ML20113F476

Special Assessment:Confirmatory Action Ltr

ML20113F476
Person / Time
Site:	Clinton
Issue date:	09/16/1996
From:	ILLINOIS POWER CO.
To:
Shared Package
ML20113F470	List: ML20113F476 U-602637, Forwards Special Assessment:Confirmatory Action Ltr , Discussing Events Leading to 960906 Shutdown of Plant Due to Reactor Recirculation Pump Seal Failure
References
CAL, NUDOCS 9609240244
Download: ML20113F476 (52)

Text

• Attcchment 1 6 . .

a k

Special Assessment l Confirmatory Action Letter Date('49, ./mber 11.1996 I

t 1

4 i

~

Prepared by: ~W / 9//c /94' Q _s Da(e Approved b + 16/ 9-/4-96 4

Date i-i 4

i i

YY l 9609240244 960916 i PDR ADOCK 05000461 g PDR

1 i

1 6 , .

Executive Summary .

l On September 5,1996, Clinton Power Station personnel, in pursuit of an increasing trend l of reactor coolant leakage, formulated plans and performed procedures to mitigate the consequences of excessive reactor coolant leakage by shifting the operation of the Reactor Recirculation System from two-loop to single-loop operation. During the course of these actions, the reactor coolant leakage rate increased to rates which are not allowed by the Plant's Technical Specifications (T.S.-3.4.5.b s5 gpm) resulting in the declaration of a Notification of Unusual Event. Subsequently, the Reactor Recirculation pump seal failed causing leakage rates to again l increase beyond limits allowed by the Technical Specifications (T.S.-3.4.5.d s2 gpm increase within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />).

On September 9,1996, the Nuclear Assessment Department began an investigation and

evalestien of the events surrounding the activities of the above described event and the follow-on

of the

cactor shutdown. Our assessment identified the following conditions:

Management personnel were not conservative in the , operation of the plant.

Numerous opportunities existed for management to evaluate uncertainties and the i need for continued reactor operation, but these were not recognized or acted upon.

Management did not properly establish, enforce or set the proper example for procedure compliance.

Oversight of the overall picture of plant conditions and actions surrounding the event was ineffective.

There was inadequate planning and evaluation of the potential consequences prior to performing an infrequently performed operation. -

= l Management tolerated long standing equipment problems that contributed to )

uncertainty of some plant conditions.  !

i

-IP personnel were not timely in identifyin(a potential procedure non-compliance and management did not sufficiently pursue other indications that would lead to i management recognition of the significance of non-conservative reactor operation.

It should be noted that IP was prompted by the NRC to pursue the issues of procedural compliance.

Many of these concerns were previously identified in letter RFP96-026 on June 10,1996, Assessment of On-line Maintenance Details of the above conclusions are contained in the  !

detailed chronology (Attachment A) and the evaluation that follows.

. - - - . - - - - .- - =-

6 , .

I. Evaluation - l d

l This' analysis of the evolution and events surrounding the September 5,1996 Reactor j Recirculation pump seal failure at Illinois Power's Clinton Power Station evaluates the following: j

! Decisions made and actions taken and their basis; contingencies evaluated related to the decisions and their consequences, management involvement and oversight.

{

1730 to 1800

Action: Pre-Evolution Briefing l Basis: The pre-evolution briefing was conducted to ensure all personnel involved in the evolution j- understood their roles and responsibilities, to review the applicable procedures, and to discuss i potential problems that might be encountered. The detailed Chronology, Attachment A, provides j attendees and topics covered.

! Management Involvement: The Assistant Director- Operations (ADO )(also acting as Director-Operations at this time) was present at the briefmg and had provided the Shift Supervisor, (SS) 4 with his written entions (Attachment I to the Chronology) prior to the brief.

j Evaluation: The brief was considered by the personnel in attendance 16 be one of the best they 4

had attended. Although contingencies were discussed, it is not clear that specific actions to be

, taken for each contingency were covered. It appears the focus was on completing the isolation i and determining whether or not the Drywell Floor Drain (DWRF) leakage rate stabilized below j the administrative limit of4.5 gpm. This limit had been established by Operations :ranagement

with the Plant Manager's concurrence in the expectations paper provided to the SS. Although the i potential for leakage to increase was discussed, it appears that it was assumed the crew would i proceed with the isolation of the loop and then evaluate the steady state leakage remaining. It j does no,t appear that the discussion addressed that an orderly shutdown or SCRAM would be i required if an upper limit ofleakage was reached. Although there are indications that the l procedure, CPS 3302.01 REACTOR RECIRCULATION (RR), for operating (including isolating i one loop) the Reactor Recirculation System was reviewed prior to and during the pre-evolution

~

briefing, it is not clear that the review included an evaluation ofwhat effect the steps of the

! procedure 1;ould have on the rate ofRR pump "B" seal leakage. If this had occurred, the crew l might have known in advance that they would most likely have to enter Technical Specification Action Statement 3.4.5a and a Notice of Unusual Event (NOUE) if they proceeded with the i evolution. This might have caused additional questions to be asked and analysis to be performed.

The personnel involved in the event believed that the singigloop isolation could be performed, but there was no consistent conclusion on whether leakage could be expected to decrease. The i mindset seemed to be that even if the "B" RR pump seal degraded fiuther during the event (which l was not really expected to occur by the participants), after isolation, the DWRF leakage would decrease, or at worse, remain the same. In general, the focus of the crew and management

appears to have been on the performance of the single loop isolation. It is not clear that the  ;

{ evolution and worse case possibilities from a reactor safety / primary coolant boundary perspective were adequately analyzed, including limits that would require the initiation of pre-determined i conservative actions.

j 1 i

l 6

. l l

I Additionally, the expectations provided to the SS by the ADO were in effect an Action Plan that )

had not received an independent technical review and contained provisions not specifically covered by existing procedure guidance.

I Furthermore, the evolution was begun knowing that equipment problems existed that could potentially impact the evolution. One of the isolation valves, the "B" RR Loop discharge valve, (IB33-F067B) had previously provided dual indication when Operations tried to close the valve leaving its status indeterminate in terms ofwhether it was fully closed or not. Furthermore, the V-notch system utilized to determine DWRF leakage was inoperable.

Finally, during the pre-evolution analysis and the pre-evolution briefing, the personnel evaluating and planning to implement the overall evolution did not sufficiently consider conditions that could prevent its success. It was suspected that some amount ofleak-by existed in the isolation valves for the "B" RR loop. If the leak-by was equal to the flow exiting the seal, this would keep the

" isolated" loop pressurized. Ifleak-by exceeded seal leakage, the limitation on leakage would continue to be the holding capability of the "B" Reactor Recirculation pump seals. Management is at fault for requesting an evolution that had a limited possibility for success. All organizations involved in the planning, preparations, and performance were at fault for not challenging the plan more aggressively.

1805-2009 Action: Reduced Reactor Power to prepare for implementing RR B Loop isolation.

Basis: Procedure CPS 3302.01 Section 8.2.1, "RR Loop Shutdawn During Plant Operations" requires that reactor power be at a level speci6ed by Management and Nuclear Engineering and evaluated against the Power to Flow Map prior to performing a pre-planned loop isolation.

Management Involvement: The ADO was present in the main control room area.

Evaluation: Reduction in power was accomplished in accordance with procedures in an orderly manner.  ;

2009 Action: Shutdown of"B" RR pump in accordance with CPS 3302.01. This included shutdown of RR Pump B (IB33-C001B) , closure of the pump B discharge valve (IB33-F067B), and fully opening the Bow Control Valve (IB33-F060B) per steps 8.2.1.3 , 8.2.1.4, and 8.2.1.5.

Basis: This is the sequence of steps for RR pump shutdoilm prescribed by the procedure.

Management Involvement: The ADO was still available in the area.

Evaluation: Based on previous experience that closure of the discharge valve resulted in dual indication and the inability to tell whether the valve had fully closed, electricians and the VOTES engineer had instrumented the valve in advance to provide more positive indication of closure.

The indication included motor current traces and stroke time. This was considered ged pre-planning for the evolution. When shut, all indications provided were that the valve fully closed including a green only position indication on the Main Control Room panel. At this time, the pump seal pressures, P1 and P2, began to equalize. The reason for this was not fully understood, nor was this expected to occur. The crew began to analyze what could be causing the 2

.-- _ . - _ .~ -.--- . .

, i j

l o .

! equalization. A theory was proposed by the System Engineer and the Engineering. Department l was tasked with analyzing this occurrence.

b

At this point, DWRF leakage had not increased and ti

e crew believed the evolution was l l proceeding as expected with the exception of the seal pressure equalization. I i

i l l 2009-2030 Action: Initiate RR Loop isolation by closure of the Reactor Water Cleanup (RT) recirculation li loop "B" suction valve (IG33-F106) per CPS 3302.01 step 8.2.4.1 to isolate the RT system from j the RR system.

Basis

This is the sequence of steps prescribed by procedure CPS 3302.01 for single loop l isolation.

j Management Involvement: The ADO was still present in the general area.

t Evaluation: At this point, the next step in the loop isolation procedure step 8.2.4.2, which 1 closes the discharge valve, had already been completed per Step 8.2.1.4. Step 8.2.4.3 was

.' reviewed, which is an "IF - THEN" statement and only requires action if the "IF" statement is i satis 6ed. In this case, the "IF" statement loop suction isolation is required 4ue to an emergency I condition (system / seal leak ). Since the SS did not consider himselfin an emergency condition j due to sealleakage, he skipped this step and proceeded to step 8.2.4.4. Step 8.2.4.4 should be a j caution prior to step 8.2.4.6, because it describes conditions to be met prior to the Control Rod

Drive (CRD) seal injection flow being isolated and does not direct any actions.

This is considered a procedural deficiency, but appears to have had no impact on the evolution, at this time.

l j

i j 2030 -

l Action: Closed the pump "B" seal staging shutoff valve (IB33-F075B).

!. Basis: This is following the prescribed sequence ofprocedure steps. The crew also thought that l this would force more cold Control Rod Drive water into the loop. -

l Manageoust Involvement: The ADO was still available in the area.

l Evaluation: This step is part of the routine sequence for isolation of a reactor recirculation loop.

However, the planning for the evolution and/or the pre-evolution brief may not have considered all of the effects that shutting IB33-F075B would have on a degraded seal. This step of the

. procedure was written under the assumption that the seal was in normal working order. Prior to I

shutting IB33-F075B, DWRF leakage (unidentified leakade per Technical Specifications) was l approximately 4.7 gpm. Twenty-five minutes after shutting IB33-FO75B, DWRF leakage l increased to approximately 5.5 gpm. 1B33-F075B diverts approximately I gpm of RR "B" seal

leak-off to the Drywell equipment floor sumps, which is considered identifiable leakage by Technical Specifications. While shutting this valve diverted flow to the "B" loop to expedite i loop cooldown to less than 250 degrees, it also increased the differential pressure seen by the RR i "B" pump upper seal (P2).- This additional stress on seal P2 may have led to the increased DWRF

flowrate seen. When DWRF leakage exceeded 5 gpm, a limiting condition for operation per j action A of Technical Specification 3.4.5 was entered. This requires restoration of unidentified i leakage to less than 5 gpm within four hours. Additionally, unidentified leakage greater than 5 4

, 3 J

l

o ,

gpm requires that a Notice of Unusual Event (NOUE) be declared in accordance with the CPS Emergency Plan per procexture EC-02 Step 4.1, LEAK RATE EXCEEDED LOSS OF INVENTORY. Had the procedure been sufficiently analyzed prior to the evolution, it might have been identified tim closure of the seal staging valve line might increase leakage to above the action level. Tins would have allowed further analysis of the merit of proceeding with the evolution and contingencies should Technical Specifications Action statements be reached.

2055  !

Action: The crew entered CPS 4000.01 REACTOR COOLANT LEAKAGE off-normal  !

procedure, entered the limiting condition for operation required by TS 3.4.5, and declared the NOUE.

Basis: Unidentified leakage above 5 gpm requires entry into Tech Specs 3.4.5 and declaration of an NOUE per Emergency Plan Procedure, EC-02, Step 4.1.

Management Involvement: The ADO was closely following the events at this time and began conferring more often with the Shift Supervisor.

8 Evaluation: The actions taken at this time were appropriate and in compliance with the

governing procedures. The entrance into the off-normal procedure, TS Action statement, and

, NOUE were an opportunity for shift management to assess overall plant conditions.

Additionally the Shift Supervisor asked the STA if he had time to perform the required

! notifications. The STA affirmed that he did and took the action and responsibility to make the notifications. In hindsight, both the Shift Supervisor and the STA indicated that if they had it to do over again, they would have had the responsibility assigned to an operator. This placed the STA in a role different than he normally fulfilled.

2116-2127 l Action Depressurized the Drywell using Mixing Compressor "A". Secured Mixing Compressor

"A" when Drywell pressure was reduced to 0.23 psig (Time 2127).

Basis

This provided additional margin between the drywell pressure and the auto scram setpoint-for this parameter.

}

Management Involvement

The ADO continued to closely follow the events and confer with the Shift Supervisor.

Evaluation: This was good anticipation and preplanning by the shift crew.

~

j L l

2130 Action: The crew returned to step 8.2.4.3.a of CPS 3302.01 and closed the RR pump suction I valve (IB33-F023B).

j Basis: The crew returned to this step because the Unusual Event was considered an entrance to

(emergency condition) this step. Because the seal staging shutoff valve (IB33-F0758) had been

previously closed in step 8.2.4.5 at time 2030, the second part of this step,8.2.4.3.b, was not l performed.

~

i l

4 4

i -

S Management Involvement: Prior to returning to Step 8.2.4.3.a, the Shift Supervisor discussed the condition with the ADO. The ADO concurred that the Unusual Event did constitute an emergency condition, whichjustified returning to the step.

Evaluation: Returning to step 8.2.4.3.a at this point and time is considered appropriate. Upon  ;

completion of this step, the Reactor Recirculation "B" loop would be isolated from the reactor.

Although activities in the Control Room during the period of time from 2055 to the present (time 1

2130) were described as hectic, upon completion of this step, the RR seal was isolated and the '

crew had time for a more methodical appraisal of the current situation. Closure of the suction i valve had minimal impact on the leakage rate. At this time, multiple indicators of complexity (

should have signaled management that a more conservative decision on reactor operations should l be considered. Although the entrance into the LCO and NOUE had increased the activity level in .

the Main Control Room, it is not clear that a heightened awareness in regard to reactor safety was l occurring. The focus ofiho operating crew and the management personn4 appeared to be on reducing the DWRF leakage.

Because the suction valve had been instmmented to ensure closure and all indications including the local VOTES instmmentation and the Main Control Room panel position indication indicated that the valve had closed, the crew had good reason to believe that the valve had fully closed.

2130-2159 Action: Main Control Room personnel monitored plant conditions, specifically, DWRF and RR "B" seal pressures for change. During this period of time, (2144) unidentified reactor coolant leakage exceeded a 2 gpm increase in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Unidentified reactor coolant leakage indicated approximately 6 gpm. Twenty-four hours prior to this, the reading was 3.8 gpm. Technical Specification 3.4.5 was re-entered for the 2 gpm leakage increase. RR "B" pump seal pressures during this period were steady at 980 psig on P1 and P2.

Basis: Personnel were monitoring DWRF leakage and seal pressures to determine whether closure of the "B" loop suction valve had an effect.

Managessent Involvement: Within this general time frame, the Illinois Power Company Vice -

President-Nuclear called the Shift Supervisor to obtain a status after having been notified via his pager that an emergency had been declared. He was m ' formed that the loop had been isolated and that we were in single loop operation. In the conversation, it was understood that if the leakage did not come down, a plant shutdown would be required. This was another opportunity for the Shift Supervisor and management to evaluate the overall implications of the event in regard to plant safety. Again, it appears that the focus was on whether or not the leakage could be reduced.

The ADO was still present and conferring with the SS periodically for updates and to provide suggestions ifrequested.

Evaluation: During this time, Technical Specification 3.4.5 was re-entered for the 2 gpm leakage increase. This should have been another indicator that it was time to reassess the situation and at least evaluate whether a shutdown of the reactor should be considered. Because the leak was basically steady state, and there was no perceived immediate risk, management attention continued to focus on DWRF leakage reduction. The plant had experienced DWRF leakage increases twice, entrance into Technical Specification Action Statements twice, and entrance into an off-normal procedure (CPS 4001.01, ABNORMAL REACTOR COOLAN f LEAKAGE), yet 5

a 4

j * . .

i none of these occurrences precipitated management or shift supervision stepping back and i evaluating the overall plant status and pot.utial concerns related to reactor safety. Management l was involved in the evolution from a perspective of assisting with ideas for leak isolation, instead of taking a broader view of overall plant safety.

I i 2159 i Action: Closed the Control Rod Drive (CRD) supply isolation to RR pump "B" (ICI1-F026B) j per CPS 3302.01, step 8.2.4.6. This completed the RR loop isolation per section 8.2.4 of CPS l 3302.01.

l Basis: The SS and ADO, with input and discussion from the crew, had evaluated that the isolated i loop was not going to depressurize through the seal with CRD flow valved in. Because DWRF l leakage was determined to be approximately the same as CRD make-up flow, it was thought that

. the leakage could not be reduced without isolating the CRD make-up flow. Additionally, the shift was focused on reducing DWRF leakage to comply with CPS 4001.01. One concern with i maintaining CRD injection was that overpressurization and further damage to the seal could i occur. Tne Shift Supervisor decided that after loop isolation pressure from CRD injection would

[ either cause pressure to increase on the seal or that seal pressure would stast to slowly decrease.

l If the seal started to pressurize, it would mean leakage also was coming from outside the loop j isolation boundaries. Ifit started to decrease, it would mean the seal leak rate was now greater j than the CRD injection rate. Inis was assuming that the loop isolation valves were not leaking i by. He expected increased pressure to either lift the injection relief valve, or require reopening the

suction valve so pressure could be released to the reactor. The potential for catastrophic seal j failure when CRD seal flow was secured and the resulting potential for drywell contamination l from the reactor water in the loop depressurizing through the seal was discussed with the RR I

System Engineer and the ADO. They decided that catastrophic seal failure was not likely and that j further degradation of an already damaged seal was acceptable. The evaluation that catastrophic j seal failure was not likely was based on the System Engineer's evaluation that isolating the CRD i injection would increase the stress on the seal and possibly incrementally increase damage, but

that the seal would probably not have a severe failure. They also decided that isolating the RR -

! pump "B". teal leakage was more important than the potential to contaminate the drywell. They F were also confident because of the full-closed position indication in the MCR and the data from

the instrumented "B" loop suction and discharge valves, that the "B" recirculation loop was  !

! securely isolated. An additional consideration was that should a catastrophic failure of the seal  !

ow;r, that the isolated finite volume of coolant contained in the loop would quickly depressurize, l

anci that any immediate increase in initial DWRF drain floderate would soon decrease. This l l assumed that the isolation valves did not leak-by.

j Management Involvement: The Shift Supervisor, with concurrence of the ADO, made the

decision to close the CRD supply isolation to RR pump "B" (ICll-F026B). They had assumed

the role of managing the evolution and trying to solve the problems being encountered, rather j than maintaining their proper oversight function.

i Evaluation: This is the most significant error that occurred during the evolution for two reasons.

i First, management, the SS, and the crew continued attempts to reduce DWRF leakage to meet the j requirements of CPS 4001.01 by isolating CRD. Isolating CRD did nothing to reduce reactor I

i 6 i

{

~ _

. . . - - - - - . .. - .-.. . .~_- - - - - - .-. -. -

e a ,. ,

I coolant leakage. If anything, it replaced the water leaking out of the seals with restor coolant leaking by the isolation valves. Thisjeopardized an already degraded seal even further.

The second reason that this was a significant error is that the conditions required by the procedure l prior to isolating the CRD injection flow were not satisfied. The two conditions dealt with allowing the loop to cool below 250 F as a protection to the seal and preventing contamination and airborne activity in the drywell by not allowing the reactor water pressurized in the isolated loop to depressurize through the seal. However, the Shift Supervisor and the Assistant Director of Operations believed that awaiting these conditions would preclude isolation of the leak per CPS 4001.01. Both conditions were evaluated, and the judgments involved had technical merit. A more conservative approach would have been to implement a controlled reactor shutdown while monitoring leakage and allowing temperature to decrease. However, it was evident that the temperature of the isolated loop could not be expected to decrease to below 250 F, nor could it be expected that the loop would depressurize in the near term. For these reasons, it was evident that the conditions required to isolate CRD flow would not be satisfied prior to entrance into the Technical Specification required shutdown statement of section 3.4.5. Because the crew and management oversight were focused on reducing the leak and were not stepping back to maintain proper oversight of overall plant safety, isolating the CRD flow appeared reasonable and justifiable.

As noted above, the crew and management present expected that even if the seal should fail, that the coolant would quickly depressurize and any increase in floor drain flow would soon decrease.

It is not clear that the crew or management evaluated the situation for the contingency that floor drain flow rate would not decrease. The shift appeared to be confident that they could anticipate and predict what would cccur and did not take the time to evaluate what the specific action should be, should the unexpected occur.

Additionally, the interviews with the Reactor Operators (RO's) indica;ed that one of the RO's made a suggestion to first throttle CRD supply back by 2 gpm. However, this suggestion, which was identified directly to the Shift Supervisor does not appear to have been evaluated for _

implementgion prior to securing the CRD injection. This would have been a more conservative  ;

approach, ahd the lower procedure limit of 3 gpm could have been tried. This would have  !

maintained compliance with the procedure and also allowed an evaluation of reduced CRD flow to the seal.

~

L 2217 Action: Annunciator 5003.01,-lK was received. This indicated a seal cooler outlet temperature alarm on the "B" reactor recirculation pump. The temperature identified was 146'F.

Basis: With isolation of the CRD complete, this was expected.

Management Involvement: It was at approximately 2200 when the Plant Manager arrived at

the Main Control Room. At this time, the crew and management present were still focused on

reducing leakage and were presently monitoring the aff'ects of having closed the CRD injection valve.

7

_ _ . _ . _ _ _ . - _ _ _ _ _ _ ~ _ _ _ . _ . _ _ _ _ ._... _. __ __

e ,

Evaluation: This was another opportunity to evaluate overall plant conditions and conservative {

actions that might be taken. Indications are that the crew and management continued to focus on  !

reducing the leakage.

l 2222-0039 (Friday, September 6,1996) ,

Action: h reactor recirculation pump "B" seal pressure decreased rapidly from approximately 950 psig to approximately 280 psig within a few seconds. A drywell air cooler drain flow alarm -  ;

was received. Drywell pressure started to increase and the "B" mixing compressor was started. l The containment was evacuated per CPS 4001.01. At basically the same time (2223), white data was indicated on the computer point for DWRF flow rate. Drywell pressure peaked at 0.45 psig ,

then decreased to a pressure of 0.12 psig as steam was being condensed in the drywell. The "B" mixing compressor was secured when drywell pressure reached 0.40 psig on its downward trend.

I h calculated value for DWRF leakage peaked at 38.1 gpm (now considered 23 gpm but at the time the more conservative value of the two was used) and continued decreasing to 14 gpm. At ,

2255 the shiR recognized the DWRF rate indications were inoperable and entered actions for leak l detections instrumentation per the Technical Specifications Section 3.4.7.

Basis: Starting a mixing compressor and evacuation of the Containment are, standard procedures when drywell pressure begins to increase. Additionally, the shia crew began using manual  :

calculations ofleakage because they recognized that the indication was inoperable.

Management Involvement: N Plant Manager and ADO were still available in the general area and were monitoring the overall actions of the crew, h crew and management present saw that the leakage was getting better and that seal depressurization was proceeding as expected. This provided ar. additional level of comfort that the evolution was under control. Although the seal had rapidly depressurized, drywell pressure started to increase requiring that a drywell mixing compressor be started, and the indication ofDWRF leakage became inoperable, the fact the calculated leakage indicated a decreasing rate influenced the crew and management present to l l l believe that success was close at hand.

j Evaluation: Initiation of the mixing compressor and evacuation of the containment were l appropriate steps to take. When the computer point for drywell floor drain flow rate became -

" white data", h was not immediately understood by the Operators what was occurring.

Discussion ensued, and the conclusion was drawn that the data was not dependable. The Shin Supervisor directed the Shia Technical Advisor (STA) to begin manual calculations for the DWRF flowrate.

Although sev_eral indicators were received that should have driven management to step back and evaluate overall re. actor safety, they were not recognized deacted upon. The focus continued to be on succer,sful loop isolation and reduction of the leakage, rather than overall conservative operation.

Additionally, it is not clear why none of the management or shia crew present understood that the DWRF leakage instrumentation in use was not designed to provide indication above 8.0 gpm.

This caused considerable confusion and several discussions occurred to try and resolve this i

unanticipated event. & ShiR Supervisor did rapidly recognize that regardless of the cause, when j the data went white, it was no longer dependable, and immediately directed the STA to begin

manual calculations.

i 8

_-.._ _ _-- . .-.-- . - . _ - - . . _ _ - . ~ . - - - - - _ _ _ - -

i + , .

l- It is not clear that at any time from the beginning of the evolution to isolate the "B" reactor j recirculation loop until this time in the chronology, that any member of the shiA crew or i management, present or contacted during the event, focused on anything other than isolation of j the loop and reducing the leakage. This is the most significant issue and consideration from l the overall evaluation. At no time during the period just described did any crew member or

management individual present or contacted during the event question whether we were doing the i'

right thing in regard to our principles of conservative decision making with reactor safety as our number one priority.

j During the evolution, the shiA crew and management uresent thought they were in control of the I i event at all times. Management oversight personnel did not step back and question whether 1

operation was being conducted conservatively in the strictest sense of the word. This also allowed i the personnel in command of the evolution to analyze the events, as they occurred, step by step, l l and provide technicaljustification to themselves on why the actions being taken were prudent and l

reasonable.

l l

0039 Action: The crew performed a Double Isolation of the Reactor Water Cleanup (RWCU)

System from the 'B' Reactor Recirculation Loop.

Basis: The crew believed that a contributor to the RR pump 'B' seal leak could be leak-by of the IG33-F106 Recirc Loop 'B' suction valve. In an attempt to further isolate RWCU from the 'B' RR loop,1G33-F100, Recirc Loop 'A' suction and IG33-F102 RWCU Recirc Suction throttle valves were closed. l Management Involvement: The Assistant Director - Plant Operations had provided a list of l management va* ions for the single loop isolation evolution, which included a contingency for ,

double isolating RWCU,' He presented this to and discussed it with the ShiA Supervisor.

Evaluation: The list of management ==*=+ ions was draAed by the Assistant Director - Plant Operations with concurrence of the Plant Manager. These ==*=+ ions were formulated to -

ensure the Shift Supervisor and management had the same focus concerning the loop isolation evolution. One ,-*= tion was to consider double isolating RWCU from the RR loops to determine if IG33-F106 could be leaking-by. This guideline was not technically reviewed to determine if current procedures supported the evolution to be performed This presented an uncertainty too the shift crew that was not specifically covered by station procedures.

During the crew discussions leading up to performing the evolution, questions raised regarding the adequacy of the procedure by crew members were not adequately addressed and resolved by shift supervision. PMSO 043, VERBAL INSTRUCTION TO CONDUCT OPERATIONS, provides direction for manipulations such as leak isolation. Following the closing of 1G33-F100 and IG33-F102, the resultant RR 'B' Pump Seal leakage did not change. However, the system ,

continued to be operated in this abnormal configuration for nearly 19 hours2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br />. This presents mixed l signals to station staff relative to management's expectations concerning procedure compliance.

)

• l I

9

o , .

l 0055-0623 .

] Action: W four hour Technical Specification Action 3.4.5.a time clock to resolve the leakage

! problem or commence a shutdown to Mode 3 expired at 0055. An orderly shutdown was l commenced per 3005.01, UNIT POWER CHANGES. This included reducing reactor power from 55% to 23% with reactor coolant flow and control rod insertion, shutdown of the B Reactor l Feed Pump, transfer of the A Reactor Recirculation Pump to slow speed, shifting from 3-element i feedwater control to single element, removing the Moisture Separators from wrvice, and dealing with a reported fire in the Service Building which turned out to be light smoke only.

l Basis: The unidentified leakage in the drywell of greater than 5 gallons per minute necessitated a ,

unit shutdown per Technical Specification 3.4.5. The shutdown was performed in accordance

with the UNIT SHUTDOWN procedure 3005.01.

l Management Involvement: The Shift Supervisor, Assistant Director Plant Operations and Plant -

l Manager discussed the shutdown evolution with Senior Management emphasis on following

! procedures and performing the shutdown carefully. The Supervisor Operations Support was l' ,

present during this time to provide management oversight of the shutdown. The Plant Manager left at 0117.

Evaluation: There were no noted problems observed while performing these varied steps in the

, shutdown sequence. The crew maintained focus, followed procedures, exercised proper 3 part

! communications, and fully implemented the STAR self cWWg technique.

I h Supervisor Operations Support filled out an OSO-086, OPERATIONS SELF j

ASSESSMENT TASK CARD, identifying the fact that the numerous feedwater heater level ,

j alarms due to the low power level and the 3 part communications between the Operator and the i Line Assistant Shift Supervisor during the acknowledgment and resetting of those alarms seemed more an exercise in verbatim compliance than a good operating practice. It seemed as though compliance to the vetions of OSO-090 EXPECTATIONS FOR CREW MEMBERS was a distraction to the operation of the plant.

0623 -

! Action: EQP 4402.01, PRIMARY CONTAINMENT CONTROL was entered due to receiving a j High Suppression PoolLevel alarm l Basis: h crew appropriately entered EOP 4402.01 when the High Suppression Pool Level l entry condition was received. Technical Specification 3.6.2.2 Suppression Pool Water level j action A requires the suppression pool level to be restored to within limits in 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />.

l Management Involvement: The Supervisor Operations Support was the Operations

! Management representative present to monitor plant shutdown activities.

j Evalcation: During the mid shift (2300-0700), the Shift Supervisor and crew were primarily

focused on shutting down the reactor and entering Mode 3. The 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> Technical Specification i time clock of specification 3.4.5.c, to be in Mode 3 by 1255 that day, was in effect. The Motor j Driven Reactor Feed Pump was started at 0550 in preparation for shifting reactor vessel feed j from Turbine Driven Reactor Feed Pump B to the Motor Driven Reactor Feed Pump which 1 requires a concentrated effort by the control room staff. At 0612, a report of a fire in the Service i Building was received by the control room staff. This diverted the crew's attention and required i

l 10 5 - .- - _

- - . _ ~ . - - - - _ . - - - - - - - . - _ . - . -_.

i ,

l

manning the fire brigade. It was later determined that no fire existed. The High Suppression Pool Level alarm occuned at 0623.

Since the April 9,1996 reactor scram and Mode 3 operation with main steam safety relief valves, 4

there has been increased safety reliefvalve leakage, necessitating more frequent use of l suppression pool cooling and water transfer to radwaste in order to maintain the required i suppression pool temperature and level. There is no evidence that the crew was expecting the '

high suppression pool level to occur. l l

l ' While the crev/s focus during abnormal conditions must be to deal with those situations, i

oversight of overall plant conditions during normal evolutions could help the crew anticipate other i needed manipulations. This is an indication that, although busy with the immediate tasks at hand

, in the control room, Shin Supervision and the Shia Technical Advisor were distracted from their l primary responsibility of oversight and review of all plant parameters, resulting in the entry into an Emergency Operating Procedure.

i 0644-0945 ~

l! Action: Continued the plant shutdown per UNIT SHUTDOWN 3006.01. This included starting  !

i and subsequently securing RHR Pump B from Suppression Pool Cooling, staning the A Electrode l l Boiler, performing a shiA turnover from mid shiA to day shiA, placing the Motor Driven Feed )

l Pump in single element automatic and shutting down the A Reactor Feed Pump..

Basis: The plant shutdown was continued in accordance with the UNIT SHUTDOWN procedure j 3006.01.  ;

j Management Involvement: The Supervisor Operations Support was present on mid shiA until i shiA turnover, and the Operations Task Coordinator was available in the control room area to

! provideinanagement oversight on day shiA.

Evaluation

There were no noted problems observed while performing these varied steps in the l j shutdown sequence. The crew maintained focus, followed procedures, exercised proper 3 part l communications, and fully implemented the STAR self d+4-N technique. -

l l

0945 4

Action: The B Reactor Operator transferred the 4160v and 6900v non-vital busses from the Unit j Auxiliary Transformers (UAT) to the Reserve Auxiliary Transformer (RAT) per 3501.01 HIGH j VOLTAGE AUXILIARY POWER SYSTEM step 8.2.1. The B Reactor Operator performing  ;

the evolution mistakenly turned the control switch slightly past the Neutral switch position when returning the switch to Neutral from the Closed position. This was done while transferring the l

. 6900v 1B non-vital bus from the UAT to the RAT.

Basis

The Unit Shutdown procedure 3006.01 step 8.4.4 addresses the transfer of the 4160v and 6900v non-vital busses from the UAT to the RAT per the 3501.01. This is part of the normal j shutdown sequence.

Management Involvement: A management representative from the Training Department was j . present and monitoring the control room activities.

j .

I1 a

l l .

e l .

I  !

Evaluation: This event was strictly operator error caused by lack of attention to detail by the  ;

operator. This resulted in the opening of the 6900v IB bus Reserve Feed Breaker, momentary deenergization of the 6900v IB bus, and the auto reclosure of the Main Feed Breaker which reenergized the bus.

The impact was the B Circulating Water Pump tripped, the Reactor Water Cleanup B & C Pumps tripped, the A Electrode Boiler tripped, Stator Water Cooling Pump A auto started, the Turbine Bearing Lift Pumps auto started, and the RPS Solenoid Inverter B trouble alarm was received. In addition, the Motor Driven Reactor Feed Pump fed off 6900v IB was running and feeding the reactor vessel, but fortunately did not trip. This would have resulted in a reactor scram. i l

This resulted in delaying the shutdown process by diverting the crews resources to resetting the '

RPS inverter trouble, restarting the B Circulating Water Pump, restarting the A Electrode Boiler which would not run (B Electrode Boiler was eventually placed into service), shutting down the A Stator Water Cooling Pump, and dealing with a Reactor Water Cleanup System restart. This complication placed added pressure on the crew to get the plant shutdown in the Technical Specification imposed 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> to hot shutdown time clock.-

The B Reactor Operator readily admitted his error and what he believed to be the cause. The Electrical System Engineer was present and verified this on the prints. This greatly reduced the recovery time. The forthrightness of the Operator is commendable.

1005-2316 Action: Continued the plant shutdown per UNIT SHUTDOWN 3006.01. This included completing the 4160v and 6900v non-vital bus transfers, restart of the B Circulating Water Pump, s.artup of the B Electrode Boiler, reducing reactor power to 17% with control rods, removing the

- generator from the grid and shutting down the turbine, shifting steam seals to Auxiliary Steam, performing a reactor scram as part of the normal shutdown procedure, returning the Reactor Water Cleanup System to service, performing a cooldown to Mode 4, and exiting the Notification _

OfUnusual Event.

When transferring steam seals to Auxiliary Steam, the IGS041 Auxiliary Steam to Gland Seal Supply Valve would not open electrically. The operators turned off the feeder breaker to the valve and manually opened it.

Basis: The sEutdown was continued in accordance with UNIT SHUTDOWN procedure 3006.01.

Management Involvement: The Supervisor Operations Support and Assistant Director Plant Operations were available in the control room area during these evolutions.

Evaluation: There were no noted problems observed while performing these varied steps in the shutdown sequence. The crew maintained focus, followed procedures, exercised proper 3 part communications, and fully implemented the STAR self-checking technique.

] The normal reactor scram per the shutdown procedure was well briefed, all rods inserted as

expected,4401.01 RPV CONTROL was entered at the appropriate entry condition (level 3) and 12

exited when reactor water level recovered with normal feedwater following the sctam. The cooldown rate was appropriately followed.

The logical corrective action to deenergize and then manually open the IGS041 valve by the operators showed good trouble shooting techniques that kept the shutdown progressing with minimal delays due to equipment malfunctions. MWR D61643 was written to document the IGS041 valve problem.

This completes the analysis portion of this report. The six major issues that are identified as a result of this evaluation are discussed in the following Conclusion section of this report.

~

l

)

egen b ,

e 13

O , i w

II. Conclusion Management penonnel were not conservative in the operation of the plant.

Numerous opportunities existed for management to evaluate uncertainties and the need for continued reactor operation, but these were not recognized or acted upon.

When leakage initially exceeded 5 gpm, and a NOUE was entered, the crew, in accordance with Technical Specifications, focused on isolating the leak within the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> action statement and did not give adequate consideration to initiating a Reactor shutdown.

Subsequent events complicated the situation at which time the crew could have re-evaluated the need for Reactor shutdown:

Leakage increased to greater than a 2 gpm increase in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> Shutting the CRD injection valve (C11-F026B) to the RR pump seal Seal Failure Drywell pressure increased requiring mixing compressor startup Continued leakage greater than Technical Specification limit following seal failure Loss of LD-27 indication for unidentified leakage.

The Shift Supervisor, Assistant Director of Operations, Shift Technical Advisor and the Plant Manager did not perform the expected oversight role in considering the overall picture of plant conditions.

Ineffective application of previous lessons learned from industry experience.

~

Management did not properiy establish, enforce, or set the proper example for procedure compliance.

Decision to isolate CRD injection valve to the RR pump seal Decision to perform double isolation ofRT and subsequently operate the system in a manner not specifically covered by the procedure.

Management provided an action plan for the evolution that had not received an independent technical review and was not specifically covered by the procedures.

1

Oversight of the overall picture of plant conditions and actions surrounding the event was ineffective. '

Shift management did not maintain proper oversight and was too involved with details regarding leak isolation.

Senior management did not maintain proper oversight and was too focused on leak isolation.

' 1 Management decisions regarding actions to operate the plant were determined not to be in '

accordance with procedures.

During the decision process, shift management did not involve all the licensed members of the operating crew and thus did not have an opportunity to hear dissenting opinions.

No one called a " time-out" but instead focused on leakage isolation, consequently more conservative options (such as reactor shutdown) were not evaluated.

4 There was inadequate planning and evaluation of potential consequences prior to j

performing an infrequently performed operation.

i

• Inadequate Action Plan (letter from Mosley to Shift Supervisor entitled, " Guideline for RR "B" Loop Isolation and RF Leakage Evaluation")

Insufficient evaluation of potential for valve leak-by RT potential leak-by on isolation (Procedure was thought to address double isolation, but did not) j -

Potential leak-by on RR isolation valves 1 MWR's existed which indicated leak by was potentially present on both trains (D63773, dated 4/8/95 and D76000, dated 3/8/96)

No upper leakage limit set for beginning an orderly shutdown.

It is not clear that any previous evolutions to .mter single loop operation, planned or

emergency, were evaluated prior to this evolution.

D.

I I

L 2

Management tolerated long-standing equipment problems that contributed tp uncertainty of some plant conditions.

The following long-term deficiencies were associated with this event:

- LD system V-notch indication routinely inoperable

- Prolonged operation with a degraded seal

- Leak-by on RRisolation valves

- Dual position indication on RR isolation valve Clinton's methods of establishing limits for operation with degraded equipment do not place sufficient emphasis on operational needs.

Long-standing equipment deficiencies, such as drywell leakage system, reactor i recirculation pump seals, and reactor recirculation pump suction and discharge valves ,

further complicated decisions to mitigate the event.

IP personnel were not timely in identifying a potential procedure non compliance and management did not sufHeiently pursue other indications that would lead to management recognition of the significance of non-conservative reactor operation. It should be noted that IP was prompted by the NRC to punue the issues of procedural guidance.

Self assessment was ineffective /madequate:

Operations did not recognize the procedural non-compliance at the critique immediately following the event.

- Procedural non-compliance was not fully recognized by the station until the I Wednesday following the event.

Management personnel did not recognize message being received from the NRC.

I

~

L 3

O , e e

ATTACHMENT A SUMMARIZED CHRONOLOGY l

Date: September 05,1996

)

Prior to Time Line of 17301800 I Prior to the evolution, the Assistant Director of Operations, acting as the Director of Plant Operations, met with the Shift Supervisor and provided a written summary of  ;

the management expectations (see Attachment 1 to this Chronology) to the Shift l

Supervisor. These expectations were identified as a tool and were not intended to I conflict or override CPS Procedures or Technical Specifications. The expectations contain the following:

1

1. As reactor power is decreased to <70%, monitor drywell floor drain (RF) l leakage for changes.

2. As the flow control valves are manipulated, monitor drywell RF leakage for changes.

3. When the "B" reactor recirculation (RR) pump is shutdown, be prepared for seal leakage to rapidly increase and expedite loop isolation, if required.  :

l

4. Once the "B" RR pump is shutdown and the loop isolated, the following may be expected to happen:

l Drywell RF leakage remains relatively unchanged. Then consider isolating the Reactor Water Cleanup System (RT) from the RR loops -

._ to determine if valve 1G33-F106 is leaking by its seat and adding to the leakage into the RR loop. If this does not cause the drywell RF leakage to decrease then CPS Management will decide what the next  ;

course of action will be

Drywell RF leakage decreases. CPS Management will decide if continued operation in single loop is appropriate or not.

Drywell RF leakage increases. If the leakage exceeds 4.5 gpm, then continue to shutdown the plant and go to Mode 4. If the leakage is j less than 4.5 gpm, then CPS Management will decide the next course of action.

1 l

l

I i

t .

) Elapsed time: 0 hr. ,

Date: 9/5/96 i

Time: 1730-1805

Plant Status:

! e Rx Power: 100%

[ e Seal Pressure P1:

l e Seal Pressure P2: 479 psig (time 1753)

• Seal Temp:

i e Drywell RF Leakage manual calculation sump fill time: 4.27 gpm (time 1744)

• Drywell RF Leakage manual calculation pump run time:4.64 gpm (time 1744)

A special briefing was held from 1730 to 1800 with the following personnel in attendance:

STA. .~

Electricians who would take valve signatures to verify proper and complete closure of the valves.

C&l Techs who would be responsible for completing the appropriate surveillances within the required time limits once we achieved single loop operation.

The system engineer.

" VOTES" engineer who would direct the electricians.

A nuclear engineer (separate from the STA) who was there to discuss control rod manipulations.

The Assistant Director Plant Operations (also filling in for the Director Plant Operations during his absence while on vacation).

• Shift Supervisor. -

Line Assistant Shift Supervisor. j

_~

• The RO's taking part in the evolution. 1 Some outside area watches.

There were various things discussed and assignments made, including:

L

1) One RO was assigned and discussed the normal shutdown procedures and process.

2) One RO was assigned and provided a briefing on the procedures and process to be used should a catastrophic failure of the RR pump seal be experienced.

3) The RR system engineer stated that the most likely point in the evolution where catastrophic failure might occur would be when the flow control valve is closed to minimum position immediately prior to 2

i o ,

shutting the RR pump down. (The system engineer cautioqed the operating crew to not delay at this point.)

4 1

-m 1

h h

M b

e 3

o . .

4) Also discussed was that during this evolution, an increase in the unindentified leakage could put the plant in an emergency plan action level. The specific leakage criteria was not discussed.

5) Procedures discussed during the briefing included the following:

3302.01 Reactor Recirculation i 3005.01 Unit Power Changes I- 4001.01 Reactor Coolant Leakage 4008.01 Abnormal Reactor Coolant Flow 4100.01 Reactor Scram

6) During the procedure discussion, the briefers did refer to specific steps and actions to be taken during the evolution. Specific i questions were asked and answers were provided with the assistance '

of the system engineer.

7) The first contingency discussed was that if catastrophic failure occurred,they would expect to see certain indication 1. These indications included:

- Greater amount of leakage

- Increase in drywell pressure

- Decrease in RR pump seal pressure

- Increase in Drywell RF leakage J) The second contingency discussed was that right at the end of the day shift operators had noted a blip on a GETARS point for the feedwater system. The contingency plan was if they have a problem with the feedwater pump / pump control circuit they would take the ,

pump off-line and operate with the one remaining pump. Personnel at

+

the briefing indicated that this action made sense.

9) The sequence of shutting valves and movement of monitoring

- ' equipment was discussed with the electricians. Provisions were made to instrument the 1B33-F067B RR Pump "B" discharge valve with VOTES equipment in order to verify positive valve closure. This <

precaution was taken since there was an existing MWR (MWR D63044) related to a dual position in[ication problem on this valve.

Since this equipment was available, it was decided to also use it on the 1833-F023B RR Pump "B" suction valve.

O 4

i l

i , .

l Elapsed time: 35 min. .

Time: 1805-2009 Plant Status:

• Rx Power: 100% Procedures Entered /In Use: 3005.01 (Unit power Changes)

Reduced reactor power in preparation for shutting down the "B" Reactor Recirculation Pump. Power was reduced from 100% to 69%.

Elapsed time: 2 hrs. 39 min.

i Time: 2009 l- ~ ,

j Plant Status:

• Rx Power: 69% Procedures Entered /In Use:

i 3302.01 (Reactor Recirculation) l e Seal Pressure P1: 985 psig

(

• Seal Pressure P2: 479 psig (time 1753) l e Seal Temperature Cavity 2: 109 *F '

I e Drywell RF Leakage manual calculation sump fill time: 4.27 gpm (time 1744) e Drywell RF Leakage manual calculation pump run time: 4.64 gpm (time 1744) i k l 4 dhutdown the "B" RR pump in accordance with the normal RR operating procedure, 3302.01, Step 8.2.1. The crew tripped the pump, closed the discharge valve . -

1B33-E067B. The procedure had an instruction to reopen 1B33-F067B in 5

minutes if it was not planned to isolate the loop. Final power level was 58%. At

this time there were electricians et the valve breaker who had the valve motor  !

l instrumented to verify proper closure. Indication for proper closure included: a) l l verify signature traces on the valve current to verify it seated properly; b) verify the time for stroking the valve to ensure full travel (spikoximately 120 seconds); c)

{- verify closure indication on control room panel. The valve was instrumented

! because there was some uncertainty that the main control room indicating lights would indicate full closure due to the existance of MWR D63044 as noted earlier.

A contingency plan was established to restroke the valve 1B33-F067B if 1

) intermediate position indication was received on the control room panel.

l Instrumentation at the valve locally would again be used to establish a trace on the 4 motor current and to determine the stroke time to verify proper and complete l closure.- If the restroking did not clear the intermediate position indication, they would consider the valve closed provided the tracer data indicated proper closure.

1 e

5

- -- . _ . . _ _ - _ - - = _--

t s , .

This contingency, to have the electricians set up to take tracer data, had_been set up on day shift in advance.

The seal P1 and P2 pressures commenced to equalize. Fully opened the RR Flow Control Valve (FCV) 1B33-F060B per 3302.01, step 8.2.1.5. The pressure across the pump seals had equalized shortly after pump shutdown, approximately 7 minutes later. The operating crew and the System Engineer were surprised when the pressure equalized. The System Engineer talked to several people about a theory that "the lower seal was not doing much such that the leakage past the lower seal was greater than what was going out the upper seal. Hence, P1 is P2.

The upper seal was carrying full pressure while the lower seal was doing nothing."

Seal temperature was approximately 130*F when the pump was shut down, i

l Elapsed time: 2 hrs. 39 min.

Time: 2009-2030 Plant Status:

• Rx Power: 58% Procedures Entered /In Use:

3302.01 (Reactor Recirculation)

• Seal Pressure P1: 985 psig (time 2013)

• Seal Pressure P2: 976 psig (time 2017)

• Seal Temperature Cavity 2: 109 'F (time 2017)

• Drywell RF Floor Laskage manual calculation sump fill time: 4.47 gpm (time 2030)

• Drywell RF Floor Leakage manual calculation pump run time:5.167 gpm (time 2030)

Commenced Loop Isolation per 3302.01, step 8.2.4. -

~ Step 8.2.4.1 Shut 1G33-F106 RT suction from RR Loop "B" Step 8.2.4.2 1833-F067B Discharge valve was already shut from section on pump shutdown, Step 8.2.1

~

L Step 8.2.4.3 Not currently in emergency condition, so did not perform this step. (i.e., did not qualify for the "lF" statement so didn't perform the "THEN" statement.)

Step 8.2.4.4 At this point, operators began to evaluate the l conditions. The step does not say to wait, it just says

! that certain conditions must be met prior to performing the isolation.

5 0

6

~

l Elapsed time: 3 hrs. ]

l Time: 2030 s

Discussed closing 1833-F0758 (RR Pump B Seal Staging Shutoff Valve) with the

] system engineer and Assistant Director Plant Operations to increase the flow to the RR loop from the control rod drive (CRD) system. (This diverts flow from the seal).

The RR loop temperature was around 530 *F and had only decreased about 10 'F. ,

The STA or one of the RO's performed a quick calculation and estimated that the )

time it would take to reach cool down would be about 6 to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. We then had a  !

l discussion on whether we would damage anything by shutting the 1833-F075B. )

The engineer indicated it would not harm the seal or the system because the valve l originally closed automatically whenever the pump was shut off. (A modification to

provide manual vent and drain capability had since been added and the valve no longer closed automatically on pump shutdown.) _ )

Shut 1B33-F075B Pump B Seal Staging Shutoff Valve per 3302.01 step 8.2.4.5 in j an attempt to increase the cooldown rate of the recirculation loop. At this time, operators were monitoring parameters to get the loop temperature down to less

< than 250*F to allow completion of steps 8.2.4.6 and 8.2.4.7, in order to complete i the loop isolation.

I 4 Floor drain flow rate was approximately 4.7 gpm. RR B" loop suction temperature I was 511

• F, seal cavity 2 temperature increased to = 150 'F. Such an increase in

! seal temperature was expected since there is no forced flow through the seal cavity j HX with the pump off.

i l -

Elapsed' time: 3 hrs. 25 min.

Time: 2055

]

Plant Status: L

• Rx Power: 58% Procedures Entered / Ira Use:

• Seal Pressure P1: 985 psig (time 2013) 3302.01

• Seal Pressure P2: 976 psig (time 2017) 4001.01

• Seal Temperature Cavity 2: 109 'F (time 2017) 4008.01

• Drywell RF Leakage manual calculation sump fill time: 5.9 gpm (time 2053)

• Drywell RF Leakage manual calculation pump run time:5.75 gpm (time 2053)

• LD-27: 5.51 gpm 7

I Floor drain flow rate increased to 5.51 gpm. Action 'A' of Technical Specification j (TS) 3.4.5 was entered which required restoring leakage to less than 5 gpm within j four hours. TS 3.4.5 requires that if leakage is not restored with in the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> time i period, the plant must be shutdown within 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />. 4001.01 Reactor Coolant Leakage Off-Normal procedure was entered at this time. 4008.01 Abnormal Reactor Coolant Flow was entered, conditions evaluated, and then exited.

l

During this time frame, the SS stayed in the MCR to provide oversight, monitor i leakage, monitor seal pressure, and monitor the actions of the crew.

{ The SS was looking for a change in LD readings. After the sump pump cycled, the next reading was about the same as the first. Once he was satisfied that leakage  :

, was not increasing at too high of a rate (he wanted to be there if conditions

! deteriorated rapidly and he was comfortable with crew actions), he went to the i Emergency Plan.

i j Elapsed time: 3 hrs. 30 min.

Time: 2100 The Plant Manager was informed by the Assistant Director - Plant Operations that

they were conducting the loop isolation, that the pump was stopped; the discharge i valve 1B33F067B was shut while waiting for the loop to coldown to 2FO*F, the l leakage had crept up to over 5 gpm, and they were in Action "A" of TS 3.4.5. ,

i Notification Of Unusual Event was declared per EC-02, step 4.1 Leak Rate Exceeded / Loss of Inventory.

j SS went to get the Emergency Plan bag, checklist and immediately returned to the '

j horseshoe. He began concentrating on getting notifications made in the correct  !

l amount of time. _ j i He called security to have security to set off the pagers for the Unusual Event. The l SS then asked the STA if he was comfortable with making the appropriate i notifications. He responded YES and took the assignment to make the notifications.

{

l NOTE: After the event, looking back, the STA say that if he had known then what

! he knows riow, he would have said NO. The SS looking back agrees that he should l have probably gotten an extra operator to make the notifications.

Elapsed time: 3 hrs. 46 min.

1 Time: 2116 1-4 Plant Status:

9 4

8  !

7

r- 1 l

j t -

• Drywell RF Leakage manual calculation - sump fill time (6.3 gpm)

Burped the drywell using the "A" Hydrogen Mixing Compressor. l l

l .

I Elapsed time: 3 hrs. 55 min.

Time: 2125 Plant Status:

State LEMA and IDNS were notified of Unusual Event declaration. ,

(Emergency Plan requires 15 minutes for State notifications,1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> for NRC l notifications). l l l i

l l

Elapsed time: 3 hrs. 57 min. _

Time: 2127 l

Plant Status:

I Secured the "A" Hydrogen Mixing Compressor. Drywell pressure: 0.23 psig.

Elapsed time: 4 hrs.

l .

Time: 2130 _

l Plant Status: )

1 1

• RX power 58%

e Seal Presure P2: 982 psig.

~

L.

Thoughts and considerations by the SS at this time included the following:

• Had proper indication that loop had been successfully isolated.

• The SS considered entering the Emergency step (and hence perform l 3302.01 step 8.2.4.3). Inputs into this consideration involved:

- are we putting control rod drive pressure into the seal cavity (1500 psig to 1800 psig).

-if staging line is open, we are removing flow at 1 gpm, if closed, no flow is being removed.

9

l l= - so, we have 3 5 gpm control rod drive flow going to the isoloted

. loop which will pressurize the isolated loop lifting the injection relief i valve at approximately 1250 psig.

- At this time, indications were: that the loop was isolated, that i control rod drive was still injecting and that the seal staging flow valve was shut.

- The SS expected to see the following:

- If the seal leak rate was greater than CRD injection rate, we would i

expect to see the seal pressure start to slowly decrease.

]. OR

! - if another drywell leak was present, which was masked earlier, and

in conjunction with the seal not passing 5 gpm, would expect to see the seal start to pressurize. (if this occurred, it would mean leakage

was coming from somewhere else).

I This would require reopening the suction valve so pressure could be  :

j relieved to the reactor.

. During this time, the SS talked to the Assistant Director Plant Operations, explained

! that we were at a greater than 5 gpm leakage, and told the Assistant Director Plant j- Operations that he felt this constituted an emergency situation and that we needed 1

! to return to step 8.2.4.3 of Procedure 3302.01, because we now met the "lF" l

~

j statement.

I l' He also judged that this qualified as an emergency because we had entered the i Emergency Plan.

Additiorvally, he felt this qualified as an emergency because the shift had entered ,

! the Reactor Coolant Leakage procedure (CPS 4001.01), which says to isolate the j leakagFin step 4.6.

i l The SS also thought he had the means to isolate the leakage. The Assistant Director Plant Operations agreed.

l

• One RO was surprised that we went back ththe step to close the suction I valve (step 8.2.4.3) 4 - The SS explained the logic to the RO and after explaining the logic, the RO

indicated that closing the suction made sense.

2 - The SS does not remember any other challenges to the procedures being used or the steps being taken or the timing of these items at any time during

j. the evolution.

i l l Operators proceeded to isolate the "B". RR loop by performing step 8.2.4.3a l 1- of 3302.01, shut 1833-F023B Pump Suction Valve. This step was now

! performed since the " Unusual Event" was considered an " emergency A

{ 10 i

condition" as referenced in this step.1B33-F075B in 8.2.4.3.b had been previously shut per step 8.2.4.5.

NOTE: Nothing in the procedure provides guidance on what to do if you get halfway through the evolution and the conditions change, which is the situation we were now in.

As noted earlier, the electricians had set up to monitor the loop suction valve j (1833-F0238). When the valve wa's closed to complete the isolation, all the l correct indications were received. l The valve timing was right i

• lt had a good trace i j The Control Room indication indicated full close.

i Seal pressure was observed to stay about the same and decrease just 5 slightly. The SS watched seal pressure for about 29 minutes and ,

indicated leakage started to go up.

~

Elapsed time: 4 hrs. '

i '

Time: 2130-2159 l Plant Status:  !

l i

l Main Control Room Personnel were monitoring floor drain leakage rate and seal pressures for change. During this period of time (2144), reactor coolant leakage '

l exceeded a 2 gpm increase in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. LD readings indicated approximately 6 l gpm. Readings taken 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> prior to this reading were 3.82 gpm. Seal pressures ,

during th' time period were steady at 980 psig on P1 and P2. Technical '

Specifloation 3.4.5 was re-entered for the 2 gpm leakage increase.

l l l

Elapsed time: 4 hrs.14 min.

Time: 2144 Plant Status:

l

• RX power 57% (2133) e Seal Pressure P2: 979 psig. (2137) ,

j e Drywell RF Leakage - manual calculation - sump fill time: 6.43 gpm. l I

NRC notification was performed for the Unusual Event declaration.

4 L

11 l

l

w Elapsed time: 4 hrs. 29 min.

Time: 2159 Plant Status:

• RX power 57% 985 psig. (2133)

• Seal Pressure P1 & P2 979 psig. (2137) e Seal Temperature cavity 2: 141 *F (2109)

• Drywell RF Leakage - manual calculation - sump fill time: 6.43 gpm.

SS thoughts and considerations during this time frame:

• The loop is isolated and not communicating with the reactqG o Leakage is about the rate of CRD injection flow (3-5gpm CRD injection flow with an overall indication of leakage at 6gpm).

Logic indicated that we are currently pumping CRD water into the drywell.

Held a discussion with the Assistant Director Plant Ops and the system engineer and discussed the following:

Pressure is going down real slowly.

Pressure started about 980psig when we shut the suction valve and isolated the loop.

- At this time, loop pressures were down slightly.

' The seal leak and the CRD injection were about equal at this point. This indicated we could have continued all night in this situation and not had any additional success at isolating the leakage.

The primary focus of the overall evaluation and the SS's objective was still -

_to try to isolate the leakage.

- The SS wanted to deviate from the normal procedural requiremet to reduce loop temperature to less than 250*F or wait until the loop is depressurized to about drywell pressure in order to comply with the Reactor Coolant Leakage procedure 4001.01 to locate and isolate the leak. I Reasoning included: L We couldn't depressurize because of CRD injection.

- We appeared to be experiencing a very slow cooldown and wriu!d have to i wait 6-10 hours to shut off injection flow in order to meet the parameters specified in the procedure for RR loop temperature to be less than 250 F.

The CRD injection appeared to be keeping the leakage high.

Subsequently, when the SS was asked by the Assessor whether there was any discussion that taking 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> to shut the CRD injection would not meet the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> criteria to get below 5 gpm. His response was that this was not a consideration. There was no discussion at this time  ;

related to not meeting the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> criteria. The primary focus was that he  ;

12

~.. - . . . ... _ _ _ ._.. .__. .._.____. ._ _ _ - _ _ _ _ _ __ . .

l-i

related to not meeting the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> criteria. The primary focus was that he l had leakage that he was causing and he wanted to stop it.

l -

As far as the SS knew at the time, no one had indicated in any way that i they were concerned that we were trying to beat the clock or in any way l prevent a shutdown.

t l

l w

due l

l L

f I

13 i i I I l

The procedure also contains a Caution identifying that seal damage will qccur at 250*F. The basis for the caution was discussed. The System Engineer believed that the Caution in the procedure was then mainly to protect the 0-rings, because in the April event, he knew that the 0-rings were not the limiting factor. The O-rings were tested at 600*F for 8 hrs. at Bingham. The SS, RR System Engineer, and the Assistant Director Plant Operations discussed the Caution in the procedure about closing 1C11-F0268, this was discussed carefully however with some urgency. They decided that catastrophic seal failure was not likely and that further degradation of an already damaged seal was acceptable. (in retrospect, the System Engineer noted, perhaps 1C11-F026B should have been shut slowlv.) The crew was hoping for a gradual relief of the pressure, because the upper seal was j expected to hold. Going to mode 4 was also discussed as the next alternative. j The System Engineer said that we would probably have some degradation on l the seal faces from thermal expansion when the CRD flow was stopped but he didn't expect a catastrophic failure. .

Even if it did fail catastrophically, there was only a finite volume of water to escape because the loop was isolated. _ .

l Did not have signature trace on the reactor water cleanup system (RT) .

isolation valve. Had discussed 1G33-F106 valve isolation because there I was a potential RT water could leak back into the RR loop. However, the I leakage in gpm was approximately equal to the CRD injection flow in opm l and with pressure coming down slowly, he thought the isolation was -i holding.  !

At this time we analyzed the 2 cautions contained in CPS 3302.01, Step 8.2.4.

This first caution identifies that there is potential to damage the seal if the l seal temperature is allowed to rise to 250*F. (The seal temperatures were

. Iow at this time, and we expected that they would increase).

- The second caution indicated that we would increase airborne activity in the _

drywell. We thought we were pumping in pressure (and leakage) and had to Dop inserting flow and pressure to let it depressurize and stop the leakage.

This would allow the reactor coolant pressure locked in the loop to depressurize out through the seal, increasing airborne activity in the drywell.

l We also considered that if depressurization occurred fast enough, we might i get depressuruation before the seal temperature got high enough to cause l l

any damage. l l - This was the only item that the SS did not feel personally comfortable with. '

He expected to contaminate the drywell and he discussed this with the Assistant Director Plant Operations.

Neither the SS nor the Assistant Director - Plant Operations thought that this would be significant or should be weighed heavily as a constraint. (The SS did not ask for an opinion from RP and to his knowledge neither did the

, Assistant Director Plant Operations.)

The SS did not think it was necessary to discuss with RP because stopping i

! the leakage was more important than possibly contaminating the drywell.

l I

l l 14

- The SS discussed this with the Assistant Director Plant Operations and then made the decision to cut off the CRD flow to the seal. This was the only known item performed that was outside of the normal procedural guidance.

This was 'outside the procedural guidance because at the time this occurred the conditions required to allow shut off of CRD flow to the seal, were not met. However, this did meet the guidance.of the Off-Normal to locate and isolate the leakage.

The SS evaluated that he could have waited 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> for the cool down of the loop to occur, but he saw no reason to wait.

- He knew we had a damaged seal and would be replacing the seal, at least

by RF-6 and was not worried about additional minor damage to the seal.

- From discussion with the System Engineer, the key reason procedure step 8.2.4.4 was there was to prevent seal damage. Since we were going to replace the seal anyway, this was not a critical issue.

SS decided, with the concurrence of the Assistant Director Plant Operations, to shut the seal injection (1C11-F0268) per 3302.01 step 8.2.4.6. The SS gave direction to LASS to shut 1C11-F0268. At this point, the SS went

back to monitoring the conditions. He expected seal pressuge to go down i aM leakage rate to decrease following pressure. This appeared to occur.

i Elapsed time: 4 hrs.' 36 min.

l Time: 2206 i The Assistant Director Plant Operations briefed the Plant Manager on plant 1 conditions which were: loop isolated, CRD injection isolated; leak rate about 5.5 i gpm, and seal pressures slowly falling. The Assistant Director - Plant Operations -

indicated loop isolations (valve closures) looked good. The Plant Manager agreed with him that it appeared the loop was isolated (based on falling seal pressures) and _

suggested the leak rate may take awhile to fall since it (the loop) would essentially depressbrize its inventory out the seal. The Plant Manager concurred in the approach, l

Elapsed t me: 4 hrs. 37 min.

I Time: 2207 Plant Status: i e RX power 57% i e Seal Pressure P1: 985 psig & P2: 967 psig. (2205)

= Seal Temperature Cavity 2: 138' F e Drywell RF Leakage manual calculation - sump fill time: 6.5 gpm O

15

l l

Time: 2207 l Security reported that all the ERO notifications were complete for the Unusual Event. i I

Elapsed time: 4 hrs. 47 min.

Time: 2217 Received a seal cooler outlet temperature alarm on the "B" reactor recirculation pump.146*F per 5003.01-1K. This was expected with CRD seal water secured.

Elapsed time: 4 hrs. 52 min.

Time: 2222 Plant Status:

Received a seal cooler outlet alarm on our recorder (annunciator on panel "High Temp - RR pump B" which sends you to the recorder).

(We were expecting to get alarms on the temperature, so this was not unexpected.)

• Doesn't remember temperature (alarms at 146*F per 5003.01-1K).

The SS had an extra RO assisting anithad him go to check the alarm.

As he came back to report the temperature, sudden depressurization occurred.

When rapid depressurization occurred, the LASS directed the crew to start the hydrogen mixing compressors, which pumps drywell atmosphere into containment suppression pool.

The evacuation of containment was also directed.

16

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

i. 1 i

i At this time, seal pressure had dropped from about 900 psig to approx 280 i psig.

- The SS again started monitoring.

• Expected the drywell pressure would go up (was at approximately 0.23 after having burped the drywell).

Also got a drywell air cooler drain flow alarm coming in. (One air cooler off

each drywell chiller has condensate metered before going to the floor drain.)

l This indicates a steam leak from steam getting condensed in the cooler and going to the floor drain. This also adds to the total drywell RF leakage which was the parameter of concern.

l

• The drywell air cooler rate peaked at approximately 3.5 gpm.

i i

• During this time (over about a 15 minute time period) the SS monitored the

, plant conditions.

e l

• Once the SS was comfortable that drywell pressure would$ot be a problem i he resumed evaluating the floor drain rates. j i

e

! Elapsed time: 4 hrs. 53 min.

i

- Time

2223 I

l- Looking at the chart recorder, the Plant Manager and Assistant Director Plant Ops saw 7.98 gpm two times in a row following the seal failure. Believing two exact i numbers to two decimal points to be bad data, informed the SS. The SS seen the .

! Display Control System (DCS) displaying white data indicating bad input data and j had the STA performing manual calculations. The LD system engineer was then

]. called.

i

~

f L Elapsed time: 4 hrs. 56 min. j I

Time: 2226 Plant Status:

• RX power 57%

• Seal Pressure P1: 318 psig. & P2: 308 psig.

• Seal Temperature Cavity 2: 174* F (2221) e Drywell RF Leakage manual calculations - sump fill time: 21 gpm 17

4

• LD-27 > pegged high 8 gpm. .

l i

Stopped B Mixing Compressor at a drywell pressure of .40 psig. Peak drywell l

. pressure was .45 psig, total run time of 4 minutes. Drywell pressure continued to j decrease to a pressure of .12 psig as steam was being condensed in the drywell.

l No drywell vacuum breaker opened.

j. This indicated to the crew that leakage was getting better and the seal p depressurization was proceeding as expected.

4 i

Elapsed time: 5 hrs. 7 min.

i Time: 2237 i

Plant Status:

j

• RX power 57%

• Seal Pressure P1: 185 psig. & P2: 187 psig.

I

• Seal Temperature Cavity 2137* F (2239)

{

• Drywell RF. Leakage manual calculation - sump fill time: 23.5 gpm i

• Drywell RF Leakage manual calculation - pump run time 38.1 gpm

!

• LD-27 > pegged 8 gpm. )

)

, The SS was informed by the STA that leakage was at approximately 38 gpm '

4 and a short time later the STA informed him it was going down. -

~

, The SS expected that the leakage rate would initially increase and then start j coming down.

<

• The Assistant Director Plant Operations said that he thought LD27

j. irttrumentation was clamped at 8 gpm. L i-About the same time, the LD system engineer called and said that LD27 did, in fact,' clamp at 8 gpm.

• Simultaneously, the oncoming STA also informed the crew that LD27 clamps at 8 gpm.

• At this time the thoughts and considerations of the SS considered the following:

18

- The SS still felt good because everything was occurring as expected.

- His first expectation that pressure would go down was occurring. l

- The second expectation was still being monitored by the SS in that he was expecting the leakage to start going down also.

• When the SS received the second set of floor drain flow data, everything continued as expected and the leakage did start going down.

l

• The following things were also occurring in this time frame. i

- The drywell cooler drain flow alarm had cleared. (This alarm clears at l approximately 1.6 to 2 gpm.)  ;

- As noted, the floor drywell RF leakage started going down. )

- Drywell pressure started coming down. (The drywell pressure had peaked at i O.45psig and the mixing compressors had been stopped at 0.40psig.) i

- Drywell pressure continued to go down and reached a value of 0.12 psig. l

- Because the volume contained in the isolated loop had depressurized through I the seal, there was a potential that a vacuum would be pulled in the drywell. t However, the drywell vacuum breakers did not activate, indicating that no vacuum was experienced. ._ l i

• At this point the SS believed that he had accomplished all objectives and again went into the monitoring mode. j i

• When relieved, drywell RF leakage was at approximately 12 gpm, down from the 38 gpm peak. l l

i Elapsed time: 5 hrs. 25 min.

Time: 2255 j Plant Status:

e RX power 57%

i e SeaTPressure P1: 121 psig. & P2: 120 psig. (2257)

• Seal Temperature Cavity 2: 142' F (2245)

• Drywell RF Leakage manual calculation - sump fill time: 15.07 gpm (2250)

• Drywell RF Leakage manual calculation - sump run time: 15.73 gpm (2258)

I '

L ,

> l

~

Recognized that drywell floor drain flow rate indications were inoperable. Entered

actions for leak detection instrumentation per Technical Specifications.

j Elapsed time: 6 hrs.  ;

Time: 2330 1 3

19

, - - - , ,v~.

i 3

Plant Status: -

1 l e Seal Pressure P1: 121 psig (2257) & P2: 99 psig. (2389)  ;

, e Seal Temperature Cavity 2: 1500 F (2333) _

e Drywell RF Leakage manual calculation - sump fill time: 11.08 gpm (2384) l j e Drywell RF Leakage manual calculation - sump run time: 14.06 gpm (2329) i Calculated value for drywell floor drain flow rate was 14 gpm based on sump pump '

{ run time.

t  :

The crew felt this flow rate decrease was also an indication that the finite volume

of coolant in the B recirculation loop was depressurizing and that the leakage would continue to decrease.

l Note: From about 2237 until turnover at about 4 322, there had been 5 dual' manual cales taken and leakage rate had decreased from a geak value of 38 gpm to between 12 and 14 gpm. This indicated that leakage was still decreasing and under control. At this time, there was about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 20  ;

minutes left on the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> Technical Specification Action statement for l isolating the leakage to within limits. ,

i

} '

! Shift Turnover 3-

! At this point, the SS began discussions with his relief including the current plant t status and how they got there. As part of the turnover, the SS indicated that they

were looking for the leakage to continue to go down and hopefully go below the 5 i gpm from the Technical Specification limit and the 4.5 opm administrative limit.

i The SS indicated that at the turnover there were no specific concerns identified i from tMB oncoming SS. The oncoming SS just wanted to know what was the plan L from here. The options provided were the following:

l i -

The first, was to see if the leakage stablized at a value below the ,

l' - administrative limit. If yes, then maintain single loop operation and l the decision would be made the nextDay on what to do next. It was somewhat anticipated that if leakage was successfully reduced, we l 3 would probably maintain the plant in single loop operation until RF-6. I I

l The second option, was if the leakage did not stabilize below the Technical Specification limit of 5 gpm, the shift could perform an l f orderly shutdown of the plant. j i

When questioned by the interviewer, the SS responded that there 1 was no discussion at any time about delaying the shutdown or 1 shutting down slowly to allow leakage to become lower.

20 E

1

1 i

Elapsed time: 6 hrs. 44 min.

Date: 9/6/96 Time: 0000 From R. Moraanstern Written Summaryl l

Approximately 0000 - After loop suction from RT double isolated, and trend didn't )

change in 2 pump down cycles, I informed LASS of this fact as I passed through '

MCR; then Gary Mosley/ Gary Setser/ Dan Andrew and I met in SS office. We concluded that Drywell RF leakage would not end up being less than 5 gpm, and that S/D was to be begun. I advised Setser to ensure Nuclear Engineer had good power reduction plan (we were in a xenon transient (minor), singlejoop, low power) ,

and that the crew had reviewed applicable procedures -- then proceed to Mode 3. ,

We discussed timing, I said I didn't thi;.x we had to hurry; plant was stable, need to j proceed carefully. Dan Andrew was to contact Outage Management (Bob  :

Gruenewald) and see if they had any specific needs with regards to time at a given plant condition. We all expected Bob to want us cool as early as possible, but we'd go at a comfortable pace. Also, I mentioned to Gary Setser that if he got to Mode 3 on his shift, to ensure we were ready for DW entry, since we wanted to know the condition of Drywell RF leakage /etc. while still hot.

Time: 0014 Calculated value for drywell floor drain flow rate steady at 10.5 gpm. Drywell RF leakaghnanual calculation.

~

L Elapsed timo: 7 hrs. 9 min.

Time: 0039  !

During the turnover process at 2330, the oncoming Shift Supervisor was advised that if leakage did not get below the administrative limit of 4.5 gpm within an hour or so, that further isolation of the RWCU system from the B recirculation loop would  ;

be possible. Waiting the additional hour would give the drywell floor drain flow rate I a chance to stabilize and still provide time to close the additional RWCU isolation 21

I 1

valves before the four hour Technical Specification time clock elapsed. CPS

.3301.01, REACTOR WATER CLEANUP, was reviewed for limitations with both recirculation loop suctions isolated as would be the case with further isolation.

Limitation 6.8.1 for maintaining bottom head drain flow between 63 and 200 gpm

, was reviewed and discussed. The crew attempted to further isolate the B RR loop from RWCU due to potential leak by on 1G33-F106 (closed at 2009) by closing 1G33-F100, Recirc Loop A Suction and 1G33-F102, RWCU Recire Suction Throttle.

Bottom head drain flow was ~189 gpm. No change in drywell floor drain flow rate l

was noted. It was decided that leak-by must be occurring from the recirculation

! loop suction and/or discharge valves and that a reactor shutdown would be appropriate.

l Elapsed time: 7 hrs. 25 min.

Time: 0055 Four hour Technical Specification Action 3.4.5.a elapsed for leakage greater than 5 gpm. Entered Action statement 3.4.5.c to be in Mode 3 within 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> (1255) and Mode 4 within 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> (1255 on 9/7/96).

Elapsed time: 7 hrs. 30 min.

Time: 0228 Completed pre-evolution brief on Reactor shutdown and commenced reducing reactor power from 55%.

Elspsed time: 9 hrs. 40 min.

Time: 0310

~

L.

Stopped the reactor power decrease at 38%.

Elapsed time: 9 hrs. 50 min.

l Time: 0320 l

Shutdown Turbine Driver Reactor Feed Pump 18.

22

l i

l Elapsed time: 11 hrs. 29 min.

l l

l Time: 0352 l

l Downshifted Reactor Recirculation Pump A to slow speed. Reactor power at 27%.

l l

Elapsed time: 11 hrs. 46 min.

Time: 0516

~

1 Commenced rod insertion to decrease reactor power from 27%. ]

l Elapsed time: 12 hrs.

Time: 0530 Completed rod insertion, reactor power 23%.

ElapseiFtime: 12 hrs. 20 min.

Time: 0550 g i Moisture Seperator Reheaters removed from service. Started up the Motor Driven Reactor Feed Pump. Shit:.>d the A Turbine Driven Feed Pump from 3 element  ;

automatic to single element automatic level control.  ;

Elapsed time: 12 hrs 42 min.

Time: 0612 l -

23

Report of a fire in the Service Building basement computer room. .

Elapsed time: 12 hrs. 45 min.

Time: 0615 Operator at fire scene determined no fire existed. There was light smoke. Six fire brigade members dressed out.

Elapsed time: 12 hrs. 53 min.

Time: 0623 _

Primary Containment Control, EOP 4402.01 entered due to Suppression Pool level greater than 19 ft. 5 in. Level increase was apparently from safety relief valve leakage, which had increased since the April 9,1996 reactor scram.

Elapsed time: 13 hrs.14 min.

Time:.0644 RHR B started in Suppression Pool Cooling to lower suppression pool level.

Elapsed time: 13 hrs. 23 min.

Time: 0653 g Exited Primary Containment Control, EOP 4402.01. Suppression pool level less than 19 ft. 5 in.

During this period, an attempt was made to place the MDRFP on the startup level controller unsuccessfully. A decision was made to allow the on-coming shift to determine if there was a problem with these controllers, and then perform the transfer. Engineering support was obtained, but troubleshooting revealed no anomolies.

1 24

i Elapsed time: 13 hrs. 48 min.

Time: 0718

'f Electrode Boiler 'A' supplying auxiliary steam. Main steam to auxiliary steam isolated.

l

)

i I

i Elapsed time: 14 hrs.

~~

Time: 0730 Shift Turnover Elapsed time: 15 hrs. 27 min.

Time: 0857 Placed the Motor Driven Feed Pump in single element automatic.

I Elapsed time: 15 hrs. 34 min.

Time: 0904 L  !

l

'A' Turbine Driven Feed Pump in manual on low speed stop. j I

Elapsed time: 15 hrs. 50 min.

Time: 0920 Commenced rod insertion for shutdown from 23% reactor power. i

Elapsed tima 16 hrs.15 min.

Time: 0945 While attempting to transfer 6900v bus 1B from the UAT to the RAT, the reserve feed breaker tripped momentarily, due to operator error, deenergizing the bus. The main feed breaker automatically reclosed reenergizing 6900v bus 18. All turbine-generator lift pumps and the 'A' GC Pump automatically started. Circulating water pump B, RWCU Pumps B and C, and the Auxiliary Boiler tripped. RPS Solenoid Inverter B trouble annunciator was also received. Entered 4200.01, LOSS OF AC '

OFF-NORMAL.

Elapsed time: 16 hrs. 35 min. _

Time: 1005 Completed transferring all AC buses to the RAT.

Elapsed time: 16 hrs. 36 min.

Time: 1006 Restarted Circulating Water Pump B.

Elapsed time: 16 hrs. 45 min.

~

L Time: 1015

'A~ Electrode Boiler will not restart. 'B' Electrode Boiler being started.

Elapsed time: 17 hrs.12 min.

Time: 1042 26 l

e e'

Performed pre-evolution brief on tripping the main turbine. .

Elapsed time: 17 hrs. 24 min.

Time: 1054

'B' Electrode Boiler in service.

Elapsed timo: 17 hrs. 30 min. ._

Time: 1100 Reactor power is 17%.

Elapsed time: 17 hrs. 36 min.

Time: 1106 Generator separated from the grid.

Elapsed time: 17 hrs. 38 min.

Time: 1108 Tripped the main turbine.

Elapsed time: 17 hrs. 55 min.

Time: 1125 27

While attempting to shift the Main Turbine steam seals to auxiliary steam,1GSO41 would not open remotely from the main control room. (MWR D61643)

Elapsed time: 18 hrs. 5 min.

Time: 1135 I

l Turned breaker off and manually opened 1GSO41.

1 Elapsed time: 18 hrs.10 min. _

Time: 1140 Main Turbine steam seals are on auxiliary steam. ,

i Elapsed time: 18 hrs. 24 min. i Time: 1154 1

Condu5ted pre-evolution brief on reactor scram. I Elapsed time: 18 hrs. 29 min. t Time: 1159 Reset main turbine to minimize cooldown following the scram.

Elapsed time: 18 hrs. 36 min.

i Time: 1206 -

28

Scrammed the reactor from 17% power as part of the normal shutdown per 3006.01, UNIT SHUTDOWN. Entered Mode 3.

Elapsed time: 18 hrs. 37 min.

Time: 1207 Entered EOP-1 RPV CONTROL due to a level 3 following the scram. Level recovered above level 3, exited RPV CONTROL.

Elapsed time: 19 hrs. _

Time: 1230 Unisolated RWCU to commence heatup of system to return to service.

Elapsed time: 19 hrs. 25 min.

Time:1255 l 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> to Mode 3 clock elapsed for Drywell floor drain leakage rate greater than 1 5 gpm?

Started RWCU Pump C.

~.

Elapsed time: 22 hrs.

Time: 1530 l

Shift Turnover Elapsed time: 23 hrs. 25 min.

29

l

%e Time: 1055 Isolated RCIC. Reactor pressure less than 150 psig.

Elapsed time: 25 hrs. 29 min.

Time: 1859 Mode 3 Checklist completed.

l l

l l

Elapsed time: 25 hrs. 50 min. __

Time: 1920 i

Opened 1G33-F100, Recirc Loop A Suction and 1G33-F102, RWCU Recire Suction I Throttle.

Elapsed time: 27 hrs. 34 min.

Time: 2104

- l Mode 4' Checklist completed.

Elapsed time: 27 hrs. 48 min, t Time: 2118 Started RHR Pump in Shutdown Cooling.

Elapsed time: 28 hrs. 20 min.

Time: 2150 30

}

M s ,

• i Notification of Unusual Event has been terminated. Drywell floor drain flow rate less than 5 gpm.

Elapsed time: 29 hrs. 46 min.

l I

Time: 2316 Entered l Aode 4 h

31

m

• ATTACHMENT 1

' GUIDELINE FOR RR B LOOP ISOLATION .

AND RF LEAKAGE EVALUATION This is to be used as a tool for dedsfons by Operations Management ,

and is not latended to conflict or ovenide CPS procedums or Tech Specs.

4

1. As Rx power is decreased to < 70% monitor.DW RF leakage for changes.

2. As the flow control valves are manipulated monitor DW RF leakage for changes. .

q .

2

3. When the B RR pump is shutdown be prepamd for seal leakage to rapid [/ increase and

, expedite loop isolailon if required.

t

4. Once the B RR pump is shutdown and the loop isolated the following may,be expected to i

happen:

l .

• DW RF leakage reinsins relatively unchanged. Then consider sotating RTfrom the RR loops to determine if 1G33f106 is lealdng by its seat and adding to the leakage into the RR loop. If this does not cause the DW RF leakage to decrease then CPS Management will decide what the next course of action will be.

• DW RF leakage decreases. CPS Management will decide if continued operation in single loop is. appropriate or not. .

• DW RF leakage increases. If the leakage exceeds 4.5 gpm then continue to shutdown the plant and go td Mode 4. If the leakage is less than 4.5 gpm then CPS Management.

will decide the next course of action. _

- . Gary J. Mosley Assistant Director-Plant Operations 9

Attachment B -

List of Condition Reports The following Condition Repora were generated as a result of the events associated with I the plant shutdown and assessment of the events:

CR-1-96-09-016, Unusual Event due to exceeding ITS leakage limit CR-1-96-09-022, Lack of procedural guidance to calculate DW RF flow up to 66 GPM CR-1-96-09-029, Windmilling ofMDRFP without lube oil l

CR-1-96-09-031, Manual operation of MOV CR-1-96-09-043, Low voltage on 4160 V safety buses CR-1-96-09-045, Missed sampling requirement for 15% power change in I hour CR-1-96-09-046, Valve initialed as being veri 6ed closed, but not installed , j CR-1-96-09-058, Single loop operation i CR-1-96-09-064, RR vibration monitors inadvertently de-tenninated j CR-1-96-09-069, PMSO-77 does not list IB33-R604 for CPS 9000.01D002 l (Not yet numbered as of 09/15/96), Procedure inadequacy- i (Not yet numbered as of 09/15/96), Failure to follep procedde CPS 3302.01 l (Not yet numbered as of 09/15/96), Exceeding wofking hours limits l

i

g. .

t L

e Attachment C -

i Penonnel Contacted '

The following personnel were contacted during the course of the assessment:

)

P. D. Yocum  !

J. Earl i R. Morgenstern l K. Cameron G. Mosley 1 R. Phares J. Hays W. Connell J. Naden i

B. Corley l R. Rippy

~

P. T. Young J. Kaineg ,

K. Sheffield J. Cunningham  !

R. Brixey R. McCubbin i J. Wells i D. Andrew J. Schottel .

)

M. Friehofer j D. Reeser i E. Rau

~ l

' D. Hodel M. Brotherton F. Perryman T. Olsen i M. Baetz i T. Doerscher t  !

J. Smith I G. Setser i

R. Giuliani l J. Spencer D. Korneman J. Wemlinger l

.k e

I

. - - . _ _ ,