startup testing. Further, from all reports, the system worked well. Training received on the simulator further established the impression l
that SG level would be controlled by the new system in an almost straight line manner with very little level oscillation.
l Communications (shift turnover) LTA l
l During shift turnover, the off-going CRS told the oncoming CRS that the digital feedwater l
system had been working great all day and that it was controlling SG level at i 4 inches l
with a 19 to 26 percent demand signal. The demand signal was being read on a digital j
indicator on the inside of the drawer with the drawer in a raised condition; the CRS was aware of the digital indicator, but the oncoming CRO was not. The oncoming CRS thought
!
i the demand signal information applied to the current reactor power of about 3 percent. The i
information actually applied to a power of 0.1 percent, but that was not made clear at shift turnover. Power had just been raised to increase heat input and SG blowdown rate to get
,
secondary chemistry into specifications.
I When the CRO discussed the SG level with the CRS just prior to placing the controller in
!
manual, the CRS observed that demand was at 30 to 35 percent with SG level at plus 25 I
inches. The CRS directed the CRO to ch'ce the controller in manual and return level to zero l
inches since the indicated demand was higher than required during the previous shift and SG l
level was also higher. The CRS and the CRO both stated that they thought the new digital j
feedwater control system was not operating as they expected.
.
l 2.2 Conclusion The inspectors' review of the facility's Significant Incident Finding Team's (SIFT) report indicated their findings were parallel with the NRC inspectors. Their report was complete and identified all the root causes. The corrective actions recommended by the SIFT for this event were; provide classroom training on the event and hands-on training with the new digital controller, reemphasize expectations for prompt communication of significant operating experience (level oscillations at low power), clarify requirements for manning the feed station while at low power, and enhance the performance of the digital control system.
The SIFT report and its recommended corrective actions were appropriate.
l
,
,
-
.
2.3 Incorrect Restoration of Senice Water Heat Exchanger On June 30,1993, the 12 senice water heat exchanger (SRW HX) was removed from service to repair a heat exchanger (HX) drain valve. The HX was isolated and drained and the 12 SRW head tank was isolated. The 11 and 12 SRW pumps were in service. After repairs to the drain valve had been completed, two safety tagging personnel went to the control room to clear tags and get permission to return the SRW HX to service. The CRS was busy trying to determine which post-maintenance testing (PMT) would be necessary to return a diesel generator to service. The CRS and the licensed safety tagger reviewed Operating Instruction (01) 15 for retuming the 12 SRW HX to service. 01-15 does not address the situation of returning a drained HX to service. The licensed safety tagger and the CRS agreed to open the inlet valve to refill the HX and permission was given to return it to senice. No further discussion concerning the evolution was conducted, and other t
watchstanders were not informed. The CRS returned to his previous task of determining which PMT to perform on the diesel.
After clearing the tags, the safety taggers proceeded to open the inlet valve on the HX without informing the control room. The safety taggers stated that they opened the valve till they heard flow noise then opened the valve an additional 3 to 6 turns. After 2 or 3 minutes, the safety taggers noted that the 11 SRW pump had tripped and decided to close the valve.
,
'
In the control room, the head tank low level and header low pressure alarms had actuated l
and the 11 SRW pump was manually tripped due to indicated low amperage. Header pressure had dropped from 100 to 50 psi, and head tank level dropped from plus 65 inches to i
plus 10 inches. The SRW system was subsequently restored to normal operation. There
,
were no safety consequences as a result of this event. There were three root causes indicated
'
in the review of this event.
Immediate Supervision, Job Preparation LTA When the safety taggers requested permission to remove tags and return the system to service, there was no preevolution briefimg or discussion among all control room personnel to discuss possible consequences of error or exactly how the system should be returned to service. A preevolution discussion could have identified the fact that the HX must be filled slowly to prevent draining the SRW head tank and to remain within the system makeup
,
!
capability.
Communications LTA Neither the CRS nor the safety taggers informed the other control room personnel that the SRW HX was going to be returned to service. Each thought the other was going to tell the control room personnel about the evolution. The General Supervisor, Nuclear Plant Operations, places the responsibility with the CRS to ensure that affected watchstande s are informed of equipment being taken out of, or returned to, service; however, he considers that
f
,
i I
!
it is everyone's responsibility if event-free operation is to be obtained. Communications were LTA when the safety taggers opened the inlet valve to fill the drained SRW HX without informing the control room.
,
Procedure LTA There was no procedure for this specific evolution available. 01-15, " Service Water System," gives guidance for returning a heat exchanger to service but does not address refilling one. During interviews with the safety taggers involved, they indicated that they had refilled the salt water side of the heat exchangers before but had not refilled the service l
l water side. They did not consider the limited amount of water in the head tank or the limited capability of the make-up system. They also stated that there is a procedure for l
l filling the salt water side of the HX, which is a routinely-performed operation. Refilling the l
SRW side of the HX is an infrequently-performed evolution.
2.4 Conclusion The inspectors' review of the facility's Issue Report, IRO-0165-623, indicated their findings
'
i were again parallel with the NRC inspectors. Their report was complete and identified the t
root causes. The corrective actions recommended in the Issue Report were to stress the importance of CRS activity coordination and briefing evolutions, stress safety tagging coordination with control room and plant operators, institute a " memory jogger" checklist for tagging activities, and evaluate need for a procedure for restoration of drained systems. The rccommended corrective actions are appropriate for the event.
3.0 IN-OFFICE REVIEW In addition to the facility reports discussed in Paragraphs 2.2 and 2.4 of this report, the inspectors reviewed the facility's repons of the other three events that occurred in June 1993.
The repons are discussed in detail in NRC Report 50-317 and 318/93-16. The event reports were reviewed to determine their comprehensiveness and if there were any common root causes that could indicate a programmatic problem. Because the facility used methods different from the NRC HPIP method, the identified root causes were translated to fit the HPIP root causes.
3.1 Overall Conclusion The inspectors found the facility reports to be comprehensive and the detailed, corrective actions recommended were appropriate for the events. There were no common root causes for all of the events. However, one root cause was determined by the NRC inspectors to be the primary root cause in three of the events. The NRC HPIP method classified the root cause as immediate supervision, with the subcategory of inadequate job preparation and prebriefs.
_
i i
'
,
i
i i
(
,
The inspectors noted that the operators, staff, and management displayed a positive attitude and were taking appropriate actions to identify common root causes and correct any
!
recognized deficiencies.
4.0 EXIT MEETINGS On July 9,1993, the inspectors met with station management to discuss the NRC activities conducted during the week and plans for the in-office review of the facility reports and comparison of NRC HPIP results with those of the facility for the last two events. On
-
July 30,1993, the exit meeting was held at the Calvert Cliffs station to summarize the NRC conclusions and findings of the inspection, as discussed above. The inspector stressed the need for management to reemphasize the need to hold preevolution briefings and discussions
to further promote an open and questioning attitude.
i
!
>
6 I
!
-
(
.
ATTACInfENT I E&CF CIIART STEAM GENERATOR LOW LEVEL REACTOR TR
.
.
,.
_ _ - - - - - - ~ - - ~
ATTACIIMENT 1 E&CF CIIART STEAM GENERATOR LOW LEVEL REACTOR TRIP
,
I
!
.
,
'
!
l
. -
-.
_
_ _ _ - - _ _ - _ _ - - _ _ _ _ - _ _ _ _ _ _ _ - _ _ - - _
._
. - - _ _ _ _ _
perat r Decides to SG LVL Oscillation Operator Decides to Place FW BYP/VLV Operator Attempts to A
Place FW GYP /VLV in Shif t Turnover in Automatic Control
- -'-
Manual on #21 SG at
- 22 So in manuai at Manuai Increasing Over Time 2s Levei 25" Level Communication FW System Response CRO-CRS Believed inadoquate Misunderstanding at low Power Not System Not Resp.
,4 Man / Machine
"
on Power Level Well Understood Properly SPAC, Training A
A
/
Turnover Training.
CRO-CRS Prior Skill of Craft No Procedure on LTA Understanding Experience (lesson plan w/ old FWCS in Manual Procedure FWCS Used instruction.
LTA practice) LTA Initial SU Test No Training on FWCS at Data Not Low Power Operations Published to Crews SPAC, Training LTA A
Training Practices LTA A
So so isoiation on Both So Fw se vtvs
^ ' * ' *
"
- 21 SG at -35" Level REACTOR TRIP i
to AUTO approx.10 min Lo LVL Pre-trip Alarm
MFW Pump Speed inct after Reactor Trip EOP-1 J
I Lack of Aggr sive Training, SG Lo LVL Trip g;
Supervision, at appr. -45" LTA Level
,
.
_
_.
u
-
.-e.
-a
.
.
ATTACIIMENT 2 E&CF CIIART SRW IIEAT EXCIIANGER RETURN TO SERVICE l
!
,..-.
. - - - - - - _. _
- - _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _ _ _ - _ _ _ _ _ - -. _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ - - _ _ _ _ - _ - - _ _ _ _ _ _ _ _ _ _ _
.. _.. _ _
_--
'
""'
CRS OKs Taggers VLV CRO Not Informed A
Repairs to 012 SRW SRW Hx Complete Lineup for #12 SRW Hx of T.aggers Activities Hx Inlet VLV #127 to on SRW System Restoration Fill Hx No Pre-Brief of infrequent CRS or Taggers Did CRS and Taggers Did Task, No Crew Coordination-Not Inform CRO Of Not Cons. der System New CRS, Task Not i
Previously SRW System Response Performed by Taggers Evolution
/
\\
Supervision, Supervision, Supervision, Communicat!ons Communications Procedure LTA LTA LTA Operator Manually CR Operator
- ""
" "'***"
A Trips NEAR LOSS OF Directs
@
'*"S P
- 11 SRW Pump Due ALL SERVICE Taggers to R
to Amp WATER Close #12 Hx
Decrease inlet VLV #127 Action Directed to Protect #11 SRW Pump B
- 12 Head Tank tvt
- 11 and #12 Completed Fill of Restart #11 SRW Lo LVL Alarm Head Tank
l
'
i i
1
. -.
I i.
l
l i
,
ATTACHMENT 3 PERSONS CONTACTED
!
Facility Personnel
.
<
- C. Cruse, Plant General Manager G. Detter, Director, Nuclear Regulatory Matters
- J. Hill, General Supervisor, Nuclear Plant Operations D. Holm, Assistant General Supervisor of Operations Training B. Montgomery, Licensing, Principle Engineer
- P. Pieringer, Supervisor Operating Experience Reviews
- L. Russell, Manager Nuclear Safety and Planning i
R. Wenderlich, Superintendent, Nuclear Operations l
l NRC Personnel
'
- M. Hodges, Director, Division of Reactor Safety, Region 1
- L. Briggs, Senior Operations Engineer i
I
- G. Galletti, Human Factors Engineer, NRR
{
- H. Lathrop, Resident Inspector, Calvert Cliffs l
- D. Mcdonald, Project Manager, NRR l
- P. Wilson, Senior Resident Inspector, Calvert Cliffs
- Denotes personnel present at the July 30,1993, exit meeting.
,