ML20210T726

Mgt Meeting,10CFR50 App R

ML20210T726
Person / Time
Site:	Haddam Neck File:Connecticut Yankee Atomic Power Co icon.png
Issue date:	09/03/1986
From:	NRC
To:
Shared Package
ML20210T709	List: ML20210T706 ML20210T719 ML20210T726
References
NUDOCS 8610090161
Download: ML20210T726 (32)

Text

.~

s

~

ATTACHliEilT 1

• * * *

• CONNECTICUT YANKEE ATOMIC POWER COMPANY * * * *

• Haddam Neck Plant September 3,1986 Management Meeting 10 CFR 50 Appendix R 5#555 8610090161 860926 PDR ADOCK 05000213 0

PDR

3 APPENDIX R MANAGEMENT MEETING SEPTEMBER 3,1986

- CONTENTS -

Page Introduction 1

t indings 2

Potential Violation Item 86-17 Control Room Halon System Test 3

Unresolved item 86-17 Halon System in Existing Switchgear Room 8

Potential Violation Item 86-17-03 -Inadequate Cooldown Analysis 11 Unresolved Item 86-17 Control Room Habitability 14 Unresolved Item 86-17 Component Cooling Water System 17 Potential Violation Item 86-17-06 -Inadequate Breaker Coordination 21 Potential Deviation item 86-17 RCS Loop Bypass Valves 24 Potential Violation Item 86-17-08 -Inadequate Emergency Lighting 27 Follow-Up Actions 29 Summary 31 Chronology of Previous Correspondence 32 l

1 l

9

' ~

INTRODUCTION During the week of June 16 through June 20,1986, the NRC Staff conducted an Appendix R inspection. At the exit interview, the Staff summarized its results and identified several action items. CYAPCO responded to the action items and provided a schedule for their resolution in a letter transmitted on July 3, 1986. The Staff subsequently notified CYAPCO of four potential violations of Appendix R involving:

(1) the control room Halon system's performance in a test; (2) cooldown capability using the steam generator vents; (3) breaker coordination setting procedures; and (4) emergency lighting adequacy.

A potential deviation was also identified concerning a commitment to lock open circuit breakers for certain Reactor Coolant System Loop Bypass valves.

Several other unresolved items were also mentioned in the inspection report.

On August 19, 1936, CYAPCO responded to its July 3 action items.

This response discussed the topics mentioned in the inspection report. On August 29, 1986, CYAPCO provided its response to the report. The discussion herein is based on these two CYAPCO submittals and the Staff's findings listed below.

wm m Om9v i

W m'w wmm1 s1 a -a

-n-_,m,-__12-u_z a,,w

-vna

2-

=

s.

FINDINGS

• Potential Violations * *

• Potential Deviations * *

• Unresolved Issues *

• I

[

l POTENTIAL VIOLATION ITEM 86-1741 CONTROL ROOM HALON SUPPRESSION SYSTEM TEST STATEMENT OF FINDING The Control Room Halon System Test failed to meet the 7% Halon design concentration for 10 minutes, contrary to a commitment made by CYAPCO.

Under normal operating conditions, the HVAC fans will be running. However, the system operational test was run for 13 minutes without the fans running.

Later in this test the fans were started and Halon concentration dipped to less than 5% in less than 10 minutes (i.e., before the 23 minute mark).

NFPA standards specify a 5% agent concentration for a control room environment.

The Technical Specification is based on a 6% concentration limit.

CORRECTIVE ACTION TAKEN CYAPCO committed to modify the Control Room HVAC system such that the ventilation system will automatically shut down in the event of a fire and Halon system activation. This modification will be completed by January 1,1987.

REASON FOR FINDING The finding is the result of a variance between the original design commitment and the practical issues of Halon system testing. The distinctions between the two are outlined below.

-4

• h! Commitment. On August 16, 1985, CYAPCO notified the NRC of its intent to modify the proposed Control Room Halon System from an "in the board" system to a " total flood" system. This change in design was due to seismic concerns and congested space problems within the Main Control Board.

This letter outlined CYAPCO's proposed commitment for the design concentration and acceptance criteria to be used for the design and testing of the proposed system. Specifically, the attachment states:

"The design Halon concentration for the Control Room will be 7%

for a duration of at least ten (10) minutes"(emphasis added).

and "The test acceptance criteria will be in accordance with NFPA 12A".

It is emphasized that the design concentration is that value upon which the discharge calculations are based. Pipe run and nozzle sizes are derived from the design value. In the original case, 7% Halon was selected as to assure the appropriate amount of Halon (5% acceptance criterion per NFPA 12A) would be maintained in the area, even if minor leakage occurred. However, due to personnel safety concerns, the 7% Halon concentration design number was reduced to 6% in order to allow longer stay times for occupants of the control room.

The acceptance criterion stated in the NFPA Standard establishes a minimum of 3% Halon to extinguish the type of hazards associated in the Control Room.

This acceptance criterion is widely used in the fire protection industry to

.f.*

establish design concentrations. This 5% concentration was the original intent of CYAPCO's stated commitment.

Testing Based on Worst Case Test Scenario. In CYAPCO's review of those circumstances, we felt that we developed a " worst-case" scenario to conservatively test the system. Such a scenario for the Control Room Halon System was determined to be one in which no ventilation was running. This determination was based on the following considerations:

1.

Due to seismic concerns and a congested space problem, system piping / nozzles needed to be located away from the Main Control Board.

This pipe / nozzle arrangement raised concerns about how effective Halon distribution would be and whether Halon would enter the Main Control Board which represents the major fire concern.

2.

Air supplied to the Control Room is returned through the Main Control Board.

If ventilation continues to run, air movement (from the ventilation system) would appear to assist the Halon in getting into the j

main board where the fire hazard exists.

i l

l 3.

Ventilation duct work outside the protected area had been evaluated and was undergoing modifications for reinsulation. It was our judg-ment that the insulation was sufficient to stop leaks in the ventilation system. This judgment was supported during the test by sampling the I

air in the Computer Room where a section of this new insulated duct work ran. No Halen leakage was detected. However, due to work-force scheduling problems, not all ventilation ducts were sealed before the discharge test (the system was sealed before plant start-up). That i

l portion cI the duct work that was undergoing repair was monitored,

.g and leakage was observed.

CYAPCO concluded that the temporarily sealed duct work provides an adequate barrier to prevent loss of Halon concentration. Ventilation which continued to operate and assist in air /Halon mixture was viewed as a better situation from a fire protection perspective, which appeared to be the overriding consideration.

With regard to the 6% Technical Specification, NFPA Code Section A-1-6.1 C iriazards to Personnel") specifies stay times for persons in certain concentra-tions of Halon without any special breathing apparatus. At 7% Halon (design concentration), the code establishes a stay time of fifteen (15) minutes. Since the area is required to be manned continuously, CYAPCO reduced the Technical Specification's design concentration to 6%, thus allowing additional stay time for operators without any type of breathing apparatus. It should be noted that the emergency procedure for control room fire requires operators to don self-contained breathing apparatus, in order to ensure that personnel safety concerns are adequately safeguarded.

ACTION TO PREVENT RECURRENCE CYAPCO will implement procedural revisions during September,1986, directing plant operators to trip the HVAC fans following Halon suppression system

~

activation.

Thhs finding highlights the importance of reconciling specifications and their bases in documenting commitments, Technical Specifications and system opera-

~

bility tests.

CYAPCO views its responsibility to be one of ensuring that such

~

information is clearly documented and keeping the Staff completely informed.

GENERIC IMPLICATIONS The unique aspects of Halon suppression systems limit the generic implications of this particular finding. However, CYAPCO views the documentation issue to have broader implications and is strengthening procedures and documentation regarding Appendix R compliance. CYAPCO also intends to work with the NRC Staff to relax the Halon concentration specified in the transmittal letter for License Amendment No. 81 which was issued on August 18, 1986 to a more appropriate level.

l l

l l

l 3

UNRESOLVED ITEM 86-17-02 HALON SYSTEM IN EXISTING SWITCHGEAR ROOM STATEMENT OF FINDING CYAPCO declared the Halon system in the existing switchgear room to be inoperable because an internal review of the test data identified differences between the acceptance test criteria and NFPA code requirements for Halon concentration. This system provides fire protection for redundant safe shut-down systems until the new switchgear room is completed in 1989.

CORRECTIVE ACTION TAKEN The Halon system may be restored to operable status by:

1.

installing new supervised electric circuits and release mechanisms; and I

2.

modifying the Halon system to achieve a 5% concentration for at l

least a ten (10) minute duration.

l l

These modifications to achieve NFPA code compliance will be completed by January 1, 1987. A discharge test will be performed after January 1,1987 f

during the first plant outage of greater than one week duration.

i

.., e REASON FOR FINDING

+,

The original installation of this Halon system was intended to provide protec-tion for the switchgear and for an exposure fire that could develop. The FPSER (Sept.,1978) for Haddam Neck, Section 5.2.6, states in part:

"The safe shutdown switchgear, motor control center, and DC distribution panel will be protected by an automatic local applica-tion Halon suppression system."

The cable system above the switchgear (at ceiling elevation) was pro-tected with Flamastic and at that time was not of concern from a fire protection perspective. The design of the Halon suppression system called for a 5% concentration. The system acceptance test revealed less than the required 5% at the ceiling levels (cable tray area). The installing vendor proposed certain correction actions to improve concentration /Halon distribution. These corrective actions were imple-mented and the system was not retested. However, the installer, who was j

also the designer, issued a letter to CYAPCO indicating that concentrations would be appropriate with the new modifications.

ACTION TO PREVENT RECURRENCE CYAPCO will perform acceptance testing of all new or modified fire suppression systems in accordance with applicable NFPA code l

requirements.

l rvAPCO is also in the process of implementing a long-term Appendix R compliance program.

This program will include monitoring the impacts of u.vleing NFPA standards and design bases. In the interim, fire protection system upgrades are being implemented using applicable standards.

GENERIC IMPLICATIONS This finding may be applicable to the Millstone units. For this reason, an evaluation of all total flooding gas suppression systems as well as all fire ye.tection systems is in progress to verify code compliance.

POTENTIAL VIOLATION ITEM 86-17-03 INADEQUATE COOLDOWN ANALYSIS FOR STEAM RELIEF STATEMENT OF FINDING The Appendix R audit results determined that the steam path capacities used to calculate the condensate inventory needed to reach RHR entry conditions did not account for the "as-built" configuration of the plant. Specifically, the steam relief capacity for the vent lines did not take into account the frictional losses from the steam generator to the release point. Consequently, the NRC concluded that CYAPCO did not have an analysis that demonstrated the ability to cooldown within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> in the event of a fire.

CORRECTIVE ACTION TAKEN Upon review of the assumptions on relief capacity, CYAPCO determined that the effect of the line losses was significant. A recalculation of the available steam relief capacity was performed that included a physical walkdown of the piping. The cooldown calculation was revised based upon the recalculated steam relief capacity. The revised condensate calculations determined that approximately 180,000 gallons of condensate is required to cool the RCS to RHR design conditions (i.e., approximately 30,000 additional gallons). However, the increase in required condensate is within the design capacity of the DWST and PWST. Hence, no changes to safe shutdown procedures were considered necessary.

' addition, the condensate stored in the Recycled Primary Water Storage Tank (design capacity 100,000 gallons) is available to be used to cool the RCS.

Abnormal Operating Procedures (AOP) 3.2-8 also details actions to provide makeup to the DWST using the well pumps or fire pumps if condensate inventory is being depleted.

REASON FOR FINDING CYAPCO does not consider this finding to constitute a violation of Appendix R craeria. The ability to cool the RCS to RHR entry conditions was an issue previously addressed in the Systematic Evaluation Program (SEP) Topic VII-3

" Systems Required for Safe Shutdown."

For the SEP, an evaluation was performed in 1979 by Franklin Research Center, an NRC consultant, that showed that the Haddam Neck Plant had sufficient vent capacity and water inventory to perform a cooldown to cold shutdown conditions from outside the control room. CYAPCO had provided information on steam venting pathways and capacities to the NRC and subsequently verified the Franklin Research Center evaluation as factually correct.

Until recently, this calculation contributed to the basis for the cooldown procedure outside the control room.

Since 1979, related evaluations performed by CYAPCO, in particular the steam generator tube rupture reanalysis, appeared to indicate that the evaluation by Franklin Research Institute may have been non-conservative.

The major concern was that the information learned from the St. Lucie event in 1980 had not been factored into the evaluation. The St. Lucie event emphasized the importance of the temperature distribution in the primary side in evaluating cooldown rates.

In view of this background, CYAPCO decided on its own initiative to repeat the calculation of the Franklin Research Institute and account for the effects of the temperature distribution of the primary system during a natural circulation cooldown. There was no reason to believe that the steam dump capacities previously verified by CYAPCO were incorrect. In addition, since the original evaluation was performed as part of the SEP, the vent capacities had received docketed NRC review and approval. CYAPCO's analysis merely attempted to repeat the evaluation using assumptions which were previously found to be acceptable to the Staff.

ACTION TO PREVENT RECURRENCE CYAPCO considers this circumstance to be a unique situation in view of the determination of the root cause.

In performing design basis calculations, CYAPCO rarely relies upon external evaluations as a source for assumptions on system performance. In terms of the design basis analysis used for the SEP evaluations, all accident calculations have been subsequently re-evaluated, including both small-break LOCA and non-LOCA accident analysis, and have been submitted for NRC review.

The assumptions have been verified by CYAPCO. Since the results of the analysis met the acceptance criteria for plant response for the Haddam Neck Plant, CYAPCO concludes that appropri-ate actions have already been completed and no other action is warranted.

GENERIC IMPLICATIONS Cooldown analyses are planned or in progress for Millstone 1 and 2.

Vent capacities will be re-evaluated and confirmed to meet the as-built configura-tions as part of the Appendix R compliance effort.

.P UNRESOLVED ITEM 86-17-04 CONTROL ROOM HABITABILITY STATEMENT OF FINDING Unresolved item 86-17-04 concerns the habitability of the Control Room 1:!b.ing a loss of ventilation (i.e., loss of CRHVAC Unit AC-3-la). Previous w.iervative calculations determined a steady state temperature of 1400F in the Control Room should HVAC become unavailable. During the audit, the Staff questioned when this temperature would be reached and whether it provided an acceptable basis for habitability and operability purposes.

CYAPCO committed to performing a more realistic calculation and to propose procedural steps, as required, to implement emergency ventilation in the Control Room.

CORRECTIVE ACTION TAKEN CYAPCO'has completed a re-analysis of Control Room temperatures following a loss of HVAC. This analysis is a more realistic mor'el of heat transfer conditions that would exist in the Control Room. This reanalysis indicates that a maximum ambient temperature of 1220F would be reached in the Control Room following a loss of CRHVAC.

Additionally, the analysis reflects the effect of utilizing forced ventilation (i.e., a smoke ejector fan) and one (1) Process Computer Air Conditioner (AC-19-la or AC-19-Ib) for Control Room cooling. Ambient temperature of 890F would be reached utilizing this cooling scheme.

As a result, Haddam Neck has implemented Abnormal Operating Procedure 3.2-M to provide actions to be taken when CRHVAC is lost. This procedure in2tructs Control Room operators to utilize temporary fans after initial loss of HVAC (calculated ambient temperature is 980F using only temporary fans) and to utilize two (2) Process Computer A/C units, if necessary, for long term cooling. The analysis also demonstrates that two (2) Process Computer A/C Units for Control Room cooling will allow the Control Room to be maintained at 850F.

REASON FOR FINDING Past operating experience indicates that Control Room temperature has never been a problem during CRHVAC outages.

Indeed, during a recent 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> CRHVAC system maintenance outage the Control Room reached an estimated maximum temperature of 85 degrees. Temperature was controlled by using portable fans and opening doors. Based on this operating experience and the conservatism in previous calculations it was judged that an actual loss of l

Control Room ventilation would not hinder the operators' ability to shutdown l

the plant. Further, it was assumed that if Control Room temperatures rose to abnormally high levels following a fire coincident with a loss of CRHVAC, the same measures would be implemented (i.e. portable fans and opening doors) as is presently done. However, it was the Staff's view that a special procedure was required to provide the operators specific instructions concerning imple-mentation of interim ventilation.

l

' ~~!^' TO PREVENT RECURRENCE This analysis involves a unique situation and is not expected to recur in the future.

GENERIC IMPLICATIONS NU is in the process of performing evaluations on vital ventilation systems at isotti Millstone Unit I and Unit 2.

These evaluations involve determining the availability of these vital systems during a plant fire, calculating maximum ambient temperatures if the systems are unavailable, and proposing and implementing procedural changes that would prescribe the appropriate steps to implement interim ventilation if the calculated ambient temperatures are unacceptable.

l l

l l

l l

i y

UNRESOLVED ITEM 86-17-05 COMPONENT COOLING WATER SYSTEM NOT INCLUDED IN THE FPER STATEMENT OF FINDING The audit team observed CYAPCO using the Comporent Cooling Water System (CCW) to ensure adequate reactor coolant pump (RCP) seal cooling. CCW is not in the Fire Protection Evaluation Report (FPER) and an associated circuit analysis had not been performed.

CORRECTIVE ACTION TAKEN in response to CCW concerns which emerged during the inspection, CYAPCO analyzed the system's availability in the event of a fire. The results of this analysis were provided to the Staff on August 19, 1986. The analysis supports 1

the existing safe shutdown analysis. CYAPCO's analysis identified four areas in which a fire could affect CCW functionality. These areas and the ultimate resolution within the new switchgear room design are described below:

i l

l Area Resolution PAB (Fire Area A-1 A)

Protect the charging metering pump.

Switchgear Room (Fire Area S-2)

New switchgear room modifications Cable Spreading Room (Fire Area S-3A) will protect at least one train.

Locker Room (Fire Area S-3B)

Control Room (Fire Areas S-1 A, S-1B, S-lC)

Disconnect EG-2B and manually start a CCW pump from the switchgear room.

Emergency Diesel Room (Fire Area D-2)

CCW unavailable. Use charging metering pump.

RE.". SON FOR FINDING The original Appendix R philosophy at the Haddam Neck Plant for providing RCS makeup and reactivity control credited cyclic on-off operation of the centrifugal charging pumps. Injection of charging flow through the normal charging path maintained pressurizer level in this mode. Operation of the charging pumps from the switchgear room was addressed in the original control room exemption request. RCP seal injection was not considered an Appendix R concern at this time. Resolution of RCP seal integrity issues is the subject of the NRC Generic Issue-23 Task Action Plan.

I Consequently, until research into the issue was completed and new require-ments, if any, were defined, safety concerns involving RCP seal integrity were j

not precisely defined.

l Recognizing the growing concerns in recent years over RCP seal integrity, l

CYAPCO reconsidered its original seal integrity assumptions as part of the 1984 Appendix R reanalysis. This reanalysis focused, in conjunction with its

g integration with other issues, led to the specification for a new switchgear room. It was determined that there were no makeup pumps protected under Appendix R that had sufficient capacity to maintain RCS inventory in the unlikely event of a seal failure. CYAPCO identified the metering pump as the best method for providing seal injection because its low capacity and variable speed would allow for continuous injection without the potential for overfilling the pressurizer. As an additional benefit, CCW pump operation would not be required and the CCW pump would not have to be protected. The final analysis for Appendix R compliance was based on metering pump operation, to the greatest extent possible, for RCS makeup via seal injection, and modifications to protect the metering pump and power source with the new switchgear room.

In April,1986, the Haddam Neck Appendix R audit was scheduled, and an in-house effort was initiated to prepare for it. CYAPCO's emphasis was directed towards demonstrating Appendix R compliance af ter the final 1989 switchgear and associated modifications were completed.

During the final preparations for the audit, plant fire shutdown procedures were reviewed and validated against the revised fire shutdown analysis. In the course of this review, it was discovered that although the metering pump would not be available for seal injection for a control room or an "A" diesel generator fire, both the CCW system and centrifugal charging pump would be.

CYAPCO decided that, as an interim measure designed to enhance the safe shutdown capability, the charging pump would be used for seal injection, maintaining pressurizer level and reactivity control. The CCW system availability was analyzed as an interim support system and factored into the plant shutdown

,...,wu'uces; however, as an interim measure, CCW availability was not docu-mented in the Haddam Neck Appendix R FPER issued prior to the NRC audit.

While CCW was considered a support system for an interim measure, additional

~ crational flexibility is gained by allowing one train of CCW and the charging metering pump to be powered from the new switchgear room. This will allow use of either the charging metering pump or the CCW/ charging pump as sources of sealinjection and reactor inventory control for various fire scenarios.

ACTION TO PREVENT RECURRENCE CYAPCO wishes to avoid circumstances which may be the cause of confusion.

He reever, until the final modifications are completed, the potential for the FPER to conflict with plant configurations and operating practices cannot be completely removed.

NU is in the process of implementing procedural configurational controls to assure that the Appendix R FPER is maintained up to date as future modifications are implemented.

GENERIC IMPLICATIONS i

The Millstone Unit 1 Appendix R inspection is scheduled prior to completion of all modifications. Consequently, there exists the potential that the FPER will not reflect the final plant configuration and operating practices in every respect. In order to reduce this potential to the extent practical, NU has issued I

draft FPERs for Millstone Unit Nos. I and 2 as part of our long term Appendix R compliance effort.

l POTENTIAL VIOLATION ITEM 86-17-06 INADEQUATE BREAKER COORDINATION SETTING PROCEDURES STATEMENT OF FINDING During the audit, it was discovered that the settings for a circuit breaker solid-state circuit protective device did not match those prescribed by the current engineering documents.

This led to the identification of a procedural deficiency in that the circuit breakers and their associated protective devices were calibrated per Preventive Maintenance Procedure (PMP-9.5-17) based upon the "as found" settings on the circuit protective devices. Therefore, if a device was set incorrectly, it would be recalibrated at the same incorrect value(s) and returned to service with an incorrect setpoint(s).

CORRECTIVE ACTION TAKEN CYAPCO reviewed the specific deficiencies associated with this finding. The action associated with each deficiency is outlined below.

i

,l_l Setpoint Change Request No. 7 CYAPCO has implemented Setpoint Change Request No. 7, to correct setpoints on circuit breaker 3C.

l l

l

1 n2-25 Breaker Test Procedure, PMP 9.5-17 CYAPCO reviewed Preventive Maintenance Procedure (PMP 9.5-17) and concluded that a revision will be required. This revision will require the Master Setpoint list values be recorded in the PMP 9.5-17 before the circuit breaker is tested to ensure correct breaker settings. This revision will be completed by July 15,1987 consistent with the schedule for similar setpoint verifications scheduled for the 1987 refueling outage.

It is noted that there were no adverse safety implications associated with these specific deficiencies.

REASON FOR FINDING Setpoint Change Request No. 7 CYAPCO's review indicates that a management oversight contributed to failing to implement the setpoint change. On May 23,1985, the setpoint change request (SCR) #7 was approved. On June 4,1935, Controlled Routing #85-715 was assigned to Maintenance to " Complete Section Ill" of SCR #7. When the assistant maintenance supervisor reviewed the routing, he thought he was asked to complete the documentation of other SCRs that affected the 480V switch-gear. These SCRs included 4, 5, and 6, and were implemented during the 1984 refueling outage under the work order for preventive maintenance of these four buses. SCR 7's documentation was completed and the Drawing Change Request verified. The supervisor then closed the routing and the SCR.

l

DB-25 Test Procedure CYAPCO did not believe there was a significant potential for breaker setpoints to vary from engineering documents. CYAPCO now recognizes that verifying setpoints, before testing, against the Master Setpoint List is a good practice, and this reverification will be implemented by procedural revision.

ACTION TO PREVENT RECURRENCE Procedural revision to PMP 9.5-17 will ensure that setpoints are properly verified.

Although motor protection and breaker coordination was not significantly affected by the SCR 7 oversight, CYAPCO believes that such oversights are unacceptable. The supervisor has been counseled, and personnel have been directed to carefully review all controlled correspondence prior to taking action.

GENERIC IMPLICATIONS CYAPCO believes the generic implications of this finding lies in the importance of management's attention to detail. NU will investigate the applicability of I

this issue to the Millstone units.

t

{

i I

I l

POTENTIAL DEVIATION ITEM 86-17-07 RCS LOOP BYPASS VALVES MCC BREAKER NOT LOCKED STATEMENT OF FINDING By letter dated September 16, 1985, CYAPCO committed to have the breakers for the following valves locked open at the motor control center (MCC) by

.bsust 14,1985: RC-MOV-510, RC-MOV-515, RC-MOV-528, and RC-MOV-577.

CORRECTIVE ACTION TAKEN As is noted in the inspection report, when this commitment deviation was found, immediate corrective action was taken. A Temporary Procedure Change (TPC) was initiated on June 19, 1986 to lock these breakers, and the valves were added to SUP3.1-126 Locked Valve Checklist.

As discussed in CYAPCO's August 29, 1986 letter responding to this potential deviation, CYAPCO has modified the September 16, 1985 commitment to have those breakers locked open. Unless a different agreement is reached during the l

September 3,1986 management meeting, we plan to remove the locks for the subject breakers during September,1986.

REASON FOR FINDING It is noted that the breakers were found in the racked-out position when I

L inspected by the audit team. Breaker positions are checked weekly by the Operations Department in accordance with Normal Operating Procedure 2.16-1.

Further, a review indicated that the four breakers were de-energized (racked out) as of August 14, 1985. This action was sufficient to address the actions needed to support Appendix R safe shutdown procedures.

At the time, a number of procedure changes were also approved to ensure that the bypass valves would remain de-energized. The valves were de-energized in accordance with the safety evaluations supporting the procedure changes. When completed, the valve motor operators were in the desired state - de-energized. Subse-quently, the commitment was made to " lock open" the breakers. The reason for the finding was the failure to adequately review licensing correspondence against the procedural controls previously implemented. CYAPCO also concurs with the Staff that failure to lock the racked-out breaker is not a violation of Appendix R but rather a deviation from a commitment.

ACTION TO PREVENT RECURRENCE This commitment deviation underscores the importance of verifying that 1

actions taken are consistent with actions reported to the NRC. Inadequate draf t correspondence review likely contributed to this oversight. CYAPCO will devote adequate resources to ensure appropriate review of correspondence to ensure comments and questions are resolved prior to finalizing correspondence to the NRC.

GENERIC IMPLICATIONS The generic implications of this finding are in the managerial aspects of plant

,.;..:3.ns. Sufficient resources will be devoted to the careful review of regulatory correspondence.

I I

i l

t I

t l

l

{

I

POTENTIAL VIOLATION ITEM 86-17-48 INADEQUATE EMERGENCY LIGHTING STATEMENT OF FINDING During the safe shutdown procedure walk-through, CYAPCO operators climbed into the charging pump cubicles to operate valves BA-MOV-32, BA-MO3-373, and CH-MOV-257. No emergency lights were in either cubicle. In addition, the operators would have to use portable lights to make instrument connections and take readings in the cable vault room.

CORRECTIVE ACTION TAKEN Upon receiving this finding CYAPCO immediately reviewed the circumstances.

This review confirmed that manual operation of the specified valves and equipment will only be necessary in the interim until the new switchgear room is completed.

At that time, a simpler alternate shutdown process will be l

l available using redundant instrumentation and equipment control from the l

l switchgear room. Until then, operators would need to enter the charging pump cubicles and the cable vault to perform shutdown actions. CYAPCO concluded l

that since the measures are interim and the final alternate shutdown procedures i

have yet to be developed, no violation of Section III.3 exists.

CYAPCO recognizes that these interim procedures merit the provision of enhancements to the interim lighting situation beyond that lighting credited

- t

,, y

' -5; the audit walkdown (flashlights). As an enhancement to lighting in the charging pump cubicles, a portable lantern has been installed by the entrance of eacn fire area, A-1B and A-IC. Additional emergency lighting will also be attached to the instrument cabinet in the cable vault.

REASON FOR FINDING None ACTION TO PREVENT RECURRENCE None GENERIC IMPLICATIONS None a

FOLLOW-UP ACTIONS A number of follow-up actions were also identified in the audit. The actions and their status are summarized below.

Action Items Status 1.

Communications - Provide Closed. Additional information was additional information provided by letter dated August 19, concerning the radio and 1986 which concluded that radio paging systems including an channel one was available for a fire in evaluation of communicating any fire area and would not be in " dead spots" by August 19, damaged in a fire. A test of this 1986.

channel resulted in good communications in all areas with no

" dead spots."

2.

Cable Vault Instrumentation Open. An evaluation was completed

- Provide an evaluation of which concluded that an improved the temporary arrangement temporary arrangement less prone to for monitoring certain plant operator error was practical. A parameters in the cable description of the new instrument vault, including a review to arrangement was provided by letter determine if a method less dated August 19,1986. Purchase prone to operator error is orders have been issued to implement practical. The results of the the modifications by January 1,1987.

evaluation should be submit-ted by August 19,1986. All modifications will be com-pleted by January 1,1987.

l 3.

High Impedance Fault Closed. Formal approval of the l

Procedure - Formally procedure was secured on August IS, l

approve the operating 1936.

l procedure governing high impedance faults.

4.

Charging Pump Cable

(

Protection - Install at least Open. Additional sprinkler heads will I

two additional sprinkler be installed by January 1,1987.

l heads in the cable tray outside the charging pump cubicles.

• a

, w

'a e

5.

Fire Watch Training -Provide Closed. " Hands-on" training for fire

" hands-on" training to fire watches was completed by July 19, watches.

1986.

6.

Miscellaneous Fire Open. The modifications will be Protection Modifications -

completed by January 1,1987.

Make the minor modifications discussed in CYAPCO's July 3,1986 letter.

l l

i O

e g, l.

g

SUMMARY

After evaluating the implications of the audit findings and their root causes, we have developed the following list of corrective measures. They are designed to prevent future problems as well as deal with the findings identified during the audit.

Hardware Modifications 1.

All hardware modifications identified during the audit will be completed by January 1,1987.

2.

The new switchgear room will be completed during the 1988-89 refueling outage. This major modification will significantly decrease reliance on existing exemptions in order to meet Appendix R and eliminate the need for interim measures.

3.

Additional modifications will be implemented in order to upgrade existing fire protection systems to current NFPA codes.

Documentation 4.

The need for adequate and complete documentation consistent with Appendix Ra and other related fire protection commitments is being emphasized at Millstone Unit Nos. I and 2.

3.

On a long-term basis, NU will proceed with license amendments to update the current fire protection license conditions for all four plants consistent with Generic Letter 86-10.

l Management Controls and Training 6.

Various options to improve the visibility of fire protection and housekeeping are being considered and will be discussed in our response to any further correspondence on the inspection findings. It is recognized that continued emphasis is required to ensure proper control of flammable materials and operability of fire protection equipment. The importance of fire protection and continued Appendix R compliance, and how this level of importance will be communicated throughout NU, will be discussed in our response.

7.

" Hands-on" training will be provided for all fire watches at Millstone.

8.

Tri-annual audits will be implemented, starting in November 1986, in order to provide an independent review of the Haddam Neck and Millstone Unit Nos.1,2 and 3 fire protection and loss prevention programs.

l l

We are confident that we will continue to improve as we move forward.

oE :N o 3.

e 9

CHRONOLOGY OF PREVIOUS CORRESPONDENCE Subject 1.

June 16 - 20,1986 Appendix R Audit 2.

July 3,1986 Letter from 3. F. Opeka to C. I. Grimes discussing follow-up action items resulting from the Appendix R audit.

3.

August 1,1986 Letter from S.

D.

Ebneter to

3. F. Opeka transmitting Inspection Report No. 50-213/86-17.

4 August 18,1986 Letter from F.M. Akstulewicz, Jr. to 3.F. Opeka issuing Amendment No. Si to the Haddam Neck Operating License regarding the control room Halon system.

0 5.

August 19,1986 Letter from 3.F. Opeka to C.I. Grimes responding to audit follow-up action Items and inspection report unresolved items.

6.

August 29,1986 Letter from 3.

F.

Opeka to S.D.

Ebneter responding to Inspection Report No. 50-213/86-17.

i l

l i

I t