Text

Loss of Inventory Event 870617-21
	ML20235U670
Person / Time
Site:	North Anna
Issue date:	08/07/1987
From:	; VIRGINIA POWER (VIRGINIA ELECTRIC & POWER CO.)
To:	;
Shared Package
ML20235U668	List: ML20235U665; ML20235U670;
References
	NUDOCS 8710140192
	Download: ML20235U670 (36)
	v • d • e

~ ~ ~

c ;-

- D. , ,

F.. 1 I

1 l

l :.

l' VIRGINIA ELECTRIC AND P'0WER COMPANY

' NORTH ANNA UNIT 1 IDSS OF INVENTORY EVENT JUNE 17 - 21, 1987 AUGUST 7, 1987 8710140192 87 g PDR ADOCK 0 PDR-'

S

___1_:____________________

i 1

l NORTH ANNA UNIT 1 LOSS OF INVENTORY EVENT j JUNE 17 - 21, 1987 ;

l t

TABLE OF CONTENTS 1

l CHAPTER PAGE EVENT

SUMMARY

I. 5 j II. EVENT DESCRIPTION 6 I

III. EVENT ANALYSIS 8 4 IV. MANAGEMENT ANALYSIS 10 V. OPERATOR ANALYSIS 12 VI. SHIFT TECHNICAL ADVISOR ANALYSIS 13 VII. SAFETY ANALYSIS 14 VIII. LESSONS LEARNED 16 IX. CORRECTIVE ACTIONS 17 i l

f Page 1 l ,-

i,

-3 NORTH ANNA UNIT 1

LOSS OF INVENTORY EVENT JUNE 17 - 21, 1987 LIST OF FIGURES FIGURE NUMBER DESCRIPTION 1 Reactor Vessel Level, Pressurizer Level, RCS . Temperature and RCS Pressure (Wide Range)'Versus Event Time (06/17/87 - 06/18/87)

' 2A - Reactor Vessel Level (Upper Range).

Versus Event Time (06/17/87 -

06/18/87) 2B' Reactor Vessel Level (Upper Range)

Versus Event Time (06/17/87 -

06/21/87) 3 Reactor Vessel Level, RCS Temperature and Pressurizes Level Versus Event Time (06/17/87 -

06/21/87) 4 Pressurizes Liquid Space Temperature Versus Event Time (06/18/87 - 06/21/87) 5 Reactor Vessel Level (Upper Range)

Versus Time 06/20/87 - 06/21/87)

[b I

i Page 2 I

M

'f

.i

, NORTH ANNA UNIT 1

' ~' ' '

LOSS OF INVENTORY EVENT JUNG 17 1987 LIST OF TABLES TABLE NUMBER DESCRIPTION 1 Mass Balance (06/17 to 06/21) 2 Mass Balance (06/17 to 06/20) 3 Mass Balance (06/21) i Page 3

{

tl L,

NORTH ANNA' UNIT 1 LOSS OF INVENTORY EVENT JUNE 17 - 21, 1987 LIST OF APPENDICES APPENDIX NUMBER DESCRIPTION 1 Sequence of Events 2 RCS Inventory Indications Page 4

I.- EVENT

SUMMARY

'During the. time frame between June 17, 1987 and June 21, 1987, approximately 17,000 gallons of RCS. inventory was removed from the Reactor Coolant System (RCS). The pathway for the loss was through the pump. backseat of the "A" Reactor Coolant Pump (RCP), up the shaft, and out through the seals into the containment sump. The pump had been decoupled from the motor for stator replacement. The inventory loss was not detected because the pressurizer.had been allowed to cool down and go into a vacuum. In this pressurizer condition, pressurizer level indication did not provide an adequate representation of RCS inventory. Indication was available via the Reactor

-Vessel Level. Indication System (RVLIS) portion of the Integrated' Core Cooling Monitoring (ICCM) System which had been installed during the refueling outage.

However, operator sensitivity to RVLIS was greatly reduced by.the fact that several work requests (WRs) were on the system, and it had been declared

-inoperable on June 19, 1987. A main control board annunciator "ICCM Failure"'

was locked in and the ICCM display diagnostics indicated " malfunction". (RVLIS was not required to be operable in this mode of operation). Volume Control

' Tank (VCT) and pressurizer level, however, were being trended.

The situation developed as a result of a management decision not to drain and vent the RCS prior to decoupling the "A" RCP. This decision was-based on schedule and ALARA considerations, minimizing RCP starts and stops during subsequent venting operations, as well as a desire to reduce the possibility of a challenge to RHR. However, potential procedural inadequacies for this mode of operation were not recognized.

It should be noted that during this entire event, inventory was sufficient to ensure operability of the RHR system. However, even though there was no direct safety concern with the inventory loss, it clearly was not planned or expected. Further evaluation to date indicates the need for improvement in several important areas. These areas include the management decision making process for entry into unusual operating conditions, operator training and procedure adequacy.

Page 5 1

II. EVENT DESCRIPTION On June 17, 1987, following a 59 day refueling outage, a problem developed !

in Mode 5 with the loop 'A' reactor coolant pump motor which required replace- l

-ment. To facilitate removal of the motor, Reactor Coolant System (RCS) system

. pressure and temperature vere reduced from approximately 195"F and 325 psig (bubble in pressurizer) to approximately 110*F and a pressurizer pressure initially at some . positive pressure determined by pressurizer metal

' temperature. Cooldown and venting of the pressurizer was conducted initially by filling and then venting through the pressurizer. power operated. relief valves (PORVs). This was, followed by a reduction of pressurizer level to approximately 80%, at which time the PORVs were closed because a' steam bubble still existed. Pressurizer level was then reduced to approximately 20% with no )

further venting or level change. The RCP motor was uncoupled approximately 10 I hours into the event at (0420 hours0.00486 days 0.117 hours 6.944444e-4 weeks 1.5981e-4 months on June 18, 1987), resulting in a small but !

expected leak path up the pump shaft of several gallons per minute. Seal

' injection flow to the '!A" reactor coolant pump was kept in service.

Pressurizer level was maintained at approximately 20% throughout this period. Once the charging pump was secured, makeup to the RCS was via a Volume Control Tank (VCT) " float". This method utilizes VCT pressure to inject water into the RCS during plant conditions when RCS pressure is less than VCT pressure. . Letdown was established from the Residual Heat Removal (RHR) system back into the normal RCS letdown path. RCS inventory was being maintained by adjunting makeup and letdown flow rates to balance losses and maintain pressurizer level at 20%.

At approximately 70 hours8.101852e-4 days 0.0194 hours 1.157407e-4 weeks 2.6635e-5 months into the event (2200 hours0.0255 days 0.611 hours 0.00364 weeks 8.371e-4 months on June 20, 1987), in response to a. request by maintenance personnel to reduce the pump shaft leakage, operators decided to lower pressurizer level to 5 to 8% to decrease static head on the seal. This evolution did not significantly reduce RCS leakage as monitored on the. Containment Sump Pump Recorder. In addition, the operator noticed a significant drop in Reactor Vessel Level Indication System (RVLIS) indication but did not evaluate the relationship of this indication to reactor vessel inventory because the RVLIS portion of the Integrated Core Cooling Monitoring (ICCM) System had previously been declared inoperable.

Rather than leaving pressurizer level in the 5 to 8% range, it was decided to raise the level back to its normal operating range but still reduce RCS leakage by reducing pressure. Pressurizer pressure was to be reduced by raising level somewhat higher than normal (i.e. approximately 50% level),

venting the pressurizer by cycling a PORV, and then lowering pressurizer level back -to its normal operating range (approximately 20% level). Pressurizer level was then raised, and a PORV was cycled. Upon venting, pressurizer level dropped and Pressurizer Relief Tank (PRT) pressure decreased to O psig (lowest reading on meter). At this time, the operator realized that the pressurizer was already in a vacuum (i.e., pressurizer pressure is subatmospherie) and that there may be a void in the reactor vessel head area based on RVLIS indication Page 6 1

L_ __ _ __ _ _

l l

l J

4 and unusual pressurizer level indication and behavior. The RCS makeup, already in progress to raise level to approximately 50%, was continued and preparations to vent the reactor vessel head were made. Approximately 17,005 gallons were required to refill the RCS. By approximately 78 hours9.027778e-4 days 0.0217 hours 1.289683e-4 weeks 2.9679e-5 months into the event (0730 hours0.00845 days 0.203 hours 0.00121 weeks 2.77765e-4 months on June 21, 1987), little or no air was being vented from the reactor vessel head, pressurizer level was stable in the 50 to 60% range, and the pressurizer was vented to the PRT. This condition was maintained until repairs to the "A" RCP were completed.

The "A" RCP shaft we.s coupled at 0612 hours0.00708 days 0.17 hours 0.00101 weeks 2.32866e-4 months on June 27, 1987, which stopped the inventory Jess through the seals. At 0626 hours0.00725 days 0.174 hours 0.00104 weeks 2.38193e-4 months on June 27, 1987, the "A" RCP ~ repair was completed, the pressurizer was filled in preparation of going water solid. The pressurizer, once water solid, was pressurized to approximately 300 psig. At 0927 hours0.0107 days 0.258 hours 0.00153 weeks 3.527235e-4 months on June 27, 1987, a reactor coolant pump was run for 15-20 seconds to sweep the steam generators of any remaining voids.

Then the RCS was depressurized, and the reactor vessel head was vented. By 1158 hours0.0134 days 0.322 hours 0.00191 weeks 4.40619e-4 months on June 27, 1987, the RCS was refilled by an addition of 865 gallons to a water solid condition and repressurized to approximately 315 psig. A subsequent venting and refill indicated that the RCS was essentially water.

solid.

I Page 7

III. EVENT ANALYSIS For analysis purposes, the shutdown of "C" RCP at 1833 hours0.0212 days 0.509 hours 0.00303 weeks 6.974565e-4 months on June 17, 1987 was chosen as the starting point for this event (or zero hours into the event) .since this plant condition is well documented and' occurred prior to the loss of RCS inventory. A detailed sequence of events is provided in Appendix

1. . RCS inventory indications available to the operators are provided in Appendix 2.

The pressurizer was initially filled and vented via the'PORVs in order to cool it down. . (Refer to ~ Figure I which shows pressurizer level (Hot Cal) during theLfirst 13 hours1.50463e-4 days 0.00361 hours 2.149471e-5 weeks 4.9465e-6 months of the event.) Note the initial increase to 100%

level. during the first 3 1/2 hours and the decrease to 20% level over the next 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . The FORV was shut at approximately 80% (Cold Cal) level', shortly after the level decrease was initiated because a steam bubble still existed in the. pressurizer. Also shown on this figure are RCS pressure and RCS temperature. Figure 2A shows a decrease in reactor vessel level from above 100% to approximately 94% starting approximately 4 1/2 hours into. the event.

This corresponds with the completion of the RCS depressurization following the cooldown. .The relatively rapid decrease in reactor vessel level continues until approximately 11 1/2 hours into the event which correlates well with the completion of pressurizer level reduction to approximately 20%. F m Figure 1 and 2, it can be seen that reactor vessel level dropped slig'+'y prior to completely depressurizing the RCS.

During these initial twelve (12) hours, the VCT was purged with nitrogen and placed on " float". Pressurizer vapor space temperature was reported to be between 220*F.and 240*F and pressurizer response indicated the presence of a steam bubble. The "A" RCF shaft was uncoupled at approximately 10 hours1.157407e-4 days 0.00278 hours 1.653439e-5 weeks 3.805e-6 months into the event.

During the next couple of days, or 64 hours7.407407e-4 days 0.0178 hours 1.058201e-4 weeks 2.4352e-5 months into the event (0600 hours0.00694 days 0.167 hours 9.920635e-4 weeks 2.283e-4 months , June 18, 1987 to 2200 hours0.0255 days 0.611 hours 0.00364 weeks 8.371e-4 months , June 20, 1987), pressurizer level was maintained constant at approximately 15% (Hot Cal) or approximately 20% (Cold Cal), and seal' injection to the "A" RCP was maintained at about 2 gpm. Figure 3 shows pressurizer level, RCS temperature and reactor vessel upper range level over the entire time of the event. Note the slow but steady decrease in reactor vessel level as RCS inventory is lost out through the "A" RCP seals. (See Figure 2B). The flow path for the leakage was through the RCP pump shaft backseat (located just above the pump thermal barrier) and out through the pump shaft. seals. Seal injection flow entered the pump at the thermal barrier just below the backseat. The backseat design rating is 1 gpm leakage at 15 psig (RCS pressure). At higher RCS pressures, the leakage is greater. j l

Figure 4 shows pressurizer liquid space temperature during the event.

Initially, the rapid drop in temperature is the result of shutting off the heaters and filling the pressurizer to approximately 65% level. Then, as level was decreased, temperature increased due to latent heating. The next large temperature decrease which occurred 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months into the event (2100 hours0.0243 days 0.583 hours 0.00347 weeks 7.9905e-4 months on June 1

Page 8

Y 17, 1987) was the result of filling the pressurizer to 100% level. Liquid temperature increased for the next 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months (until 0600 hours0.00694 days 0.167 hours 9.920635e-4 weeks 2.283e-4 months on June 18, 1987) as the level was lowered to 20%. .Then a steady decreasing liquid temperature, caused by ambient losses, occurred until 73 hours8.449074e-4 days 0.0203 hours 1.207011e-4 weeks 2.77765e-5 months into the event (0000 hours0 days 0 hours 0 weeks 0 months on June 21, 1987). The- subsequent rapid temperature drop is due to the makeup initiated to refill the RCS. Pressurizer vapor space temperature data is not available, but a steam bubble at a positive pressure did exist in the pressur-

- izer during the initial portions of this event.

As the vapor space temperature decreased, the sceam bubble cooled and pressurizer pressure also decreased, since the pressurizer (and RCS) was not vented. Ultimately, the pressurizer pressure dropped below atmospheric pressure. The effect of' this pressurizer pressure reduction was to prevent normal filling of the RCS loops from pressurizer inventory. Consequently, net inventory was being lost from the RCS which resulted in voiding the steam generator tubes and reactor vessel head. Pressurizer level instrumentation did not indicate this net loss to the RCS because the level in the pressurizer was not changing. As'a result, a net loss of approximately-17,005 gallons occurred from the RCS without operator recognition because in this condition (vacuum in pressurizer), pressurizer level is no longer a valid indication for RCS

- inventory.

The event was terminated when operators received indications that the j

' pressurizer was in a vacuum (i.e., pressurizer pressure was less than atmospheric pressure). These indications resulted from efforts to reduce RCP seal Icakage which occurred from 70 to 73 hours8.449074e-4 days 0.0203 hours 1.207011e-4 weeks 2.77765e-5 months into the event (2200 hours0.0255 days 0.611 hours 0.00364 weeks 8.371e-4 months on June 20, 1987 to 0130 hours0.0015 days 0.0361 hours 2.149471e-4 weeks 4.9465e-5 months on June 21, 1987). These efforts consisted initially of lowering pressurizer level to decrease static head. After a shift turnover at approximately 71 hours8.217593e-4 days 0.0197 hours 1.173942e-4 weeks 2.70155e-5 months into the event (0000 on June 21, 1987), the on duty SR0s became concerned about maintaining such a low pressurizer level.

They decided that they could accomplish the same objective of reducing leakage by decreasing pressurizer pressure rather than by reducing static head (i.e.,

pressurizer level). Therefore, they decided to raise pressurizer level to approximately 50%, vent the pressurizer and then lower pressurizer level back to its normal 20% level. However, when the pressurizer was vented at approximately 73 hours8.449074e-4 days 0.0203 hours 1.207011e-4 weeks 2.77765e-5 months into the event (0130 hours0.0015 days 0.0361 hours 2.149471e-4 weeks 4.9465e-5 months on June 21, 1987), it was immediately noticed that both pressurizer level and PRT pressure dropped. This indicated that the pressurizer pressure was already subatmospheric. The RVLIS recorder trace was also checked. The operators understood the implications of the pressurizer being in a vacuum on pressurizer level indication accuracy, and correctly concluded that they probably had a void in the reactor vessel head.

They continued the already in progress RCS makeup and also vented both the reactor vessel and the pressurizer until the RCS was refilled. The RCS was refilled and the event terminated at approximately 77 hours8.912037e-4 days 0.0214 hours 1.273148e-4 weeks 2.92985e-5 months from the initiation of the event (0600 hours0.00694 days 0.167 hours 9.920635e-4 weeks 2.283e-4 months on June 21, 1987). Figure 5 shows the RVLIS upper range recorder trace from approximately 70 hours8.101852e-4 days 0.0194 hours 1.157407e-4 weeks 2.6635e-5 months into the event (2200 hours0.0255 days 0.611 hours 0.00364 weeks 8.371e-4 months on June 20,1987), through event termination.

i f.

Page 9 ,

L__________________ __ i

. IV. MANAGEMENT ANALYSIS The decision to ' perform the "A" RCP stator replacement without draining down to approximately 40 inches above centerline of the RCS loops and venting the RCS was ultimately made by station management. There was a desire on the part of station management not to drain down and vent. This desire was based on schedule' considerations as well as the desire to minimize 1) radiation exposure from installing and removing vent path piping and spool pieces 2) challenges to RHR at drain down conditions, and 3) the number of RCP starts that would be required to fill and vent the RCS prior to starting up. j A management decision was also made to maintain seal injection flow to the "A" RCP after decoupling. The basis for this decision was.to ensure that unfiltered RCS did not leak through the seals. However, management was not fully aware of the leakage path for a decoupled RCP or the static head' requirements for minimizing leakage through the seals. Management was not entirely cognizant of the exact ' amount of leakage through the seals being experienced during the entire event. However, management did understand, as a

. basis for decision making, that a small amount of leakage was pussible.

The Operations management concurrence with the decision not to drain was based on a review of the controlling operating procedure (1-0P-3.4) which had provisions for being. in Mode 5 with a positive pressure steam bubble in the pressurizer. However.. subsequent direction to the operating shift was to maintain pressurizer level but not a positive pressure steam bubble. The adequacy of 1-0P-3.4 for this condition was not specifically evaluated.

Once the significance of the event was understood by the on shift SR0s, they contacted the Superintendent Operations at approximately 73 hours8.449074e-4 days 0.0203 hours 1.207011e-4 weeks 2.77765e-5 months into the event (approximately 0130 hours0.0015 days 0.0361 hours 2.149471e-4 weeks 4.9465e-5 months on June 21, 1987). The Superintendent Operations consulted with the shift SR0s on refilling and venting the RCS as well as reviewing the need for deportability to the NRC under 10CFR50.72. The Superintendent Operations also consulted with the Supervisor-Nuclear Safety Engineering on deportability. The event ras considered to be nonreportable but a station Deviation Report was initiated.

The Assistant Station Manager (Nuclear Safety and Licensing) was subsequently contacted and briefed between 0900 and 1030 hours0.0119 days 0.286 hours 0.0017 weeks 3.91915e-4 months on June 21, 1987 by the on duty STA, on duty SRO and the Superintendent Operations. The Assistant Station Manager considered the event to be potentially both significant and reportable and contacted the Station Manager at approximately 1030 hours0.0119 days 0.286 hours 0.0017 weeks 3.91915e-4 months on June 21, 1987. T!.e Station Manager convened a review meeting at North Anna Power Station'at approximately 1200 hours0.0139 days 0.333 hours 0.00198 weeks 4.566e-4 months on June 21, 1987 and also notified the NRC Senior Resident Inspector. A briefing meeting was held at the station with the NRC Resident at approximately 1430 hours0.0166 days 0.397 hours 0.00236 weeks 5.44115e-4 months on June 21, 1987.

The event was determined to be nonreportable under both 10CFR50.72 and 10CFR50.73. However, station management has decided to submit a volunteer License Event Report to the NRC and to make an INP0 Nehwork entry describing Page 10 l

6

\

t 5

the event. Management'also initiated a detailed' engineering review and Human Performance Evaluation System (HPES). evaluation of this event. Finally a " Case ;

Study" will be prepared for dissemination to other nuclear utilities.

l 1

l i

l f

f.

r Page 11

1

, 1 ,

yl. r -

A ":.

, ,,' :V.-. OPERATOR ANALYSIS.

Each;L.of the five operating shifts were'on duty.for at least one shift of j eight hours' duration during;this event. 1Each shift maintained RCS inventory.by! j

' balancing makeup: and letdown to : maintain'a. constant pressurizer level. No

~

q

licensed operator. independently. evaluated pressurizer . pressure ~ conditions or. -

4

_' verified the RCS inventory balance through other indications. The' leakage rate :;

S from. the; RCP, seals .vas being determined periodically. by visual1 observation from s unlicensed operators in the containment.

Several licensed'. operators. noticed the decrease in reactor vessel 11evel as f jindicated'by RVLIS. iHowever ino actions were 't'aken because 1) pressurizer

.* 1evel andLthe makeup:and. letdown' balance were normal,'2).the-ICCM trouble alarm

",- 1had been in prior to the' trip of'"A" RCP and during this entire event, 3) RVLIS

had failed its operability. Periodic; Test (1-PT-44.7) on June ~ 19, 1987 and been declared inoperable, 4) previous experience with RVLIS at low RCS pressure conditions had indicated"that RVLIS was unreliable. (susceptible to gas binding Lof the. capillary. tubing between 'the RCS and the pressure bellows).

.. joperator assessment'and corrective actions taken during this l event when.

' additional ' indications - became ~ - available were performed- correctly'and in a timely manner.

l I

t Page 12

'r'

-[____________ _ _ . _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ . . _ . _ __

\

VI. SHIFT TECHNICAL ADVISORS ANALYSIS STAS are not required for an unit when it is in either Mode 5 or 6. As a result,. STA attention was inadequate. STAS had not been specifically trained or required to perform an RCS inventory mass balance in Modes 5 or 6.

i l

l Page 13 I

i J

p b

VII. SAFETY ANALYSIS In order to determine the_ impact on safety, if any, a RCS inventory 1 mass balance was performed to determine if RHR would have been lost, and a f safety analysis (10CFR50.59) was performed to determine whether operation with ~the pressurizer in a vacuum (i.e., pressurizer pressure less than atmospheric pressure) created an unreviewed safety question.

The'RCS inventory mass balance was set up between an equilibrium plant condition prior. to the event (1200 hours0.0139 days 0.333 hours 0.00198 weeks 4.566e-4 months on June. 17, 1987) and an equilibrium plant.condit' ion after the event (0623 hours0.00721 days 0.173 hours 0.00103 weeks 2.370515e-4 months on June 21, 1987).

At the equilibrium condition prior to the event, no voids existed in the reactor vessel or steam generator and two RCPs were running. At the equilibrium plant condition after the event, the RCS had been vented and the pressurizer PORVs were open with RVLIS upper range indicating above 100%.

Between these two conditions. RCS inventory was both added and removed.

(See Table 1.) In order to determine the net loss during this period, the above. mass balance was divided into two subintervals. The first interval (1200 hours0.0139 days 0.333 hours 0.00198 weeks 4.566e-4 months on June 17, 1987 to 2400 hours0.0278 days 0.667 hours 0.00397 weeks 9.132e-4 months on June 20, 1987) covers the timeframe when losses to the 'RCS exceeded makeup. (See Table 2.) The second interval-(0000 hours0 days 0 hours 0 weeks 0 months on June 21, 1987 to 0623 hours0.00721 days 0.173 hours 0.00103 weeks 2.370515e-4 months on June 21, 1987) covers the timeframe when makeup to the RCS exceeded losses. (See Table 3.)

The maximum loss of RCS inventory (from Table 3) was approximately 17,005 gallons.

With a RVLIS upper range indication of 79% (lowest indication recorded), it can be calculated that 638.5 cubic feet of the reactor vessel head had voided. However, a net loss of 17,005 gallons translates to a net void of approximately 2447 cubic feet. Since only 638.5 cubic feet are accounted for in the reactor vessel, it is assumed that the remaining voiding (approximately 1608.5 cubic feet) occurred in the steam generators.

The- 1608.5 cubic feet would translate to approximately 73% of the steam generator volume (assuming 7% of the tubes. in each of the three steam generators are plugged.) l It is known that the makeup on June 21, 1987, to refill the RCS may not have completely filled some of the RCS high point areas because a RCP was -)

I not started to conduct a fill and vent operction. The fill and vent operation was subsequently conducted between 2200 hours0.0255 days 0.611 hours 0.00364 weeks 8.371e-4 months on June 26, 1987 and 1158 hours0.0134 days 0.322 hours 0.00191 weeks 4.40619e-4 months on June 27, 1987, and additional 115.6 cubic feet of void was filled. Assuming this additional 115.6 cubic feet of voiding existed in the steam generators prior to the June 21, 1987, the maximum net inventory loss would have resulted in 78% (rather than 73%) of the steam generator volume j being voided.

j l

l Page 14 j I

- _ _ _ _ _ _ _ _ i

i<

Based on RVLIS indication,'at the lowest. point there was still at least 5 feet of water above the centerline of the RCS loop. (RHR takes suction of tne hot leg lop pipe below centerline of the RCS centerline). Also, there was'at least 8 feet of water above the top of the reactor core. As a result, there was no impact on RHR e.,d core cooling during this event Operation with the pressurizer pressure less than atmospheric pressure l

.was also assessed from a safety standpoint, and it was' concluded that operation in this mode did not constitute an unreviewed safety gr Stion.

It should be noted that had the event continued undetected past June 21, 1987, the pressurizer would have eventually " burped" (i.e. vented when RCS 3evel in the loop dropped below where the press -rizer surge line enters the top of the _ loop pipe) and level would have dropped in the pressurizer.

Operator action (as actually occurred on June 21, 1987) would have been to restore level in the pressurizer.

l l

l Page 15 ;

i l

VIII. LESSONS LEARNED The root cause of'the event'was'the failure of Operations management to. '

adequately evaluate. and control plant. . conditions. The- . controlling

, procedures, both for Operations and Maintenance, used during this event- ,

,specified initial conditions'which were subsequently changed or never met.

Specifically, 1-0P-3.4, required ~a positive,-pressure steam bubble to be -

maintained in the pressurizer, and MMP-C-RC-28 required either the loop to be drained below the pump flange- (with loop l isolation valves closed and disc- .

. pressurization system in service, o the pressure in the RCS on the pump be-

~

r maintained less than 15 psig (30 foot static head) when. lowering the pump shaft. The requirement for a positive pressure steam bubble condition was subsequently eliminated. .The 15 psig (30 foot static head) requirement was

not met at 20% pressurizer level or with seal injection from the VCT which was ' being maintained between115 and 30 psig (i.e. . the RCS should-have been-drained'to near, centerline and vented.with seal injection secured).

A contributing factor to this event was that the operator's. lacked.

complete understanding.of.when pressurizer level is a good-indication of RCS inventory. 'As a result, operators failed to fully' utilize all available indications.;

Page 16

J p

y.

IX. CORRECTIVE ACTIONS 1

iCorrective ' actions were. identified to address procedural, training and RCS inventory. indication inadequacies during Mode 5 operation. A summary

, breakdownLof corrective action is presented below.

- Training L4 censed operators and licensed operator trainees and shift. technical

' advisors will receive classroom and simulator training on all phases of Mode.

5' operations '(drain down, steam bubble and water solid conditions) and on Mode _5 relevant operating procedures.

Theory instructions will be expanded to specifically address the relationships between Control Room indications and RCS inventory in both normal operating and abnormal transient conditions.

RVLIS (ICCM) will be installed on the simulator and integrated into both normal and transient simulator exercises.

Operators will be . instructed on primary maintenance evolutions which open the-RCS or result in leakage from the RCS.

Major design changes installed as a result of the TMI accident (e.g.

RVLIS/ICCM, reactor . head vents, pressurizer PORV/ safety valve open

. indicaticus, etc.) that assure RCS inventory is adequately maintained will .

be periodically reviewed during Operator Continuous Training.

Procedures Operating procedures used' in Mode 5 will be reviewed and revised as required. The' review will include OP-1.1. 1.2, 3.3, 3.4, 3.5, 5.6, and 11.3. OP-5.6 will be expanded to correlate RVLIS percent full values with actual height above RHR suction, top of the core, and the bottom of the core. The validity of RVLIS ranges (Full, Upper and Dynamic) will be specifically stated for all modes of operation and all combinations of running and nonrunning RCPs.

A simplified Operator Curve will be developed (if possible) to

. correlate Hot Calibrated and Cold Calibrated pressurizer level indications.

Maintenance (mechanical and electrical) procedures used for RCP maintenance or repair will be reviewed for adequacy.

Page 17

__.________________m

i i j Indications of RCS Inventory i

A study was performed and recommendations made on .RCS inventory indications. The recommendations will be reviewed by Operations and Technical Services, and final recommendations will be provided to management.

Operations Operations will develop a system to ensure that initial maintenance procedural conditions are maintained (unless properly deviated) as long as the procedure is in effect.

STAS will be required to independently verify adequate RCS inventory in Modes 5 and 6.

Management Station personnel will be reminded, via a memo from the Station Manager, of the importance of evaluating the procedures they use with respect to plant conditions which er 'st during, or could result from, the performance of the procedures. Furthermore, this memo will reemphasize that employee concerns over the applicability of procedures to plant conditions and/or assigned tasks should be brought to the attention of station management promptly.

An Operations Directive will be issued to operations personnel emphasizing the importance of considering non-routine evolutions that are not specifically described in procedures or the UFSAR as requiring a safety evaluation in accordance with 10 CFR 50.59 and/or being considered as a special test. Special tests reqire a safety evaluation prior to performing the test.

Page 18

TABLE 1 MASS BALANCE (06/17 to 06/21)

MAKEUP 06/17 8,941 GAL 06/18 to 06/20 2,754 GAL 06/21 19,747 CAL

, TDTAL ADDITION 31,442 GAL 1

IRSSES l

l DIVERTED TO GAS STRIPPER 13,940 CAL SEAL LEAKAGE TO SUMP 12.228 GAL i

SHRINK EFFECT 1,574 GAL IDENTIFIED LEAKAGE 2,376 GAL PZR LEVEL CORRECTION 1,480 CAL TOTAL LOSSES 31,598 GAL O

.s

c

=),,

~

h!

, TABLE 2-MASS BALANCE (06/17 to 06/20)

MAKEUP 06/17 8,941-GAL

~ 06/18 to 06/20 2,754 GAL TOTAL' ADDITION 11,695 GAL

IASSES DIVERTED TO GAS STRIPPEk 13,940 CAL SEAL LEAKAGE TO SUMP 11.055 GAL SHRINK EFFECT 1,574 GAL

. IDENTIFIED LEAKAGE 2,131 GAL TOTAL IASSES 28,700 GAL NET LOSS 17,005 CAL l

l-i p

1.

TABLE 3 HASS BALANCE' (06/21)

MAKEUP 19,747 GAL IASSES SEAL LEAKAGE 1,173 GAL' IDENTIFIED LEAKAGE 245 GAL PZR LEVEL CORRECTION .1,480 CAL.

i MTAL LOSSES 2,898 CAL

-NET MAKEUP 16,849 GAL l

j 1

l

g y

g g

g

~ 1 g

T I

l v

e g e

g N L ;

U R Z

P g g

- g

- g N g O

I 3

3 T l l

A 7 _ _ l T8 / l S8 l

I 1

1

/ l ER6 . l RE- 1 l

UW8/

G 7

l 1

I O7 1 e

i i

F P/

6 r u

s I

A I

s e I r

N P I N S I A

C R

' I I

I H p I

T m e

I I

R T s

I O C R

I I

N .

S I

L V I I

R I j

i l

- - - ~ - - - - , - -

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 2 0 8 6 4 2 0 0 0 3 3 3 8 6 4 2 0 8 4 2 2 2 2 2 2 1 1 1 1 1

,,5

l

~

3 1

' , 2 1

1 1

E ,

p M

- 0 I g T p1 i

s ,

v ,

,9

~ L8 7

/

E3 i

1 i 8 n V/ w o

AE6 -

i d

t u

2L 7 , 7 h s

E8 / Ps E G7 i

C 1

i R

U R N/ 6 i 6 i

C A

i e

G R1 i c

I i 5 in T s F RI N U

i i

s r

E S i

u o

PP i 4 H PN A i

i U i i 3 S

I I

i L e V i 2 R i i

e i

1 i

i I

O 5 4 3 2 1 0 9 8 7 6 0 0 0 0 5 4 3 2 1 0 1

0 0 9 9 9 9 9 9 9 9 9 9 1 1 1 1 1

[} 3 g$g l

5 l

i i

a i

i

- i i

i E i i

M I

i T

i i '

7 8 i s/

v1 i'

8 2

/

i 2E6 i G78 i

0 E N/ 7 i i 5 R

U A1 / i R6 i

G i 0

I R 1 i 4 F ET i PN I

i P SU i UP A i 3 i

0 S N I i i

L V

i 0

i 2 R i i

i i 0 1 i

i i

- - - - - - - - - 0 2 0 8 6 4 0 0 9 2 0 8 6 4 2 9 9 9 9 8 0 8 6 4 1 1 8 8 8 8 7 7 7 C$UU5E e$o:n IDO3 Ujm m u-

i i

~

i i

i 1 i T

i i

I N ,

U i i

- i N

i i

O I

i i

T i A 7 i

3 T8 i

/

S1 2 i i

ER6 /

w ~.

i RE-i i

UW8/

G 7 i I O7 1 i

FP/

A 6

A^ _

i i

N s i

N I

L v

A tii a

/ l i

e H v e i T I R

i R p Z P

i O T m

e i

N S C

R

- i i

~

- - - - - - - - - ~ -

0 SO 706 05 04 30 20 9

1 I 1 1 1 1 1 0 0 0 1

1 0 9 0 80 70 60 504 03 02 0 0 0 1

1 1 yS u"

,l

0 0

1 P

M E

- 0

_X ' 8 T-7 n

w o

E /8 d

- t C12 u

h S

A/

0

' 6 P

P6 C

4 -

S-R E -

- O' R

D 87 U -

e G.

I

' 0 c 4 n l

F U/8 -

i S

Q1/

s

- r I - -

u L6 o

H

- - 0 R e r

'2 Z t u

a -

=-

P r - -

e 4 p -

m -

T e -

o l

. l 0 0 0 0 o 0 0 0 0 5 4 3 2 3

m...

. ~b fwtl l l'

W E

M I ,

T .

s v 7 L8

/ ,

E1 2 ,

V/ , 0 5E6 - , 3 L

E E8 7 ,

R /

G0 U 2 ,

G N/ 6 ,

I A ,

FR1 T RNU I

i 0 E i PP SA i PN i U i S

I i L

V i

R i 0 i 3 0

, 2 i

i 2 0 8 6 4 2 0 8 6 0 0 4 2 0 8 6 4 2 0 9 9 9 9 9 8 8 1 1 8 B 8 7 7 7 7 7 n$ - $C$ e a $ $ > a:

- [aw

c.,. - - -

l i

TABLE 1- l MASS BALANCE j (06/17 to 06/21) J MAKEUP l l

06/17 .

8,941 GAL l 06/18 to 06/20 2.754 CAL 06/21 19,747 CAL TOTAL ADDITION 31,442 CAL LOSSES DIVERTED TO CAS STRIPPER 13,940 CAL SEAL LEAKAGE TO SUMP 12.228 CAL SHRINK EFFECT 1,574 GAL IDENTIFIED LEAKAGE 2,376 CAL PZR LEVEL CORRECTION 1,480 GAL TOTAL LOSSES 31,598 CAL l

Page 25

TABLE 2 MASS BALANCE (06/17 to 06/20)

MAKEUP 4.

06/17- 8,941 CAL. !

06/18 to 06/20 2.754 GAL TOTAL ADDITION 11,695 GAL LOSSES DIVERTED TO GAS STRIPPER 13,940 GAL SEAL trArACE TO SUMP 11.055 CAL SHRINK EFFECT 1,574 CAL IDENTIFIED LEAKAGE 2,131 GAL TOTAL LOSSES 28,700 GAL NET LOSS 17,005 GAL Page 26

I TABLE 3 MASS BALANCE-(06/21) i MAKEUP 19,747 CAL -j i

l j

1 LOSSES l

SEAL LEAKAGE. I,173 GAL IDENTIFIED LEAKAGE 245 GAL PZR LEVEL CORRECTION 1,480 CAL TOTAL LOSSES 2,898 GAL NET MAKEUP 16,849 CAL Page 27 )

_ _ _ _ _ _ _ . _ _ _ _ . . _ _ _ 1

ro L L L L L u C L L L L u r r oS p / p o eh f O 0 O O 0 m O O O 0 0 m c t t n R R R R R o R R R R R o e ni I C C S S. C C S S S C C C R I w

~

d n

u o

4 r d g ) n e l a d r a o o c )

M t l o t a S r m d o c T . o ; e g h

(

E a f g f d m d H p g n i

s o n u

u u

i s %

l o

1 T o E ~

o p n c p 0 c X

4 i o r a 0 (

IF % 5 t 0 i g v 0 1 O 5 2 a 2 t 2 %

1 r 3 a s t 3 n 0 ME FC =

1 a

p e i c i n e e h a 0 1

FR I s e r d A m r t AE 2 l4 r u n 1 n u o U N a0 p s i - i s r t Q u3 s P a s e E , q1 n e ,

- t e t l S % e . i r A C n r a a 5 0 P 1 R o P e u 8 e p - c r q t s u S P t S g e g = al t C - r p g C k ra a A R C o u n C R l l

/ 2 u e 1 R p t i l e e c H kq h - ; - e a k - ; v v c ae P FF 1 r e a P F e e e e S - *

h e -

l l 4 c ld C C 50 f s r C 2 a e R R 92 o n S b R 8 r r 6 p di - 1 4 a C - 1 e e s ef d 1 p i R d 1 z z n

= ii e -

pp i c e p i i e rp s f t c d mm r i d c d e

m e

r u uo 2 a i n n e ee T r e n p H g t e e t TT t p e T s s nd m r o c p m p s sV S T ei m a SR t e o m o S e eR C C d n o t CZ u l t o t C r rO R V I u C S RP A E S C S R P PP t

ne 0 0 em O 0 5 5 vi '

ET 5 5 l

ae 0 0 0 8 8 2 3 9 9 1 3 3 0 0 um 3 2 5

0 7

5 1

1 2

4 2

5 2

0 3

2 4

3 8

0 0

ti 1 cT 0 0 0 1 1 1 1 1 1 1 1 1 0 0 A

7 7

_ e 8 8

/

t /

_ a 7 8

_ D 1

/

1

_ /

_ 6 6

_ E$-

(

a I

i:j

n t o p g i e w e o t c eO D r w L ar i R e g g g g eO

- mu vS .. d o o o o iR O 3 g g ro r t r L L L L vC r R C

o o o oS eh n o eh L L L f

n t t i c 0 O 0 O t t

/

O O ni~ a e R R R R ni R O O I I w M R C C C R R R C I w S C C C

)

3 e -

d y e d e o l o r a g M e t u l % i ne t o 9 v acl a s -

c 1 rni l s

( t et m e e e S % ) a rrn i x v r . l e

T %0 o py N 08 l

a l eeu pf r o l r i

h E 2 c T a 1T y f pf e p % R w E uil p 8 Z Rm X

yl t T Db a /i s s l e' o C S a (

- P r n V IF et 5 ; p o R DO h

(

V I c g e d t a L .i y l O N am A d V k o s eo l P EE mi % e l / r t i st a FC i x 1

- 5 R "B p c a a a g FB h p c dl AE xo or P

1 i

s n" a i m l e

r ep ru 3 ee l v U rp C r l od 5 r p v o ce ce Q pp R e b nt e t ek 7 yl E pa a t ao 6 l d a c S a

- t a n n o S m 4

R 1 t e - t R C 7

t e s m "A . )

l 9 d Z oa) l g r e e n Hd Z

P R ed 1 nP t n g t" S. o Oe o ve a V a i ,

asT i Id g eu f e lR c l l B t r( t Ld n t .

l n o ,s eO p e 1 se a a i i Ni vPd u v - t% r s r p0 Rt p a e l o e P n nt3 mO e u5 Z n u h r ld o c l - o oi t

n e P o e t ec n a 2 w k y C k i o R s( u r Hi Ct i mn e rH o 'a l ;

a wt Z o Pll e) - c t sa cnh c

n g

s l me t V

R .

m d zl 1 e aat o T g C e e i a 0n d d a Od ) R n v j r c 0o e em d

e e t

e rc d n dt r 5l t ci P e s

S C

Pi c u ei e e y l a nx u

.l l sd r rl t H d a d a R d n u) p sl ul eea eg i eo ee ( ei dlR m eo ca t ve S d t mr mp l r st eaZ o rc ee ner C d0 i

n op cc ye T s n R cP C P( S s El g R A5 I C a Cd C ao V C c t 0 0 .

ne 0 0 0 5 0 5 0 5 5 5 5 5 5 5 0 em 7 7 0 5 7 5 7 0 7 2 0 vi 1 1 1 3 ET 7 1 9 1 1 3 4

3 8

4 9 2 3 2 8 4 6 7 7 7 7 7 l

ae 5 0 0 0 5 um ti 1 o0 2 0 0 7

3 5

5 4

1 0

0 3 0 2 0 0 2t6 4 6 6 3 1 3 1 o4 3 cT 0 0 4 9 2 1 0 0 0 0 0 1

1 t6 2 A 1 2 0 0 0 0 0 7 7 e 8 8 7 7 t / /

8 8 a 8 9

/ /

D 1 1

0 1

/ 2 2 .

/ / /

6 6 6 6 b S x> -

(l i l

ro C C L L C L. L L oS / / /

f O O 0 O O O 0 0 n R R R R R R R R I S S C C S C C C y

l R e Z yt h P l a wg d . yem ou e- dd l t i l o a eax S . pl .l f r T R pe t oe . t mo h I

Z ov nch e air f t B P re e g mxp f -

1 T dl vnxnoR a s i op xra ot k u E d I n ld s d r n op ao IF a ee n a% m o rpo e l g o y e1 ol l pat BO d vt ea l r h rl l p n BE e a l c l a

a yf u f

a g

adh eg l i ak FC i g

m xl r pi ea FB h rd i R ed ABU x en zi 8 r

p t eS aat I 6 o ph f o se l

Q R i i mnL 0 re p E rn 9 o vieV 4 ddl om S n ua 1 t xvR 3 a a t o e s 4 R o 2 el c sr e so '

H Z rsn h es f w et f O P pai s vf o t r . o I pg p xef t P e P%) L oa e u Rl o C b 0l p t es e AR

1 a u s d ma k mRt 3 d d c e m d eoe a oZ n 0" e ey k a ass r m rP e 3C .

l tld a r ea c f w 1 " l a

l nel m g h e n ) - -

a et o r .i S s S V& e t vac ) 0 xcl C asI C s s m( S 0 R ee% R gel n di C 6 d v1 ( tV H "B P i ax% R d e d uR d" C eo6 ( d elld T en . e R t h r3 e t e e C t i % sm n p T d nvl t V nm e %0 2

loro "A e x pf C d eeao V Rao V A Vl cN 3 V21 1 Cf" t 5 0 ne 7 5 em 7 vi 7 ET 7 7 l

ae 0 3 3 2 4 5 8 5 0 8 8 um ti 1

6 2

6 2 o3 7 t 7 5

3 2

4 2 o2 1 t4 0

8 5

8 2

8 1 1 cT 0 0 0 0 1 1 1 1 1 A

7 e 8

_ t /

- a 1

_ D 2

_ /

6 Eu

.[ .M n;

Notes:

L1) : Prior to. 6/17/87,' reactor had been shutdown.for 60 days for a refueling routage. i 2): Unit 1 remained;in Mode 5 during entire event.

3) RHR^ remained.in service during. entire event.

-4); PZR level transmitters vented on 6/25/87'resulting-in an approximately 1-2 percent . reduction in indicated PZR level.

'5) , Cold cal PZR-level reads several percent higher than hot cal andLis more

-accurate at the RCS-temperature levels during this event. .However, at higher? PZR levels,; hot cal' reads higher ~ than1 cold cal. The relationship is not linear, and in general, hot cal reads higher than cold cal.

6)A11 RCS inventory additions during the event.were makeups to. the. v,olume control tank (VCT).. Each makeup was recorded in'the control-room operator logs for the analysis period. 'A summary of the additions is provided' below:

Date Time Total Makeup'

, 6/171 1730 126.1 6/17 1812 342.1 6/17 1907 788.5 6/17 1926 1,818.1 6/17 2005 2 5.1 6/17 2026 24.0 6/17 .2038 21.1 6/17- 2055 24.8:

6/17 2100 5.771.0 6/17 1200 to 2400 Total 8,940.9 6/18 1710 371.0 6/19 0239 325.5

'6/20 0208 781.6 6/20 0500 555.1 6/20 1133 428.5 l 6/20 1235 292.6 L

6/18 0000 to 6/20 2400 Total 2,754.3

l. Total additions to RCS are:

6/17 total gallons '8,940.9 6/17 to 6/20 total gallons 2,754.3 6/21 makeup 0112 to 0640 19,747.3 Total Additions 31,442.5 Page 4 L_-_=___--___

.i-

~

{

' APPENDIX 2 I

INDICATIONS RANGE ~0R INDICATOR TYPE

l

- DIRECT

Pressurizer Level L

-; Hot Calibrated 0 - 100% Level L-459-460-461

- Cold Calibrated 0 - 100% Level

.L-462-

~

Rx Vessel Level Indication. System (RVLIS)

- Wide Range 0 - 100% Level'

, - Upper Range 65'- 100% Level

RCS Standpipe 'Tygon Hose'with a. rule attached to measure water level above centerline.

(Monitored by local operator or'TV. camera in Control Room.)

INDIRECT

/

' RHR

- Pump Amps Ammeter

- Flow (FI-1605) 0 - 8500 gpm

RCS

- Loop trassures 0 - 3000 psig Wide Range 0 - 600 psig Narrow Range

- RCP Ampn Ammeter

RVLIS

- Dynamic Range 0 - 100% Full a P

Indications are monitored in the Control Room unless otherwise noted.

Page 1

- _ - _ _ - _ _ - - .

ML20235U670

Text

SUMMARY

SUMMARY

Navigation menu

Search