ML20197F054

Insp Repts 50-010/97-24,50-237/97-24 & 50-249/97-24 on 971016-1122.Violations Noted.Major Areas Inspected: Operations,Maint,Engineering & Plant Support

ML20197F054
Person / Time
Site:	Dresden
Issue date:	12/19/1997
From:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
To:
Shared Package
ML20197E973	List: IR 07100016/2011022 ML20197E968 ML20197F054
References
50-010-97-24, 50-10-97-24, 50-237-97-24, 50-249-97-24, NUDOCS 9712300127
Download: ML20197F054 (31)

See also: IR 05000010/1997024

Text

__ _ _ _ _ _ _ _ _ _ _ - _

- - _ _ _ _ _ _ _ _ .

- _ _ - _ _ __ ___ _ _ _ - ________-__ ____--______ - ______-____

-___ -_.

'

.

U.S. NUCLEAR REGULATORY COMMISSION

'

REGIONlil

Docket Nos:

50 010, 50 237, 50 249

License Nos:

DPR-02, DPR 19, DPR 25

Report Nos:

50-010/97024(DRP), 50-237/97024(DRP),

50 249/97024(DRP)

Licensee:

Commonwealth Edison

Facility:

Dresden Nuclear Station, Units 1,2 and 3

Location:

6500 N. Dresden Road

Morris, IL 60450

Dates:

October 16 November 22,1997

Inspectors:

K. Riemer, Senior Resident inspector

D. Roth, Resident inspector

J. Roman, Illinois Department of Nuclear Safety Resident

inspector

B. Dickson, Resident inspector in Training

J. Ellis, Operator Licensing Examiner, Region ill

Approved By;

M. Ring, Chief

Reactor Projects Branch 1

'

i

I

9712

127 971219

PDR

A

K 05000010.

G

PDR - i

. . . .

__

..

..

-

.

. . -

- -

-.

.

-

- . _ _ - .

-

. _ . _ . - -

- - . - - -

. -

. - - - - -

.

,

i

EXECUTIVE SUMMARY

Dresden Generatin Nation, Units 1,2 and 3 -

i

NRC Inspection Reports No. 60-010/9702d t #iM; 50 237/97024(DRP); 50 24g/g7024(DRP)

{

This inspection included routine resident inspect 6on with augmentation from the lilinois

,

Department of Nuclear Safety and NRC Region Ill.

.

Oserations

The inspectors concluded that the licensee's operational plan for dealing with the

_

l

increased unidentified leakage in the drywell was good. The licensee commenced taking

actions well ahead of TS limits. The planning and execution reflected a careful and

conservetive operating environment (Section 01.2)

!

When faced with conflicting indications between the alarm and the local reading, licensee

personnel relied on the less conservative of the two indications and did not declare the

standby liquid control (SBLC) system inoperable. The licensee operated the piant for

more than 34 hours3.935185e-4 days <br />0.00944 hours <br />5.621693e-5 weeks <br />1.2937e-5 months <br /> without an operable SBLC system, well beyond the allowable

-

8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. Lack of knowledge of the exact locations of the temperature sensor and switch

contributed to this decision. A violation was issued for failing to comply with TSs

(Section 02.1).

The SBLC temperature annunciator procedure did not provide advance woming of

i

exceeding a TS limit. The licensee was aware of this, but had chosen to delsy

implementing a change to the procedure based on operations monitoring of the local

temperature of the SBLC system (Section 02.1).

4

,

The licensee generally performed routine operations in a safe manner (Section 04.1).

+

The inspectors concluded that overall performance during the shutdown and startup was

good. However, the inspectors identified examples of some minor problems in

knowledge of system status, communication, and procedural use (Section 04.2).

The usual response of operators to annunciator alarms was to follow the appropriate

,

annunciator procedure. However, in two instances, the inspectors observed that crews

failed to follow the annunciator procedures completely when an alarm recurred

(Section 04.3).

The Unit 2 single loop activities witnessed by the inspectors were performed well. The

inspectors observed operators enforce three way communicationsi follow plant

!

procedures, and coordinate activities well with other involved departments

(Section 04.4).

-

_

The operators responded well to material condition induced challenges and transients

+

(Section 04.5),

The inspectors concluded that the operations department showed a lack of a questioning

attitude during the September 7 operability surveillance run to declare the Unit 3 high -

pressure coolant injection (HPCI) system operable. The shift displayed a weak

s

2'

.

. _ _ . _ . - . , . _ . . _ _ _ _ _ _ . _ _ . _ _ _ . _ _ _ _ . - - _

.

.

._

l

.

knowledge of the operationalimplications of the HPCI turbine exhaust and vent system

-

and how it was affected by the status of the Unit 3 HPCI system (Section 08.1).

Maintenance

The work perfcrmed on the Unit 2 feedwater control system was poorly planned. The

+

failure to provide adequate work instructions directly challenged the operators. The

history of feedwater system work and resulting transients showed that feedwater work

needed more stringent reviews than those provided (Section M2.1).

The surveille. ice activities observed were satisfactorily completed and met the

+

procedure's acceptance criteria (Section M2.2).

Due to a lack of detail on a sketch provided by engineering and a lack of a questioning

attitude by a maintenance worker, the licensee demonstrated poor performance and work

practices regarding the installation of Temp Alt lll 1197 (Section M4.1).

The inspectors noted several examples of maintenance rework following the forced

+

outage activities (Section M4.2).

For the most part, the self assessments were of sufficient scope and depth. The

+

self assessments were hard hitting and critical of the malatenance process. The

exception was the self assessment on unauthorized modifications. This self assessment

did not have sufficient depth to leam the workers' knowledge of administrative processes

in place to control unauthorized modifications, as proved by an NRC violation issued

shortly after the self assessment was performed (Section M7.1).

The Q&SA organization was satisfactorily monitoring the activities in maintenance. The

audit reports and surveillance were complete, thorough, and critical. The field monitoring

reports were an indication that Q&SA personnel performed sufficient field monitoring

activities (Section M7.2).

Enoineerina

The licensee did not have all vendor information related to emergency diesel generator

.

cylinder test valves. Applicable information sent to another Comed site had not been

forwarded to the licensee. The actions taken in response to a pavious NRC-identified

violation for failing to incorporate vendor information regarding cylinder test valves were

not sufficient because additionalinformation received by Comed 2 months before the

licensee's response to the Notice was not incorporated (Section E2.1).

Plant Support

The licensee improperly determined that a problem experienced at Braldwood Station

was not present at Dresden Station. After discussions with the inspectors, the licensee

concluded the problem was applicable and the licensee formulated corrective actions.

This issue, combined with the licensee's failure to test lights adequately, caused the

inspectors to consider the overall material condition of the lights to be marginal

(Section F2.1).

3

__

.. _ - ___

_ = _ _ - _ _ _ _ _

_ ___ _ _ .

_ ._. _ _ _ . _ ._- _ _ _ _ _

.

Report Details

4

i

Unit 2 was maintained at full power except for short uuration load drops to support routine

surveillance tests throughout the first part of the inspection period. On November 16, the

licensee decreased power to 25 percent and tripped one recirculation pump motor generator

(MG) to perform MG brush replacement and to enter the drywell to add oil to the 2A recirculation

>

pump motor. On November 18, the licensee performed maintenance on the feedwater system

and inadvertently caused a feedwater transient. The licensee held power at 650 MWe pending

,

investigation of the issue.

Unit 3 was near full thermal power at the beginning of this inspection period. Full thermal power

on Unit 3 was not achieved because the main turbine control valve positions were limited to an

average of 85 percent open with no greater than 90 percent open on any one control valve, and

feedwater flow was limited to g.735 Mlbm/h (instead of the approximately g.8 Mibm/h at full

power) as a result of a review of the fuel cycle analysis performed by engineering personnel.

3

These limits remained in effect until the end of the inspection period, On November 1, Unit 3

power was decreased to 300 MWe to facilitate drywell entry to investigate increased drywell

'

leakage. The licensee identified unisolable leakage from the 3B reactor recirculation loop

discharge flow element and manually shut down Unit 3 for repairs. On November 6, the licensee

started up Unit 3.

,

i

I

1. Operations

01

Conduct of Operations

01.1

General Comments

l

Using inspec tion Procedure 71707, the inspectors conducted frequent reviews of ongoing

plant operations. Overall, the conduct of operations was safe and in accordance with

procedures.

During the inspection period, one event occurred for which the licensee was required by

10 CFR 50.72 to notify the NRC, The event and the notification date are listed below;

November 1

(Unit 3) TS required shutdown because of pressure boundary leakage, A

leak of 0.32 gpm from a weld on a tap-off of the B recirculation loop flow

element was found during a drywell entry,

01.2 (Unit 3) Response to Drvwell t.eakaoe

a.

101pection Scope (71707)

The inspectors monitored the licensee's response to indications of a leak in the drywell.

4

l

l

l

. . - . _ . - _. _ . _ _ . , ,-

.

......_,_..,_,,r.m

. -

. -

,

. _ , , , , . _

. - -

, , _ .

-

.

b.

Observations and Findinat

On October 17,1997, radiatinn protection personnel reported to operations that Unit 3

drywell activity increased by a tactor of three. On October 18, operations determined that

liquid going to the drywell floor drain sump increased from 0.50 to 0.67 gpm over

24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Dresden UFSAR Section 5.2.5.5 stated,"in the case of a steam leak,

essentially all of the leak will be routed to the floor drain sump as condensate from the

drywell coolers." The leak rate quickly increased to 0.83 gpm by October 19.

Operations established shutdown limits on total unidentified leakage and changes to

unidentified leakage that were conservative compared with the requirements of TS 3.6.H.

(less than or equal to 5 gpm of unidentified leakage; less than or equal to 2 gpm increase

in any 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period).

The licensee then developed and executed a " Unit 3 Drywell Leakage Plan." The plan

included monitoring, reviews of industry avents, and sampling using the air sampling

manifold system. The licensee also assigned senior reactor operators as owners of the

plan.

Air samplin0 and chemical analysis showed the leak to be primary coolant coming from

the 3B recirculation pump area. Operations determined that a drywell entry was

necessary to find the exact source of the leak.

Operations reduced power to 300 MWe on November 1, then entered the drywell and

readily identified a non isolable leak on the 3B recir.:ulation flow sensing line. The leak

was pressure boundary leakage, so operations immediately commenced a unit shutdown

in accordance with TSs.

c.

Conclusions

The inspectors concluded that the licensee's plan for dealing with the increased

unidentified leaksge was good. The licensee commenced taking actions well ahead of

TS limits. The planning and execution reflected a careful and conservative operating

environment.

02

Operational Status of Facilities and Equipment

02.1 (Unit 3) Standby Li. quid Control System (SBLC)

a.

Inspection Scope (71707)

At 0609 hours0.00705 days <br />0.169 hours <br />0.00101 weeks <br />2.317245e-4 months <br /> on October 27,1997, the SBLC low temperature alarm annunciated in the

control room. The licensee did not determine that the alarm was valid until 1530 hours0.0177 days <br />0.425 hours <br />0.00253 weeks <br />5.82165e-4 months <br /> on

October 28,1997. The inspectors reviewed the licensee's investigation and

troubleshooting.

5

-

.

b.

Observations and Findinos

,

identWication of the TS entry conditions

At 0609 hours0.00705 days <br />0.169 hours <br />0.00101 weeks <br />2.317245e-4 months <br /> on October 27,1997, the SBLC low temperature alarm annunciated in the

control room. Per the annunciator procedure, the alarm was set for 78-83'F. By

contrast, the TS required that the suction piping temperature be greater than or equal to

83'F.

After the low SBLC suction piping temperature alam1 was received, the operators

checked local indication in accordance with the Dresden Annunciator Procedure (DAN);

the local temperature indicator showed 90'F. The operators accepted the local Indication

as valid and assumed that a problem existed with the alarm or the alarm's temperature

switch, so operations contacted maintenance to troubleshoot. In fact, the alarm was

valid, as will be further discussed below. Therefore, the operators chose not to believe

an annunciator without proving that the annunclator was invalid.

Both operations and system engineering personnel were unaware at this time that the

local indication came from a different location on the SBLC suction piping than the alarm

sensor. The locallndication temperature sensor was closer to the SBLC tank and was

showing the correct temperature for its location. The temperature switch that provided

the alarm signal was also correctly alarming because the SBLC suction temperature at

the switch's location was below the alarm setpoint.

During troubleshooting on October 28,1997, the licensen used a surface pyrometer and

found that the SBLC suction piping temperature was only 80'F while the local indication

showed 89'F. Operations then declared both SBLC subsystems to be inoperable. With

both SBLC subsystems inoperable, TS allow 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> to restore at least one subsystem or

12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> after that to be in at least hot shutdown.

Immediate Corrective and Compensatory Actions

Maintenance adjusted the controllers for the heat trace to raise temperature above the

83'F TS requirement. The temperature was increased to above the TS requirement at

16t5 on October 28,1997. Maintenance also observed that the heat trace controllers

appeared to not be wired correctly.

The licensee ran an operability surveillance for the SBLC system, and the system passed.

On October 28,1997, at 2033, operations began to monitor the suction piping

temperatures with a surface pyrometer once a shift on both units.

On October 31,1997, at 0930 during the shift rounds, a non-licensed operator found that

the SBLC suction piping was 80'F. The non-licensed operator (NLO) did not recognize

the significance of the extra readings, so the NLO failed to report the condition to the unii

superviser for more than i hour. The 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> delay was a significant portion of the

8-hour LCO. The licensee adjusted the heat trace controllers and increased the

temperatures to above the TS requirement at 1220 hours0.0141 days <br />0.339 hours <br />0.00202 weeks <br />4.6421e-4 months <br /> on October 31,1997. This was

within the B-hour TS action statement.

6

_ .. _. _ . ... _ _ _ _ . _ _ _ _ _ _

. __ ._ ._ _ . _ . . _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ . .

-

.

1

Cause of Temperature Problems

i

1

The onset of cooler outside temperatures caused the ambient temperature of the SBLC

system area in the reactor building to decrease. This allowed the suction piping to cool.

j

Maintenance personnel concluded that two of the three heat trace circuits had their

i

'

controllers crossed such that the thermostat was sensing temperature on one section of

piping, but controlling the heat. trace on another section of piping. The section of the pipe

that had experienced the low temperature was affected by the wiring problem. The

i

4

licensee corrected the error by the aftemoon of October 31.

i

e

,

i

cause of Procedure Problems

,

'

The alarm setpoints in the annunciator procedure DAN 902(3) 5 G-6 were below the

TS setpoints. The alarm came in at 78'F, whereas TS required 83'F minimum

'

temperature. The low SBLC suction piping temperature was changed as part of the

'

TS upgrade program from 80 to 83'F in June 1996. A decision was made then to defer

i

-

'

the setpoint change until mid December of 1997. However, even without the TS upgrade

'

program, the old alarm would still have been below the old TS requirements. The

deferral was made because the licer$see believed that routine operator rounds'

temperature monitoring (which monitored local Indication, not the true suction

temperature) was sufficient.

'

'

\\

'

Licenses Event Report

On October 31,1997, the licensee identified that unit 3 was outside TS compliance when

'

SBLC suction temperature was below 83'F (PlF# D1997 07873). The licensee

.

subsequently published Licensee Event Report No. 50 249/97-01100, "SBLC was

Inoperable from suction Line Low Temperature due to a Wiring discrepancy in the Heat

'

i

trace Controller Circuit."

The licensee event report (LER) discussed the correcting actions and the event causes.

- ;

4

The LER did not discuss the operators' failure to believe the valid annunciator,

in the LER, the licensee stated that personnel reviewed various SBLC work performed

from 1988 to present and were unable to identify when the miswiring took place.

Event Significance

The UFSAR Section 9.3.5 and the TS bases stated that the SBLC system temperature

was required to be maintained at least 10'F above the saturation temperature of 62'F to

guard against boron precipitation, Since the lowest observed temperatures were above

the 62'F precipitation temperature, as were reactor building ambient temperatures,

precipitation did not take place. The inspectors therefore concluded that the safety

' significance was small.

,

,

! However, the event was significant for operations. . It demonstrated a willingness to

"

assume that a valid alarm was invalid without any supporting data. Additionally, even

n

7

'

>

4

L

sz-x-

=.s

• -~re - s

re o e se - -eme

on w r

e--

-

,<--r--

-

=.se r;= w ew w-u-eme.-,er---

--..-

.-..w--nas

-

r-wr-*

ww---.m

e

3-ee

-

n-t-, - -

w

w---me-m,=,meir-e---w

-e--em

wre.-m-en-3-.,r

-

• e.w--te=m---W

after operations started shiftily checks, operations did not assure that the personnel doing

the checks understood the significance, as demonstrated by a failure of an NLO to report

'

exceeding the TS minimums immediately upon discovery.

Dresder. TS 3.4.A.1.b stated that with both standby liquid control subsystems inoperable,

at least one must be restored to operable status within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> or be in at least hot

shutdown within the next 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

Contrary to the above, between October 27,1997,0609 hours0.00705 days <br />0.169 hours <br />0.00101 weeks <br />2.317245e-4 months <br /> and October 28,1997,

1615 hours0.0187 days <br />0.449 hours <br />0.00267 weeks <br />6.145075e-4 months <br />, both standby liquid control subsystems were inoperable for greater than

8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> and the plant was not brought to at least a hot shutdown condition. This was a

violation (VIO 50-010; 237; 249/97024-01).

c.

Conclusions

An error in wiring led to the SBLC suction piping heat trace not keeping the appropriate

sections of piping above TS minimum temperatures. An installed temperature switch in

the SBLC system correctly sounded an alarm in the control room. Upon investigation,

operations found the local indication to be beyond the TS limits.

When faced with conflicting indications between the starm and the local reading, licensee

personnel relied on the non-conservative of the two ir:dications and o.J not declare the

SBLC system inoperable. The licensee operated the plant for more than 34 hours3.935185e-4 days <br />0.00944 hours <br />5.621693e-5 weeks <br />1.2937e-5 months <br /> without

an operable SBLC system, but the LCO for SBLC was only 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. Lack of knowledge

of the exact locations of the temperature sensor and switch contributed to this decision.

A violation was issued for falling to comply with TSs.

The SBLC temperature annunciator procedure did not provide advance waming of

exceeding a TS limit The licensee was aware of this, but had chosen to delay

implementing a change to the procedure based on operations monitoring of the local

temperature of the SBLC system.

04

Operator Knowledge and Performance

04.1

(Units 2. 3) Routine Operations

a.

Inspection Scqpe (71707. 83822)

The inspectors conducted frequent reviews of ongoing plant operations in the control

room and in the plant. The inspectors also discussed plant status and

pendingevolutionswith shift personnel in the control room.

b.

pbservations and Findings

During routine operations the licensee met procedures and TSs (except for the SBLC

issue discussed in Section 02.1 of this report). Control room manning was adequate and

the operators were not overburdened. No problems were found with indications or valve

lineups. Usually, shift personr'el were aware of plant conditions and operational

requirements were listed. The inspectors observed that the licensee continued to

practice good communications.

8

_

-

.

c.

Conclusions

The licensee generally performed routine operations in a safe manner.

04.2 (Unit 3) Forced Outaae (D3F24) gnd Startuo for Pressure Boundary Leak at

.

t

a.

Inspection Scope (71707_)

The inspectors monitored the performance of operations during activities associated with

the shutdown for and startup from the forced outage (D3F24) to repair pressure boundary

leakage.

b.

Observations and Findinos

Overall, the performance of operations during the Unit 3 startup was good. The

operators' actions were generally characterized by careful panel monitoring, good

communications, and good command and control of the plant. The operators were

challenged by equipment problems during the startup and responded generally correctly.

However, the inspectors noted that the operators' responses to some frequently alarming

annunciators degraded (see Section 04.3 for additional discussion).

Control Room Performance

Overall, the control room performance was good. The inspectors observed that the

lice .Ne was carefully following procedures, maintaining good awareness of the plant,

k..d 8. eping the control room atmosphere quiet and professional.

The unit supervisors (US) and shift managers (SM) held crew briefs at regular intervals to

assure that the crews were fully aware of plant status and plans. The briefs conducted

were thorough. The crews maintained three-way communicetions during the startup.

The performance by the nuclear station operators (NSOs) was generally good. The

NSOs performed the procedures as written. The NSOs were generally very attentive to

the plant indications during the startup. The NSOs preformed frequent and detailed front

panel walkdowns.

On November 1, during the Unit 3 shutdown, the inspectors questioned the US about the

status of the shutdown cooling system. The US was unaware that 3B shutdown cooling

pump was unavailable (the pump was in a testing status). Pumps A and C were available

and the licensee was meeting the TS. This showed an example of inattention to detail

during tumover.

During the startup on November 5, the inspectors reviewed a copy of the startup

procedure and noted that one step sent the user to a wrong step during startup. The SM

reviewed the procedure and concluded that the error was a typographical error. The US

informed the shift manager that the typo in the autherized copy of the startup procedura

hao already been noted and corrected. The inspectors considered the typo to be minor

inattention to detail during procedure revision.

9

.

.

.

l

Also, during the startup on November 5, the inspectors identified an instance of failing to

,

complete an annunciator procedure. This issue is discussed in Section 04.3.

Field Performance

The inspectors observed part of the performance of a procedure for starting the steam

jet air ejector.

The non licensed operators (NLos) performing the procedure followed all the required

procedure steps and administrative requirements, including radiation protection

requirements.

The procedure was not frequently performed. The licensee took advantage of the startup

to train additional NLOs on the required actions. The inspectors consWered this to be

prudent. The NLO doing the task provided careful and knowledgeable instruction

The evolution was not successful due to material condition. The Unit 3 hydrogen

analyzer system could not be placed in service because one train was blocked and the

other train apparently had a slug of water introduced during the shutdown. The NLO

attempting to place the system in service identified the water intrusion problem and

correctly reported the condition to the system engineer (who was pre ent) and to the Unit

Supervisor.

The licensee could not restore either system, so chose to startup the plant directly into

the TS LCO for inoperable hydrogen analyzers.

The inspectors concluded that the NLOs performed the evolution in a careful and

controlled manner. The NLos showed a good ability to recognize abnorrnalindications

for system operating parameters that are entry level conditions for TSs.

On a different shift, the inspectors noted one instance of an unexpected alarm caused by

NLOs energizing plant equipment. The NLOs did not inform the control room immediately

before energi:ing the equipment, but instead radioed the control room immediately after

causing an alarm. After the inspectors asked if the NLOs had informed the control room

to expect ar? alarm, the unit supervisors remir'ded all NLOs to provide advance notice

immediately before energizing equipment.

c.

Conclusions

The inspectors concluded that the overall performance during the shutdown and startup

was good. However, the inspectols identified examples of some minor problems in

knowledge of system status, communication, and procedural use.

04.3 (Units 2. 3) Response to Annuncictgn

a.

Inspection Scope (71707)

The inspectors monitored the operators' use of Dresden Annunciator Procedures, in

particular, the inspectors checked for compliance with annunciators that were repetitive.

10

.

. _ _ _

.

.

,

b.

Observations and Findinas

The normal response of an operator to an annunciator alarm was to complete the actions

listed in the Dresden annunciator response procedure (DAN). The inspectors observed

many instances of operators correctly reviewing and executing the appropriate DAN.

On October 20, the inspectors observed an oncoming NSO receive an alarm on the gland

seal condenser, then clear and reset the alarm without referencing any procedure. The

off going NGO had told the oncoming NSO that the alarm had been repeatedly coming in

due to a material condition problem.

The inspectors informed senior licensee management about the failure to use the

annunciator procedure. The licensee Quality and Skfety Assessment (Q&SA) department

had written problem identification form (PIF) # D1997 07317 on September 28 for a

similar issue. The licensee determined that the significance level of the PIF was

" Condition Not Adverse to Quality" and had closed the issue.

On November 5, during the startup, the inspectors observed the following: the operators

were at a point in the startup that required frequent changes to the range setting for the

intermediate range monitors (IRMs). The material condition of the IRMs was challenging

the operators because the changing of the range switch for IRM 14 was causing IRM

downscale alarms to occur. Shortly after the first spurious downscale alarms, the

inspectors reviewed the backpanel indications and observed one IRM with a "Hi" alarm lit,

and another with alamis for both "Hi" and "Hi-Hi" lit, but the readings for all IRMs were

below alarm setpoints. No IRM associated annunciators were in alarm on the control

room front panels. The inspectors reviewed the annunciatormsponse procedure

(DAN 902(3) 5 C-5, "lRM Downscale," Rev. 06) and noted that it required review of

back panelindications as a subsequent action for the alarm.

When the operators next changed the range of IRM 14, the IRM downscale alarm again

came in. The operators briefly discussed the alarm, noted it was the same IRM, and

concluded that the range switch contacts might be the source and that a work request to

investigate must be written. Although the operators reviewed the annunciator response

procedure, no operator was sent to recheck the backpanelindication. Because the

operators did not check the backpanel, they were unaware that the backpanel indications

showed signs of spiking high.

When the NSO next ranged IRM 14. a half scram due to "Hi-Hi" level occurred. The

operators then reviewed the back panels and saw and cleared the IRM channel alarms.

In addition, the operators bypassed the noisy IRM and called maintenance and

engineering for assistance.

After the half scram, the inspectors asked the unit supervisor if the alarm response

procedure for the previous IRM downscale had been completed. The US said,"No," and

acknowledged that it should have. The inspectors asked the NSO who was responsible

for ranging the IRMs if anyone had reported completion of the alarm response procedure

to him, and he said *No." The inspectors discussed the concem of completing the

annunciator response procedure with the shift manager, and the shift manager

reemphasized to the operators that DANs need to be followed completely, even for

expected or nuisance alarms caused by equipment problems, The shift manager also

11

- - .

_

_

_

__

_.

.

_

.

,

!

l

directed that backpanel tours be done on 15 minute intervals rather than on the usual

1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> interval to detect any other IRM spikes,

.

Procedure DAP 07 50," Conduct of Safe Operations," Rev. O Step 6.3.2 stated," . . .

after receiving unexpected alarms, reference appropriate procedures." Step 6.3.3 stated,

" . . when responding to annunciators . . . follow the annunciator procedure as written."

.

Step 7.2.3 stated to announce an alarm and * Pull out the annunciator procedures (DAN)

l

and take the appropriate actions at directed by the procedures."

The operators' actions on October 26 and November 5 did not meet DAP 07 50. In both

instances, the repetitive alarms caused by material condition problems caused the

operators to become complacent about following the annunciator procedure. Although

the actions were contrary to procedures, no violation was issued because this issue

would be expected to be encompassed by licensee corrective actions to a recently issued

violation for operators not following procedures (50 237; 249/97019-02)

c.

Conclusions

The usud response of operators to annunciator alarms was to foll0w the approprihte

.'

annuriciator procedure.

However, in two instances, the inspectors observed the crews failed to follow the

annunciator procedures completely when an alarm recurred. in one instance the

consequence of not following the procedure was the operators being unaware that an

IRM was spiking "Hi" And "Hi-Hi" as well as downscale.

04.4 Unit 2 Sinole Looo Operations

e,

inspection Scope (71707)

The inspectors monitored the response by the licensee to the discovery of uneven

wearing of *e brushes on the 2B reactor recirculation motor generator (MG) set. The

,

replacement of the brushes required the licensee to secure the MG set, and thus to enter

single loop operations on November 16.

The inspectors observed and reviewed licensee plane for the evolution, the plant

operations review committee (PORC), control room activities, and maintenance work in

the field.

b.

Observations and Findinos

UnN 2 Single Loop Activities

Licensee personnelidentifiod that several brushes on the 2B MG set were wom much

more than the others. The licensee developed plans tu replace the bmshes and inspect

the rings on the MG set. The MG set had to be secured to perform the task. The

licensee elected to do the task on-line by tripping one recirculation pump and operating

the plant in single-loop.

12

. .

,_

_

- _ _ _

_

_ _ -

. _ _ . _ _ _ - _ _ _ _ _ _ _ _ . - _ _ . _ _

_ _ . _ _ _ _ _ _ . _ _

.

i

,

Historica.ly, tripping a pump on-line and restarting was a significant event for Dresden.

As documen6ed in inspection Report No. 50 237/95004(DRP); 50 249/95004(DRP), in

i

<

1995, the 25 teactor recirculation pump tripped because of a technician inadvertently

j

closing the contreeler for the 28 recirculation pump motor generator set temperature -

control valve (TCV) while attempting to repair the 2A TCV Operations reviewed the

.

requirements for a restart of the pump and found that the bottom head drain temperature

!

was needed. However, the bottom head drain had been clogged for mar *y years. The

j

,

operating crew concluded that sitomate temperature indications could be used to meet

,

.

the TS and procedural intent. This determination was made without discussion with

j

senior operations management, station management, or engineering personnel, and

against the advios of an independent site engineering group (ISEG) engineer. This

4

resulted in escalated enforcement action. Since 1995, the licensee improved the material

condition by cleaning out the bottom head drain, and improved the overall conduct of

!

operations. However, the licensee had not performed single loop operations since the

1995 event.

j

The planning for the brush replacement was thorough. The licensee used a wide range

of resources, including the vendor.

{

,

i

.

The review of the plan was also thorough. Thc PORC thoroughly discussed the planned

!

3

maintenance work, possible contingencies, and operational requirements for a single loop

l

i-

condition,

j

.

On November 16,1997, the licensee removed the 28 recirculation loop from service to

perform the brush replacement activities. The inspectors verified that the operators

!

complied with the T8 requirements and closely follow the approved plan. The inspectors

!

,

vonfied that plant parameters were within T8 requirements for recovery of the idle loop

.

and that the operators propert*, retumed the loop to service. The inspectors also noted

>

close coordination and communications between operations department personnel,

maintenance and engineering personnel, and the designated project manager for the

i

evolution.

,

c.

G9DQlusion

1

The Unit 2 single loop activities witnessed by the inspectors were performed well. The

l

inspectors observed operators enforce three way communications, follow plant

'

[

procedures, and coordinate activities well with other involved depa 1ments.

04.5 Qperators Response to Material Condition Induced Transients

E

a.

Inspection Ecoce (71707)

?

- The material condition of the plant caused challenges and transients during the

Inspection period. The inspectors observed and reviewed operator performance in

t

response to these events.

c

3

,

5

l

13

t

!

t

.w_...___.,~.-_2..________.._________

_ _ . _ _ , , , _ , , . . .

, . _ _ _ . _ _ _ , _ . . . .

-

.

b.

Qp.ittyggng.and Findinog

Reactor Feed Pump Ventilation

On November 17,1997, Unit 2 operators noticed an increase in the 2C reactor feed partp

(RFP) stator temperature. The temperature of the 2C stator temperature was

apprnximately 80'C (crocedurallimit of 85'C) and slowly rising, while the stator

temperature of the otner runninD RFP (2B RFP) was approximately 46'C. Operators fp

the field identified that the ventilation damper for the 20 RFP had failed to open.

Operators started the 2A RFP and secured the 2C RFP. The operators placed the 2C

RFP in a Standby lineup and caution carded it for emergency use only. During

subsequent troubleshooting, the damper operated as expected and operators retumed

the RFPs to their original configurations.

On November 18,1997, the operators noticed that the stator temperatures of the 2B and

2C RFPs were trending upwards. An oper2 tor dispatched to the field identified that the

ventilation dampers were not in their expected positions. The individual pump vent

dampers were in the correct positions, however, the recirculation damper was full open.

The operators wired the recirculation damper closed and the exhaust damper open in an

emergency attempt to lower stator temperatures (this was procedurally allowed to protect

equipment). The stator temperatures trended downwards and stabilized around GO'C for

both of the pumps. The licensee documented the occurrence via problem identif: cation

forms (PlF) D199708145 and D1997 08101.

The inspectors concluded that the operators performed well by identifying the unexpected

temperature increase and restoring the RFPs and RFP ventilation system to a stable

condition.

The inspectors noted similaritPs between the problems encountered on August 12,

August 13, and the November 18 ventilation problems, in all cases, the operators were

required to perform emergency downpower maneuvers when reactor feedpump

ventilation problems occurred.

2A2 Flash Tank Level Control Problems

On November 17,1997, control room operators noticed level swings in the 2A2 Flash

Tank. Since the level swings were increasing to the point of automatic opening of the

bypass valve, the operators were concemed with the potential tripping of the heater string

and subsequent impact on feedwater temperature. Licensee personnel performed a

heater bay entry rnd identified that the positioner arm oa 'he level control valve was

broken. Maintenance personnel repaired she level conWI valve and the heater string was

retumed to a normal status. Licensee personnel documented the occurrence via

PIF D1997-08160.

The inspectors concluded that the operators performed well by identifying the unexpected

level swingt early enough to allow for corrective action before automatic tripping of the

heater string.

The inspectors noted that this issue was similar to problems operators encountered with

heatet level controls previously documented in inspection reports 97012 and 97013.

14

- - . -

. -.

-

- - _ - .

-

-_

-

.

l

i

2A Feedwater Regulating Valve

'

On November 18,1997, licensee maintenance personnel were doing work on the 2A

feedwater regulating valve (FRV) when operators notice >d that the 28 FRV had switched

,

'

from submetic control to manual centrol. The operators entered trcnsient level control,

but experienced some d fficulty in matching steam flow and feed flow to stabilize level

since the low flow FRV was still in automatic and was attempting to control level.

Operators were subsequently able to place the low flow FRV in manual control and

restore reactor level to its proper level. During the transient, reactor level went as high as

32 inches (normallevelis 30 inches) and as low as 22 inches. The operators had

direction to insert a manual reactor trip (scram) if level dropped as low as 20 inches. An

inadequate work package and deficient work instructions caused the reactor water level

'

transient (reference Section M2.1 of this report for additional discussion of this issue).

The operators responded well to an unexpected reactor water level transient caused by

improper maintenance work in the field. The operator's successful recovery of water

level, and restoration to the proper level band, demonstrated effective corrective actions

started after operators mishandled a levt,l transient in July 1997 (see inspection

Report No. 50 237/97016(DRS) for more information).

c.

.Qgnqlgig

The operators reonded well to material condition induced challenges and transients.

l

However, the challenges themselves were similar in nature to previous occurrences and

Indicated ineffe:tive licensee efforts to correct known plant deficiencies. The inspectors

were concemed about the equipment problems discussed above since material condition

issues continued to challenge operators. The items above were repeat items and were

similar in nature to items previously documented in NRC intpection reports. The

inspectors concluded that licensee effort to address known material condition deficiencies

,

were not completely effeJive.

08

Miscellaneous Operations issues

!

08.1

(Unit 3) Hioh Pressure Coolant inlection (HFCl) System

1-

a.

inspection Scope (71707)

On September 5, the licensee declared the Unit 3 high pressure coolant injection (HPCI)

l

system inoperable due to a malfunctioning level control! alarm switch in the HPCI gland

seal condenser hotwell. The licensee issued LER 50 249/97-009 to document the event.

l

The inspectors discussed the LER with the licensee, and reviewed and observed

!

subsequent operability surveillances performed on September 7 and September 8 to

l

declare HPCI operable.

l

l

b.

Ob6ervations and Findinas

On September 6, the licerisee isolated the HPCI system steam line in preps,ation for

completing repairs on the HPCI system gland seal condenser hotwell. On September 7,

after replacing the level control / alarm switch in the HPCI gland seal condenser hotwell,

15

l

- - .

.

-

.

the licensee placed the Unit 3 HPCI system back in service and attempted to complete

DOS 2300-03, "High Pressure Coolant injection System Operability Verification."

Immediately after opening the steam supply shutoff valve (3 23013) many exhaust drain

pot high level alarms were received in the control room causing the NSO to trip the HPCI

system turt>lne.

Step I.1 of DOS 2300-33 required the licensee to drain the HPCI exhaust drain pot. The

inspectors reviewed the completed surveillance procedure and noted that unit supervisor

administering this surveillance had initialed the procedural steps as ' conditions met"

(C/M). However, if the condition was met, then the HPCI system would not have tripped.

The inspectors found no entry in the NSO logs nor the US logs that documented the drain

pot being drained. During subsequent interviews, the licensee stated that the US

believed that the condition was met based on information from the previous shift's US.

Specifically, the previous US said that the exhaust drain pot had been drained on two

prior shifts. The inspectors Mquested a copy of the completed procedure in which this

activity was performed to verify this information, but the licensee could not find the

completed procedures. The licensee was unable to produce any documentation to

support of the use of" conditions met"instead of draining the drain pot.

'

According to the system engineer, the drain pot had partially filled with condensed steam

(from the OOS HPCI steam line) which slowly leaked past the steam supply shutoff valve

(3 2301 3). With the drain pot partially filled, there was not enough volume to receive the

condensate normally experienced during HPCI operation, so equalization problems

occurred. The exhaust drain pot was quickly filled, ultimately causing the turbine casing

to fill via the casing drain lines. The cystem engineer added that an addition 61 contributor

may have been the short time (one hour) between clearing the OOS (opening utoam

valves 2301-4 and 23015) and restarting the surveillance (opening valves 23013 and

the turbine stop valve). After proper drainage of the HPCI system exhaust drain pot, the

operability surveillance was completed successfully.

Step F.8. of DAP 09-13 *Procedura! Adherence" stated that " Condition Met (C/M) should

be entered IP an Individual finds that the requirements of a procedure step are already

satisfied, e.g., the step calls for starting a pump; however, the pump is already running."

Based on interviews with the licensee and the review of licensee documentation, there

was no clear evidence that the procedural step was already satisfied. The licensee is

required by TS 6.8.A to implement applicable procedures recommended in Appendix A of

Regulatory GV.de (RG) 1.33, Rev. 2, Feb.1978. Administrative procedures goveming

procedure adherence are recommended in RG 1.33. Contrary to this, the licensee failed

to follow DAP 09-13 guidance for determining if a condition had been met, and

consequently had to trip the HPCI system during a surveillance test manually. Although

the actions were contrary to procedures, no violation was issued because the violation

would be expected to be encompassed by licensee corrective actions to a recently issued

violation for operators not following procedures (50-237; 249/97019-2)

16

..

-

c.

Qanslution

The inspectors concluded that the operations departmerW showed a lack of a questioning

attitude during the September 7 operability surveillance run to declare the HPCI system .

operable. The inspectors also concluded that the shift i;'c,d a weak knowledge of

the operational implications of the HPCI turbine exhaust and vont system and how it was

effooted by the status of the Unit 3 HPCI system,

11. Mainlanance

M2-

Maintonenee and Material Condition of Faellities and Equipment

M2.1 (Units 2 & Si trW of Feedwater Raoulatina Valve (FWRV) Maintenance

a.

Inanection Boone (62707. 93702)

The inspectors reviewed the licensee's response to a level transient that occurred on

Unit 2 on November 18,1997. The review included assessing the licensee's response

and preliminary root causs.

b.

Observations and Findinos

On November 18,1997, with Unit 2 at approximately 80 percent power, maintenance and

engineering were allowed to do maintenance on the 2A FWRV. The work was to address

soms instability exhibited by the 2A FWRV in positions between 30 - 80 percent open by

replacing the linear variable differential transformer (LVDT). The licensee repleoed a

solenoid on the valve's operator during the single-loop operations of November 18, but

the valve continued to exhibit erratic operation, so the licensee was next trying the LVDT.

,

The licensee used WR 97012985 01, "2A FW REG VLV; Valve Exhibits instability -

between 20 and 52 percent replacement.' When a lead was lifted from the LVDT, the 28

,

FWRV swapped from automatic control to manual control. The swap caused a loss of

automatic reaclor pressure vessel (RPV) level control, and cycling of the low-flow

feedwater regulating valve. The operators were challenged, but recovered RPV level

-(see Section 04.4).

The preliminary investigation by the licensee, and presented to the Plant Operations

'

Review Committee (PORC) on November 20,1997, determined that the response of the

feedwater control system was correct. The system was designed to place the FWRVs in

manual mode if the LVDT current loop opens. The automatic modo swap was designed

to fall the valve "as-is"in the nont of a problem with the quality of the signal. This -

feature had not been identified in the work package. The work request was reviewed by

a senior reactor operator before execution, but the SRO also did not realize that the work

on the 2A LVDT would affect the control of the 28 FWRV.

The licensee's investigation team informed :he PORC that the type and number of

reviews done for the WR were correct per DAP 15-06, but the reviews were inadequate,

one of the proposed corrective actions was to have feedwater level control system

corrective type work have a technical review. -

17

.

..l

--

,

- _ _ _ . . __ .______

_._

-__-____

l

h

a

r

,

i

i

!

r

The inspectors noted that some precursor events suggested the need for greater

attention to foodwater level control system procedures, in inspection Repori No. 97012

Section M1.1, the inspectors documented the June is unexpected opening of the

SA FWRV due to a procedure revision error. The inspectors also noted that

PlF# Digg7-07698, "3A FWRV Found in Test Mode," Novenhor 2,1997, desenbod the

- i

i

operators' discovery that the Unit 3 A FWRV had unexpectedy gone into ' Test" mode

i

following the shutdown of Unit 3. The system engineer informed operations that, "the tog

valves will automaticalh go into twt mode on a loss of LVDT signal," and theortrod that -

i

the rapid response of the FWRV following the scram may have caused a momentary loss

!

of LVDT signal to the controller. This had not occurred before, so the iloonsee replaced

!

i

the LVDT oscillator card to reduce the likelihood of a re:urrence. The inspectors

!

recognized that there are a ome differences between the Unit 2 and Valt 3 FWCS,

!

(according to the system engineer, placing Unit 3 FWC8 into test before lifting the lead

would have been sufficient; the Unit 3 LVDTs would not have to be separately placed in

test). However, the inspectors concluded that the incident documented in the PlF

,

demonstrated that LVDT c9n impact the valve configuratinn,

f

,

The history of feedwater level control system problems at Dresden showed that feedwater

maintenance needed thorough reviews. On May 31, igg 6, during testing of the

_

-

feedwater level control system, Unit 2 experienced a level transient of sufficient

!

'

- magnitude to cause the operators to shut down the plant. The event was described in

Inspection report g6006 Section 02.1 and LER 237/g6-00g-00, " Manual Reactor Scram

!

due to Lowering Reactor Water Level due to Automatic Foodwater Level Control System

4

j

Design Deficiency." The LER stated:

"The decisions made during the FWC8 [feedwater control system) testing

/

t

2

!

should have been more conservative . . . The review and assessment of

l

the risks and consequences associated with change could have been

i .

mote thorough. Indicatior.s of the need to strengthen the conservative

i

safety culture are as follows:

b.

It was riot recognized by the testing team that the on line

,

configuration function of the Bailey Network N was a potentially

j

untested function and that it should have been tested prior to

'

relying on it for the logic configuration change."

d.

A review with independent personnel outside the test team was not

[

performed and may have identified the importance of placing the -

'

Master Station into manual prior to performing the evolution. "

Many of the statements in the LER could be directly applied to the November 18

'

transient. In the November 18 transient, the licensee did not recognize that changing the

configuration of the LVDT (by lifting lead) affected the FWCS, and no independent review

of the WR was performed that could have identified the importance of placing the control

- stations h manual,

Dresden Station TS 6.8.a required that written procedures be established, implemented,

i

'and maintained covering the applicable procedures recommended in Appendix A of

.

RG 1.33, Revision 2 February 978. Appendix A of RG 1.33, Revision 2 February 1978,

referenced procedures for the repair or replacement of equipment. The

-

.

18

i

e

w

sue.==

e-

w-wr=+-*--

w--'e

em

-e

--M--

e-vo--,,,r---Nte

-vv-,

-

vv - - - . - + = - = . -

s.$--e,www-.n-y-ww.w

+ - - - * +

v-4+

www

we r= r-- e e==u--

w n -e m er v e ee v=%.e +w rece--ee,ce, w-w irr

wev*

4

work request 97012985-01 was inadequate to perform the repairs of the feedwater

control system. Consequently, an unexpected reactor pressure vessellevel transient

occurred, and the operators had to take manual control of the feedwater level control

system.

Although the preparation of the procedure was inadequate, no violation was issued

,

because the licensee had not yet responded to a recently issued violation for inadequate

procedure and corTective actions for tha previous violation would be expected to

encompass this issue (50-237; 249/97019-01).

c.

Conclusions

The work performed on the Unit 2 feedwater control system was poorly planned. The

failure to provided adequate work instructions directly challenged the operators. The

history of feedwater system work and resulting transients showed that feedwater work

needed more stringent reviews than those provided.

M2.2 Surveillance Performance

a.

Inspection Scope (61726)

The inspectors witnessed and performed a documentation review for surveillance

DIS-263-05, Rev. 8, " Unit 2 Anticipated Transient W.9out Scram (ATWS) Recirculation

Pump Trip and Attemate Rod Injection (RPT/ARI) and Emergency Core Cooling System

(ECCS) Level Master Trip Unit (MTU) and Slave Trip Unit (STU) Channel Functional

Test."

b.

Observations and Findinal

During the performance of the surveillance, communication was established between the

control room and the auxiliary electric equipment room. Three way communications were

used in the performance of the surveillance. The inspectors researched the Updated

Final Safety Analysis Report (UFSAR) and TE s. The puriodicity of the surveillance was

compared with the TS requirements. The survoillance acceptance criteria listed in the

procedure met the TS requirements. The surveillance was completed in a professional

manner.

c.

Conclusion

The surveillance was satisfactorily completed and met the procedurn's acceptance

criteria. The surveillance was completed in a timely manner and met regulatory

requirements,

M4

Malntenance Staff Knowledge and Performance

M4.1 Condensate and Feedwater System

a.

inspigtion Scope

The inspectors reviewed the worn instructions and interviewed licensee personnel

regarding the completion of Temp Alt Ill 1197 (Reference WR#970082873), and

19

-- _ _

-.

.-

.-.

-

_ --

. - - - -

- .

.

reviewed the information about URI 50-010; 237; 249/97013-04 provided by the licensee

i

in Hs reply to a Notice (ref. JSPLTR: 97 0195).

b.

QhattYah9fundf0f091

On August 3, in an attempt to alleviate repeated draining and flashing of the

3D2 HP heater high level switch (3 3541 558) reference leg, the licensee attempted a

l

temporary alteration (Reference Temp Att lll 1197 and WR#970082873). Maintenance

'

personnel connected the reference jeg to the high-pressure side of the heater rather than

the low pressure side as diagramed in the work instructions. The temporary ahoration

'

work failed its acceptance criteria during performance of functional testing. The licensee

discovered that the reference leg had been connected to the wrong pipe.

Results of investigation and inspector interviews of maintenance and engineering

personnel suggest workers became confused due to labeling problems during installation

and a poor sketch from engineering included in the work instruction.

I

The workers' action to continue of with work despite this confusion demonstrated poor

work prtctices, since DAP 05 08 " Control of Temporary System AHerstions" stated that,

i

"Whi!e performing a temporary aHeration, if field conditions are different from shown in

wort, psckage then stop and contact the preparer." The inspectors verified that this had

+

not occurred.

'

The licensee reported that the maintenance workers who performed the work were

coached by the department supervisor and the engineer who mad < the poor , ketch

agreed that the sketch should have provided clearer guidance. Tin licensee initiated a

Nuclear Tracking System (NTS) ltem (NTS 237100-97-01304A) to coach the engineer

regarding engineering standards. The involved workers were contract employees and

were no longer working at the site.

The licensee was required by TS 6.8.A to implement applicable procedures

recommended in Appendix A of RG 1.33, Rev. 2, Feb.1978. Adherones to administrative

procedure goveming temporary alteration is recommended in RG 1.33.

Since this self revealing and corrected violation occurred while working on

nonsafety-related equipment, and did not adversely affect the safe operation of the

reactor, the violation is being treated as a Non Cited Violation, consistent with

Section Vll.B.1 of the NRC Enforcement Policy (NCV 50 237/97024-02(DRP)).

c.

Conclusion

Due to a lack of detail on a sketch provided by engineering and a lack of a questioning

attitude by a maintenance worker, the licensee demonstrated poor performance and work

practices regarding the installation of Temp Alt lll 1197.

<

M4.2 Rework (62707)

The inspectors noted several examples of maintenance rework following the forced

outage activities.

20

-

. . - . . .

-

.

-

-

. . . - . . - . .- . . . . . -

, - --.

.

.

The licensee identified that the correct post maintenance testing (PMT) was not

performed following the weld repairs on the recirculation piping. As a

consequence, the licensee had to redo the weld repair; almost 300 additional and

unnecessary millirem were received by station personnel as a result.

Prior to the shutdown, the number 3 bypass valve showed conflicting position

indication. Position indication repair was included in the forced outage schedule.

After the outage, however, the valve still showed conflicting posit.on indication,

Prior to the forced outage, intermediate range monitor (IRM) 16 experienced

o

erratic indication. The licensee's repair efforts were unsuccessful; following the

outage maintenance work, the IRM still displayed erratic indication. The licensee

subsequently c'etermined that moisture was present in the IRM cable and the IRM

would require further repairs during the next refueling outage.

The consequences of the maintenance rework issues were not severe; husever, they

were indicative of weak maintenance performance. In the case of the IRM, the scope of

the work was not correctly diagnosed; a situation similar to previous concems about

correct problem diagnosis documented in prior NRC inspection reports.

M7

Quality Assurance in Maintenance Activities

M7.1

Licensee Self Atitssments Activities (40500)

a.

Inspection Scope (71707. 83822)

The inspectors reviewed three licensee self-assessments in the maintenance area. The

self assessment's scope, depth, and conclusions were reviewed.

b.

Observations and Findinat

The following licensee self assessments were reviewed:

1.

NIf assessment of the maintenance process for the potential of unauthorized

modifications. The assessment dates were January 7,1997, to January 29,1997.

The objective of the assessment was to review a sample of completed corrective

maintenance work requests from the past four years for unauthorized

modifications.

A ssmple of 315 work requests was reviewed. The sample size was determined

using American National Standard Sampling Procedures and Tables for

Inspection (ANSI /ASQC Z1.4-1993). One unauthorizea modification was

identified in which a cover was bolte1 over a temperature switch. The

assessment also found one instanca where a vendor manual was not properly

updated. Using a table in ANSI /ASQC Z1.4-1993, it was determined with a

99.85 percent confidence level that corrective maintenance work requests

completed during the period did not result in unauthorized modifications. Based

on the confidence level it was concluded that no further actions were roquired.

21

__

. _ . . . _ . _ . _ _

. . _ . . _ . _ . _ . _

__.

_ . _ . _ _ _ _ _ _

-

.

The self assessment did not identify that maintenance personnel t'eoded

additional training in modifications. Shortly after the self assessment, the NRC

- Issued a violation for an unauthortrod modification (an installed and unattended

digital voltmeter on a safety-related battery).. The response to the violation stated

,

the unauthorized modification occurred due to the individual n0t being fully aware

j

of the requirements for temporary modifications. The response to the violation

!

stated maintenance p%f onnel would be trained on the temporary modificat6on

!

procedure. The inspeaors therefore concluded that the self-assessment failed to

!

dentify the training weakness.

!

,

2.

Self assessment of the conduct of maintenance. The assessment dates were

i

'

February 1,1997, through February 28,1997. The objective of the assessment

l

was to determite if work practices used during the conduct of maintenance were

[

in keeping with t,.e highest industry standards,

j

Procedures, policies, the problem identification form (PIF) data base, .

I

main'enance monthly performance reports, and other documents were reviewed

'

by the self assessment team. In addition, interviews and in the-fleid obtervations

were performed.

There were 28 findings listed in the self assessment. They could be summartzed

as findings of work start delays and procedural nonoompliance events durit's

planning, scheduling, work performance, and work documentation. There were no

,

strengths documented in the self assessmem. The report stated that all areas

i

evaluated during the self estessment were found to need improvement. The _

report further stated, "The overall performance trend found in most cases was

either stagnated, cyclic, or had a declining performance trend."

3.

Self assessment of maintenance cerformance indicators. The assessment dates

,

were September 15 through September 26,1997. The objective of the self

assessment was to perform an effectiveness determination of the maintenance

performance indicators.

Interviews with maintenance management were conducted to get their impression

of maintenance performance indicators.

l

The findings of the self assessment were that the maintenance department has

failed to: 1) take ownership and communicate the expectations and requirements

for the measurement of performance indicators, and 2) provide for the adequate

resources for properiy maintaining and evaluating the performance indicators.

'

There were no strengths documented in the self assessment.

c.

Conclusion

For the most part the self assessments were of sufficient scope and depth. The

self assessments were hard hitting and critical of the maintenance process. The

exception was the self assessment on unauthorized inodifications because it did not

3

identify that some workers' knowledge of modification requirements was deficient.-

'

'

22

i

=.- .- .

= - - .. -

-

.. - -.- - - -.:-.- -.

=

--

.

.

_

.

M7.2 Lleenste Quality and_ Safety Assessment Activities f40500)

a.

Inspection Sco,pe f71707. 84822)

The inspectors reviewed licensee Quality and Safety Assessment (Q&SA) activities in tne

maintenante area. Documents reviowed included audit reports, completed surveillance

reports, and field monitoring reports.

b.

Observations and Findinot

Q&SA audit reports are documentation of in depth reviews of se'ected areas. The audits

are usually conducted by a team of Individuals. In the maintenance area the inspectors

reviewed the fo! lowing audit reports:

QAA 12 96-09,

Unit i decommissioning

QAA 12 9610,

ASME code / maintenance / contractor activities

QAA 12 97 02,

Staffing and training

QAA 12 07 0/,

Corrective actions

QAA 12 9710,

Outage activl'..a

QAA 12 9715,

Station blatkout system

The audit reports contained many negative findings including corrective action records

(CARS), PlFs, and recommendations. The audit reports also included some positive

comments. The reports did not back away from criticizing licensee practices and did not

focus on unimportant details.

The Q&SA surveillance reports are of much less depth than aun,ts. They are usually

performed by a few people and focused on a specific atea or topic. The following

surveillance reports were reviewed:

QAS 12 96 01,

Vendor technical information program

QAS 12 96-04,

instrument out of tolerance trend

QAS 12 96-06,

Material condition on unit 2 east low pressure coolant

injection comer room

QAS 12 96-17,

GE HGA relay investigation

QAS 12 96-26,

Electrical bus 331 maintenance

QAS 12 96-33

Reactor building ventilation

QAS 12 96-30

Rework / repeat work

QAS 12 96-43,

in-place repair of 3B reactor recirculation pump motor

QAS 12 97-16,

Commitments for Meriin Gerirt circuit breakers

QAS 12-9719,

Scheduled adherence during Ihe refueling outage

QAS 12 97 25,

NRC Generic Letter 96-01," Testing of Safety Related Logic

Circuits"

QAS 12 97 26,

Corrective actions to improve electronic work control

The surveillance reports, while of less depth than the audit reports, still contained many

good recommendations and findings. The reports were critical and well balanced.

The Q&SA field monitoring reports (FMRs) documented the observations the made while

touring the station. They were short and stated what the Q&SA individual observed and

23

_ _ - _ - _ _ - _ _ - - - - -

.

concluded. The inspectors reviewed 112 FMRs made from June through October 1997.

'

The inspectors considered the number of reports to show an active Q&SA organization.

c.

Qonclusion

The Q&SA organ!zation was satisfactorily monitoring the cctivities in maintenance. The

audit reports and surveillance were complet6, thorough, and critical. The field monitoring

reports were an indication that Q&SA personnel performed sufficient field monitoring

activities.

M8

Miscellaneous Maintenance issues

M81.1 (Closed) LER 50-249/97-09-00: HPCI System Declared Inoperable Following Gland Seal

Leakoff Condenser Hotwell High Level During to Drain Pump Stop Switch Failure. This

LER documented the self-revealing failure of the Unit 3 HPCI system during routine

surveillance testing. The failure occurred on September 5,1997, and the HPCI system

was restored following repairs on September 8. This issue was discussed in

Report 97019 and in Section 08.1 of this report. This issue is closed.

i

M8.2 (Closed) URI 50-010: 237: 249/97013-04f DRP): Failure of Temporary Alteration 111-11-97

to Mset Acceptance Criteria Requirements During Performance of Functional Testing Due

to Maintenance Error. This item was discussed in Section M4.1 of this report. This item

is closed.

Ill. Enaineerina

E2

Engineering Suppott of Facilities and Equipment

E2.1

Enaineerina Support of Emeraency Diesel Generators

a.

Insoection Scope (37551)

The inspectors reviewed the licensee's awareness of and compliance with the

recommendations for torquing cylinder test valves on the emergency diesel generators.

The inspectors also reviewed the licensee's corrective actions for an event that took

place last year when a cylinder test valve was ejected during a Unit 3 EDG surveClance

test.

b.

Observations and Findinas

On November 24,1996, a cylinder test valve was ejected from the Unit 3 EDG during a

surveillance test. Sections 02.2 and E4.1 of Inspection Report No. 96014 documented

the inspectors' review of the issue, and the discovery by the inspectors that the current

vendor information regarding the cylinder test valves was not entered into the licensee's

Vendor Equipment Technical Information Program (VETIP). The NRC documented the

failure in violation VIO 50-237; 249/96014-02.

On November 12,1997, personnel at the LaSalle station found a loose cylinder test

valve. During review of the LaSalle event, the inspectors at LaSalle found vendor

24

-__________-____ _ _ _ -

.

--

.

..

.

-.

- _ - - - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

<

.

I

information that was not incorporated into LaSalle procedures, although it was in the

LaSalle vendor manuals. The information was a letter dated January 2,1997, from Klene

Diesel Accessorie s, Inc., to the diesel system engineer at Quad Cites Station. The

letters subject was, "V-24A/AX-8SA, Indicator Valve / Adapter Assemt:ty Installation in

Electromotive Engines," and the letter transmitted the recommended installation

instructions of the cylinder Indicator (or test) valve and adapter.

The inspector discussed the contents of the letter with the system engineers at Dresden,

and the system engineers said that they had not received the letter. However, they

believed that the routine work request to verify the torque on the cylinder test valves

developed in response to the November 24,1996, event was sufficient.

The inspectors reviewed WR # 970087465-01, "[ Unit 2) Engine Standby Diesel

Generator, Verify Torque on Cylinder Test Velves", last completed October 1,1997. The

task required " Torque packing nut to 65 ft/lbs, plus or minus 1 ft/lb. Torque values

acquired from V TIP D1164." V-TIP D1164 war not related to cylinder test valves, but

V-TIP D1163, pages 215, stated that for the original cylinder test valves,"If a cylinder

test valve is leaking, check that packing nut . . . has been torqued to 81-88 Nm (60-65 ft-

Ibs), if nut has been overtightened, change seat, , , . , and correctly torque packing nut."

The January 2,1997, letter from the cylinder test valve vendor stated that "ths adapter

should be installed into the engine and tightened with 45 to 50 ft-lbs of torque, and then

the packing gland nut should be tightened against the packing gland enough to hold it in

place. The V-24A Valve is then installed . . . and tightened to 75 ft-lbs of torque."

The licensee was unable to find any documentation that showed that the cylinder test

valves were installed in accordance with the January 2,1997, letter. The licensee wrote

action requests to verify the torques, and the verifications had not been completed at the

end of the inspection period.

Inspection Report No. 50-237/96014(DRP) issued Violation 50 237/96014-02 for failing to

incorporate vendor information regarding the cylinder test valves into the VETIP manuals.

The licensee's response (JSPLTR: 97-0045, dated March 6,1997) stated that the failure

was attributed to a system engineefs failure to take action to update the technical manual

when a new component we= installed in the dieselin January of 1996. Also, as of

February of 1997, the venovr stated that no technical manual change will be issued

addressing the replacement valve as new diesels will be delivered with the old style

cylinder test valve installed. The response did not state if the vendor of the new-style

valves .nlanned to issue any vendor information. The response also discussed Dresden-

site specific actions taken to assure all vendor information received at Dresderi was

incorporated.

The system engineer gave the inspectors a copy of E-mail from Comed corporate

engineering regarding the LaSalle cylinder test valves that concluded that "the deviation

from the vendors procedure in this case would not affect the operability of the diesels."

The inspectors noted that the licensee's diesels had all passed their surveillances (except

for operations-based problems documented in report 97019), and that no visual evidence

of loose cylinder test valves was found during routine walkdowns, and corcluded that the

lack of incorporation of vendor information did not result in an immediate operability

concem.

25

_ _ _ _ __

-

- - - - _ _ _

_

-_ - _ _ _ _ _

_ _ _ .

_ _ . .

._

.

I

The bspectors determined that the response to the Notice was insufficient because

vendor information regarding the cylinder test valves was still not captured into the

i

VETIP The unincorporated information was dated January 2,1997, and was sent to a

Comed site more than two months before the licensee issued the response to the Notice.

Therefore, the licensee should have identified the vendors information before issuing the

response. Acccrdingly, the status of violation VIO 50-237/96014 02 will remain OPEN

pending review of how the vendors information was sent to one Comed licensee, but not

another.

c.

Conclusion

The licensee did not have all vendor information related to emergency diesel generator

cylinder test valves. Applicable information sent to another Comed site had not been

forwarded to the licensee. The actions taken in response to a previous NRC-identified

violation for failing to incorporate vendor information regarding cylinder test valves were

not sufficient because additionalinformation received by Comed 2 months before the

licensee's response to the Notice was not incorporated.

IV Plant SuDDort

F2

Status of Fire Protection Facilities and Equipment

F2,1

Safe Shutdown Emeroency Liohtino Material Condition

a.

Inspection Scope (62707)

The inspectors reviewed Inspection Followup item (IFI) 50-237; 249/97019-03. The

inspectors also reviewed emerging problems with all Unit 2 safe shutdown (SSD)

emergency lights being declared inoperable due to a missed surveillance test.

b,

Observations and Findinos

The inspectors reviewed problems with the installation of batteries within many

emergency lights, The lights were missing either the battery tray and/or the pressure bar.

Within the emergency light assembly, the battery rests on the battery tray and is pressed

against the front of the unit by the pressure bar Without either of these items the battery

is free to move around inside the emergency light assembly This could affect the

seismic qualification of the emergency lights. After being notified of the problems with the

emergency light, licensee personnelinitiated a PlF (# D1997-07316) and notified site and

design engineering for assistance with the seismic concems,

While interviewing electrical maintenance and engineering personnel, the inspectors

asked if they were aware of a similar problem in the past at Braidwood station

(ref. Braidwood PlF# 456-201-97-0601 dated March 6,1997). The licensee first M vm

that the problem at Braidwood was a different problem for which Dresden had been

evaluated. The licensee personnel mistakeniy believed the problem from Braidwood was

with the "J" hook that secures the SSD emergency light to the mounting shelf. Arter the

inspectors described the Braidwood problem clearly to Dresden personnel, the licensee

concurred that the problem was applicable to Dresden.

26

_ _ - _

..

.

.

_-.. .-. - ...- -.-

-- ~ ._- - - - . -

. _ - - - .

.- -

- .

. .

f

~e.

.

b

i

l

r

L

The inspectors reviewed a memo (Doc ids 0005537204), written by Comed Design

Engineering, which states the SSD emergency lights do not have to be seismically

,

qualified but only have to address seitmic interaction concoms. T he memo states the

missing battery tray or the pressure bar would not aNoct the seismic interaction of the

SSD emergency lights.- From discussions with the Design Engineering Structural Lead

-

' who approved the memo, engineering judgement was used to decide that the battery

would not be thrown from the SSD emergency light assembly during an earthquake.-

-

Electrical maintenance personnel stated that, based on there being no seismic cencom,

. they plan to determine the requirements for property installing the batteries in the SSD -

. emergency lights, during the performance of the next quarterty SSD emergency light

,

surveillance. They then plan to correct the battery installation during the following

i

j-

quartetty surveillance,

- During a Q&SA audit it was oetermined that the battery discharge surveillance on all the

unit 2 SSD emergency lights was over due. With the surveillance past its due date, the

,

licensee declared all SSD emergency lights on unit 2 inoperable. Dresden Administrative

I

L

Technical Requirement (DATR) 3.2.3.1 was entered. The DATR stated that inoperable

SSD emergency lights must be retumed to operable within 7 days or to establish backup

lighting, if this could not be done, then the equipment illuminated by the SSD emergency

!

lights must be treated as inoperable. The station decided to replace the batteries in all

l;

- the unit 2 SSD emergency lights with new batteries rather than do the discharge test.

- The batteries were successfully replaced within the 7-day DATR. During the battery

replacement, the parts requirement document to correct the battery installation was

generated.

Also, during the Q&SA audit it was identified that certain areas in the plant, where manual

actions are required in SSD procedures, did not contain SSD emergency lights. These

areas are required to have SSD emergency lights. The licensee began fire watches in

the effected areas until temporary SSD emergency 1:ghts could be put in place. Work is -

l

in process to put permanent SSD emergency lights in place in these areas.

i-

c.

ConclusioD1

i

r

The licensee improperly determined that the problem experienced at Braidwood Station -

!

was not present at Dresden Station. The licensee determined through engineering

j~

judgement that the battery being improperly restrained inside the SSD emergency lights -

did not aNect the seismic interaction of the equipment. The licensee planned to install the

batteries in the SSD emergency lights correctly. The licensee dealt with SSD emergency

light issues from a Q&SA audit quickly. However, due to all the problems with SSD

. emergency lights, the overall material condition of the lights was considered marginal.

'

,

27-

,

=,-,e

..r.

---

w

..-

u- .

.

.

.

.

m

m

.

.

. .

.

m

.

_ . , _ _

. . .

_

_ _

_ _

_..

_.

. . _ .

-

-

. _ .

.

.

!

V, Mananoment Meetinas

- X1 Exit Meeting Summary

The inspectors presented the inspection restits to members of licensee management at

the conclusion of the inspection on November 22,1997. The licensee acknowledged the -

findings presented. The inspectors asked the licensee whether any materials examined

during the inspection should be considered proprietary. No proprietary information was

identified.

,

i

.

28

.

-

._.

.- . .-

.

-

-

-

-

- -

-

-

-

_ _ . -

. __.

.

._.

_

-.

.

. ~ . -_

.,

-

-

m.

.

PARTIAL LIST OF PERSONS CONTACTED

Licensee

• G. Abrell >

NRC Coordinator-

• L. Weir

Design Engineering Superintendent-

• D. Ambler

Acting Regulatory Supervisor

a

• J. Tenz

System Engineering Safety Group Lead

• R. Peaks

Programs Engineering Supervisor

• S. Perry.

Site Vice President

• B. Holbrook

Training Manager

• S. Kuczynski

Shift Operatoins Supervisor

• R. Whalen

Plant Engineering

'

'

• D. Willis

Electrical Superintendent

• W. Lipscomb

SVP Staff '

• C, Richards

Audit Supervisor

,

.

~ *M. Friedmann

Lead Health Physicist

• Present at exit meeting of November 21.

!

i

i

5

29

,

.-,

%

.-

. ~ - - - - - - - - , , .

- - - _ . - -

-

--

.

. -.

_

-- .

.

-

.

.

. .~

.

INSPECTION PROCEDURES USED

~

- Inspection Module: 71707f Operational Safety Verification

Inspection Module: 83822

Radiation Protection -

Inspection Module: 62707

Maintenance

inspection Module: 61726

Surveillance Observations

inspection Module: 40500

Effectiveness of Licensee Controls in Identifying, Resolving, and

Preventing Problems

ITEMS OPEN, CLOSED, AND DISCUSSED

Opened

50-249/97024-01

VIO

SBLC inoperable in excess of TS times.

50-237/97024-02

NCV Failure to install temp att correctly,

Closed

50-249/97-09-00

LER

HPCI System Declared inoperable Following Gland Seal Leakoff

Condenser Hotwell High Level During to Drain Pump Stop Switch

Failure.50-010;237;249/

97013-04

URI

Failure of Temporary Alteration 111-11-97 to Meet Acceptance

Criteria Requirements During Performance of Functional Testing

Due to Maintenance Error.

50-237/97024-02

NCV Failure to install temp alt correctly.

DitMSMd

50-237;249/

97019-03

IFl

Review of the seismic requirements '- tr,e emergency lights.

50-249/97-011-00

LER

SBLC was Inoperable from Suction Line Low Temperature due to a

,

Wiring discrepancy in the Heat trace Controller Circuit

237/96-009-00

LER

Manual Reactor Scram due to Lowering Reactor Water Level due

to Automatic Feedwater Level Control System Design Defic:ency

"

50-237;249/

96014-02

VIO

Failure to enter EDG test valve tech info into VETIP

30

-

. .

.

.

.

.

.

.

... _.

-

-

. .

. - . . - - . .

.

. .

.

- . . -

,

1

-

LIST OF ACRONYMS USED

DAN-

Dresden Annunciator Procedure

DAP

Dresden Administrative Procedure

,

DATR

Dresden Administrative Technical Requirement

DEOP

Dresden Emergency Operating Procedure

DGA

Dresden General Abnormal Procedure

.

DOA'.

Dresden System Operating Abnormal Procedure

DOP

Dresden System Operating Procedure

HPCI

High Pressure Coolant Injection

IFl

Inspection Followup item

IPE

Individual Plant Evaluation

IR

.

Inspection Report .

ISEG

Independent Site Engineering Group

ISI

inservice inspection

LCO

Limiting Condition for Operation

LER

Licensee Event Report

LPCI

Low Pressure Coolant injection

NCV

Non-Cited Violation

NLO

Non-licensed Operator

- NOV

Notice of Violation

'

NRC.

Nuclear Regulatory Commission

NRR

Office of Nuclear Reactor Regulation

.

<

L

NSO .

Nuclear Station Operator

NSWP

Nuclear Station Work Procedure -

'

OE '

Office of Enforcement

Ol

Office of Investigations -

OOS

Out-of Service

PIF

Problem identification Form

PORC

Plant Operations Review Committee

RUFSAR

Revised Updated Final Safety Analysis Report

F

Q&SA

Quality and Safety Assessment

QC

Quality Control

TS

Technical Specification

'

VIO

Violation

WEC

Work Execution Center

WR

Work Request

.

1 ~

31

i

.

.

.