IR 05000338/1987021

Insp Repts 50-338/87-21 & 50-339/87-21 on 870621-0710. Violations Noted.Major Areas Inspected:Introduction & General Description of Loss of Coolant Sys Inventory Event & Contributing Factors Which Complicated Event

ML20237L166
Person / Time
Site:	North Anna
Issue date:	08/11/1987
From:	Cantrell F NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II)
To:
Shared Package
ML20237L133	List: IR 05000338/1987021 ML20237L131 ML20237L149
References
50-338-87-21, 50-339-87-21, NUDOCS 8708270444
Download: ML20237L166 (23)

Text

.

, ,,

pH Rfoug UR!!TED STATES 1 4 'o fdUCLEAR REGULATORY COMMISSION

Tg REGION ll

, ~$

'8 j- .101 P lARIETTA STREET, A%ANTA, GEORGI A 30323 ,/

.

....

Report Hos.: -50-338/87-21 and 50-339/87-21 Licensee: Virginia Electric & Power Company-Richmond, VA 23261 Docket Nos.: 50-338 and 50-339 Facility Name: ' North Anna 1 and 2 l Inspection Conducted: June 21,1987 - July 10,1987 Team Members: J. L. Caldwell L.' P. King L. E. Nicholsen A. R. Long C. L. Vanderneit Approved by: d'h.4 ([ Mef F. S.~Cantrell, Secti

// b Date Signed Division of Reactor P ects SUMMARY Scope: This ~ isDa special inspection covering the loss of reactor coolant l system inventory event which occurred during June 17, 1987 to June 21, 1987,

.and involved -the following areas: Introduction and General Description of the Event, NRC Response to the Event, Sequence of Events, Potential ~ Safety Issues Associated with the Event, Contributing Factors or Factors which Complicated .

the Event, Reactor Vessel Level Instrumentation System Training, Enforcement and Conclusion Results: Two violations were identified: Violation of Technical Specification 6.8.1.a with four examples and a failure to perform a 10 CFR 50.59 safety evaluation. (See Paragraph 9)

,

i L ege2ggesage 1 G t _ _ _ _ _ _ _ _ - _ _

-

'

.

REPORf DETAILS Licensee Employees Contacted

• W. Harrell, Station Manager

• C. Driscoll, Quality Control (QC) Manager

• M. L. Bowling, Assistant Station Manager

• R. O. Enfinger, Superintendent, Operations M. R. Kansler, Superintendent, Maintenance J. A. Stall, Superintendent, Technical Services J. R. Hayes, Operations Coordinator D. A. Heacock, Engineering Supervisor J. P. Smith, Superintendent, Engineering D. B. Roth, Nuclear Specialist J. H. Leberstein, Engineer

• G. G. Harkness, Licensing Coordinator Other licensee employees contacted include technicians, operators, mechanics, security force members, and office personne * Attended exit interview Exit Interview (30703)

The inspection scope and findings were summarized on July 10, 1987, with those persons indicated in paragraph 1 above. The licensee acknowledged the inspectors findings. The licensee did not identify as proprietary any of the material provided to or reviewed by the inspectors during this inspectio (0 pen) Violation 338/87-21-01: Violation of TS 6.8.1.a with four examples for inadequate procedures and lack of procedures. (paragraph 9)

(0 pen) Violation 338/87-21-02: Faliure to conduct a 10 CFR 50.59 safety evaluation prior to conducting a test or experiment on the reactor coolant syste . Introduction and General Description of the Event On June 21, 1987, the licensee informed the resident inspector of a situation discovered on Unit 1, involving the inadvertent formation of voids in the reactor vessel head and Steam Generator (SG) tube are The licensee had already taken action to return the unit to a stable condition  !

I by filling the Reactor Coolant System (RCS) and venting the reactor vessel

, head to the top of the pressurizer. The pressurizer was also vented to

! the Pressurizer Relief Tank (PRT) via the Power Operated Relief Valve I (PORV). The licensee reported that it took approximately 17,000 gallons to refill the RCS and the reactor vessel was verified full by a 100%

reading on the Reactor Vessel Level Indication System (RVLIS). The other

.

__.________._______M

.

l l

.

'

conditions of the unit on June 21, 1987, were RCS temperature of 110 degrees F, RCS pressure of approximately 20 psig and a pressurizer level of approximately 36L The inspector responded to the station to review the event with the licensee and verify the unit had been placed in a safe and stable conditio Background information and a description of the event are provided in the following discussion:  ;

On June 17, 1987, the licensee had commenced a heat up on Unit I to change modes from Mode 5 to Mode 4. The unit was being restarted from a 60 day refueling outage. The plant conditions were RCS temperature of 195 degrees F, RCS pressure of 320 psig, pressurizer temperature of 425 degrees F with a steam bubble and both "A" and "C" Reactor Coolant Pumps (RCP) operating. Approximately 25 minutes after the "A" RCP was started, it tripped automatically due to an electrical ground on the motor. The licensee determined that it would be necessary to cool down and depressurize the RCS to allow the "A" RCP motor repair The initial pla'n was to drain the RCS to a level approximately 40 inches above the loop nozzle centerline to allow removal of the "A" RCP motor and prevent leakage out of the RCS via the RCP shaf Later the evening of June 17, following a review of the RCP design, the licensee determined that it would not be necessary to fully drain the RCS j and the decision was made to maintain a level in the pressurizer of '

approximately 20%., This decision was based on the RCP design which provided a backseat for the shaft to rest, once uncoupled from the motor.

The inspectors later were told that Westinghouse had informed the licensee that the RCP backseat would allow a maximum of one gpm out of the RCS as long as RCS pressure at the pump was less than 15 psig. This information was contained in the maintenance procedure MMP-C-RC-28 but does not appear to have been communicated to the station operator RCS pressure was maintained between 15 to 20 psig during the even ,

The licensee also decided that for added protection, seal injection would be maintained to the "A" RCP at a pressure higher than RCS pressure to ensure any leakage would be into the RCS not out of the system. However at that time the licensee's understanding was that seal injection water entered the RCP above the backseat. Following the event the inspector discovered that based on a RCP technical manual drawing it appeared that the seal injection water entered the RCP below the backseat. The licensne discussed this situation with Westinghouse and confirmed that seal injection water was actually entering the "A" RCP below the backseat.

Therefore with a seal injection pressure of approximately 30 psig, the

"A" RCP seal injection was potentially unseating the shaft and creating the leakage out of the RC ,

] l I

-

'l

. 1 i

Along with the decision to maintain a level in the pressurizer, the licensee decided not to install the spool pieces and hoses necessary to fill and vent the RCS following maintenance. This decision was based in i part on the elimination of the radiation dose which would be received j during installation of the venting equipment and to minimize the time necessary to restart the unit following the "A" RCP maintenanc The fact. that water level would be maintained in the pressurizer and the RCS  ;

was not going to be opened convinced the licensee that venting following 1 maintenance would not be required. However, the unit was depressurized and based on events associated with pressurizer drain, the requirement to vent the vessel head following maintenance was determined to be necessar )

Therefore, following the event, the head was vented per 1-0P-5.5 prior to resterting the uni .The instructions, provided to the operators to place the unit in a condition which would allow the RCP motor maintenance, were to follow 1-0P-3.4 step 4.15, cooldown and depressurize the RCS for the performance of maintenance where a pressurizer bubble would not be maintained and the RCS would not be opened. This procedure did not describe in detail how I the operators would depressurize the pressurizer or the final conditions of the pressurize On the evening of June 17, the operators raised and lowered water level in the pressurizer to cool and depressurize it. Later the following morning the pressurizer was filled solid and cooled using auxiliary spray. Several hours following the fill procedure, the operators initiated action to lower the pressurizer level to 20%. This level reduction was performed with the PORV closed because the operators felt based on vapor space temperature that a steam bubble still remained in the pressurizer. Therefore the pressurizer was drained and allowed to cool aown with the PORV closed and the pressurizer unvente This appears contrary to the intent of 1-0P-3.4, however, as will be discussed in section 9 of this report, it appears that the procedure was inadequat A review of the pressurizer vapor space temperature trend was not possible because that temperature is not logged or trended by the compute However, the data for pressurizer liquid space temperature was available and indicated that the temperature of the liquid in the pressurizer would not support a steam bubble. The temperature of the pressurizer liquid just prior to the drain was approximately 185 degrees F (See Figure 4).

This demonstrates that the pressurizer bubble most likely began to collapse following the initiation of the drain of the pressurizer. Also the motive force for the drain was provided by the Residual Heat Removal

.

(RHR) pum Therefore, it was not necessary to have a positive pressure i in the pressurizer to reduce the level to 20%. A low pressure or vacuum could have begun to be established in the pressurizer anytime following or during the drain down. Additionally, during the pressurizer draining ,

! evolution the operators were not required to monitor RVLIS or perform a I mass inventory balance to determine if the volume of letdown was equal to or greater than the change in pressurizer leve .

---

.-t___ . _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _

.

'

Following the pressurizer drain, RCS inventory was maintained by monitoring and maintaining pressurizer level constant. This was performed by adjusting letdown flow which was through the RHR pump back to the Chemical Volume and Control System (CVCS) to match makeup which was via the Volume Control Tank (VCT) through the RCP seal injection line Make up was obtained by maintaining a nitrogen pressure of :pp oximately 30 psig on the VCT. Prior to June 17, when the nitroge , nk1ket was established, a hydrogen blanket had been maintained on the v' . The RCS system was not degassed prior to establishing the nitrogen blanke Since the RCS was not going to be opened the licensee was not required to minimize hydrogen in the plan RCS samples indicated a hydrogen concentration of 6.4cc/k The inspector also reviewed the shift logs for the letdown flowrate and makeup flowrat The operators consistently logged a greater letdown flowrate than makeup flo cate (approximately 20 gpm vs.18-19gpm). This along with the knowledge that the RCS had at least a 2 gpm leak out of the system could have alerted the operators to the loss of RCS inventor However the meter for letdown flowrate was 0 to 150 gpm with the smallest (i.e. the first) mark on the meter being 20 gpm. Therefore the operators could have been mislead into thinking the discrepancy was in the meter accurac Just prior to the pressurizer level manipulation, the "C" RCP was secured and the operators used RHR to cooldown the RCS. Following the securing of the "C" RCP, RCS temperature dropped rapidly. Along with a review of the RHR temperature data, the rapid drop of RCS temperature demonstrated that the core had very little decay hea The operators continued to maintain system status with the RCS temperature relatively constant at 110 degrees F, RCS pressure relatively constant at approximately 20 psig and pressurizer level relatively constant at 20%

during the period between June 18 and June 20. The only major indicators that were changing were RVLIS which was gradually dropping and pressurizer liquid temperature which was also dropping. Either of these instruments could have alerted the operator to a problem. RVLIS would have told the ,

operator directly that vessel level was dropping and pressurizer liquid temperature would have warned an operator who understood the initial conditions of the pressurizer (i.e. hot with a possible steam bubble and the PORVs closed) that a low pressure or vacuum condition was being established in the pressurizer. An additional indication available to the operator would be an inventory balance. None of these indicators were monitored by the operators to back up pressurizer level which was being used by the operator as the sole indicator of RCS inventory. Also based on the apparent stability of the RCS parameters there was no reason for the operator to suspect the indicators they were monitoring.

.

,

l

.

--_---.-___-__._-___J

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

.

.

+ 5 During the night of June 20, Westinghouse personnel requested operations to reduce the leakage out of the "A" RCP seals. The first attempt to reduce the leakage involved lowering the pressurizer level to approxi-mately 5% to reduce the head of water on the "A" RCP shaft back- sea This was performed without a procedure by increasing letdown which used the RHR pump to provide flow into the CVCS. Apparently unknown to the operators at this time was the condition of the pressurizer which was at a lower pressure than the rest of the RCS, therefore, the effect of the operation of the RHR pump was to actually decrease RCS inventory greater than twice the volume reduced in the pressurize The RVLIS trace (see figure 1) shows a rapid vessel level decrease near the end of the event which is explained by this lowering of the pressurizer leve This RCS level reduction was also verified by the licensee using a mass i inventory balance between the volume of water diverted compared to the volume reduction in the pressurizer from 20% to 5%. The volume diverted  ;

was greater than twice the volume change in the pressurize ;.

At this point in time the reduction in RCS inventory was not due to the leakage out of the system but due to a forced pump down via the RHR pump i If the conditions had been right (i.e. a lower starting point in the  !

vessel and a higher pressur'zer level) and pressurizer level reduction had

~

continued long enough, RHR pump cavitation may have occurre However, based on the low decay heat of the core, the operator's instructions for loss of RHR and the abnormal procedure 1-AP-11 which uses the Low Head Safety Injection (LHSI) pumps to back up RHR, if all RHR is lost, the i total loss of decay heat removal capability for any extended period of time is highly unlikely. Also the licensee never actually came close to lowering vessel level below the top of the loop nozzles as indicated by the RVLIS trace. The inspectors reviewed the logs on RHR motor current, discharge flew, and temperatures and none of these instruments indicated that the RHR pump ever cavitate This reduction in pressurizer level evolution did not cause a reduction in the leakage rate because the leak was coming from the "A" RCP seal injection instead of the RCS. The fact that the leak was not coming from the RCS was not recognized by the licensee. Also, the condition of the  ;

pressurizer was not recognized or understood. Therefore, an additional i attempt was made to reduce RCS pressure thereby reducing the leakage by  ;

the ' A" RCP seals. This attempt involved the raising of pressurizer level to 50%, opening then closing the PORV and draining the pressurizer back to 20%. This would put the pressurizer into a vacuum conditio There are several problems associated with this and the previous pressurizer draining evolution. Both were performed without procedures and both had the potential of placing the RCS in a worse conditio The second attempt to reduce RCS pressure as described above, started at 0113 on June 21. This attempt was not completed because after the pressurizer level was increased to 50% and the PORV was opened, the PRT went into a vacuum and a reduction in pressurizer level occurred. The operators recognized this as an indication of void formation elsewhere in

_ _ _ _

.

'

the RC The operators began to look at other indications of RCS i inventory and found RVLIS was indicating approximately 79% vessel leve The operators isolated letdown and continued makeup to the RCS. The licensee management was notified and a decision was made to install the -

reactor vessel head vent spool piece and attach it to the pressurize A total of over 19,000 gallons were added to the RCS to bring the RVLIS- l back to 100% indication and a pressurizer level of 50L The vessel head vent.was opened and pressurizer level dropped to 36%, RVLIS returned to 100L The licensee performed a mass balance of the additions and losses  :

.from the RCS:from June 17 to June 21. This mass balance indicated a net-loss of approximately 17,000 gallons from the RCS which is equivalent to the 79% void in the vessel as indicated by RVLIS and an additional 73% of the total steam generator tube volum Following the event' the operators received instructions to monitor RVLIS, vent the vessel head at least once a shift and maintain the pressurizer  ;

vented to the PRT via the PORV. This condition was maintained until the

'

completion of the "A" RCP motor repair. Following the motor repair RCS pressure was increased to allow jogging the "C" RCP. The operators then vented gases from the reactor vessel head for approximately 30 second The additional makeup to the RCS following the venting was determined by the licensee to be approximately 865 gallons which is equivalent to approximately 5% of the total steam generator tube volume. Therefore the maximum steam generator tube voiding during the event, if RVLIS was an accurate indication of the vessel voids, was 78% of the total tube volum The licensee also evaluated the potential for damage to the PRT, after ,

having been placed in a vacuum. A review of the FSAR indicated that the PRT was designed to withstand a vacuum condition. The PRT was inspected and found to have a leak in a rupture disc. The rupture disc was replaced I prior to.the unit being started u Following the event, the accuracy of the pressurizer level instrument was questioned. There was a concern that with a vacuum in the pressurizer, the level instrument reference leg may have been affected. The licensee filled the reference leg and monitored pressurizer level. The level only dropped approximately one percent indicating that pressurizer level was accurate during the even . NRC Response to the Event The Senior Resident Inspector reported to the station on Sunday evening, June 21, 1987, following the notification by the licensee. The inspector discussed the event with the licensee and verified that the unit had been placed in a safe and stable conditinn. The inspector notified Region Il staff of the event first ',hing Monday morning, June 22. The regional staff requested a conference call with the licensee and following'the call decided to send the Resident Insper%r from the Surry Station and a

.

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

>

,

,

,

,

.y j I

,, 4

. l

'

li . ,  ;

"

regia ial' inspector from the Operational Programs Section. The regional

-

staff also notified the Project Manager in NRR on Monday, Junr2 22, and he participated in the conference cal On Tuesday, June 23, the Resident Inspector from tb Surry Station and the. Region II Inspcctor arrived on site.'

?# On June 24, the Director.cf the Division of Reactor Projects, Luis Reyes, and the Section Chfof in charge of North Aiya, Floyd Cantr, ell, arrived on site to receive a briefing from the licensee c1.the event and other previously planned topics. The Project Managed Leon Engle, from NRR in ,

Washington also reported to the site on June ,TA and attendsd the j licenseeb' briefings. The following Monday, June 29, an additional

'

e inspbtor from the 0,;erational Programs Section reported to the station

<- to continue the followup inspection of the eve & Sequence of Events

,

'

<

Information Date Time

'

^

_ Source

.s .

l- 6/17/87 0135* RCS Sample Indicates Hydrogen = CR0 Log

.O G.acc/kg 0520 VCT gas space sample indicates Hydrogen = 85%, Nitrogen = 15% CR0 Log  !

!

,

'0730 Completed RCS Leak Rate with the following results: Identified leak rate equals 1.254 gra, unidentified equals 0.1304 SR0 Log t158 ' Commenced RCS heatup in preparation for Mode 4 using the "C" RCP. RHR l is Still operatin SR0 Log 1218 Str t 1 RC-P-1A, ("A" RCP) CR0 Log

,. 1242 RCS T'emp 195 degrees F; RCS f .' ' Pressure 320 psig; PZR Temp 420 l

<

degrees F with a steam bubble in Computer l

'the pressurizer j

l 1243 Auto Trip "A" RCP, indication of SR0/CR0 Log 4

'

motor ground 1259 Electricians report "A" RCP has a SR0 Log motor ground j l

j

-

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

_ _ - - . _ __ __ _ __-_- _-_____ _ _ _--

.

.g

.

1309 Secured RCS'heatup for "A" RCP work SRO Log 1421- Commenced breaking containment vacuum ~CR0 Log 1833- Stopped."C" RCP (Time O for Trend Plots; see figures 1, 2, and 3) CR0 Log RCS. Temp 182 degrees F; RCS Pressure Computer 320 psig 1907 Initiated filling and cooling of PZR CR0 Log 6/18/87 0000 Pressurizer level greater.than 100% Recorder (hot calibration).

0000 Pressurizer level equal to 100% (cold Interview calibration) and PORV open with SR0 0115 .VCT gas space sample indicates

. Oxygen = 0%, Nitrogen = 64.4%,

Hydrogen = 29.8% with a Nitrogen blanket on the VCT CR0 Log 0215 Reduced PZR level to approximately Interview to 20% (cold cal). Closed PORV at with 0600 approximately 80% PZR level.(cold SRO cal). Per the operators, vapor space temperature indicated a steam bubble (NRC comment - Liquid space temperature did not support this)

0420 Completed uncoupling "A" RCP- Maint. Dep Pressurizer level greater 15% (hot l cal) 19% (cold cal) Recorder 0605 Secured "B" Charging Pump, established  !

'VCT float via seal injection to all 3 RCPs with a 30 psig Nitrogen overpressure on the VCT CR0 Log 1 6/19/87 0337 Entered action statement on RVLIS j upper range level transmitters "A" and "B". Difference greater than 3%  !

(T.S. not applicable until Mode 3) I This was discovered during the performance of 1-PT-44.-7 by an operator. The operator logged 89%

vessel level in the survie11ance  !

test CR0 Log

)

0455 RCS sample indicated Hydrogen concentration = 6.5cc/kg CR0 Log

j

!

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

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

'

.

1914 Added 500 grams LIOH to primary (approximately 50 gallons water added CR0 Log 6/20/87 2200 Lowered PZR level to 5-8% in an Interview to attempt to reduce RCS pressure to with CR0 2400 reduce the leakage from the "A" RCP sea (NRC comment - This appears to be the only time in which the loss of RHR was possible.)

6/21/87 0113 Commenced makeup to RCS to raise PZR SR0/CR0 Log level to approximately 50%. The second attempt to reduce leakage by the "A" RCP seals 0130 Cycled PORV; PZR level decreased; PRT pressure decreased. This was the

- first indication recognized by the operators that the pressurizer was in a vacuum end there was a potential for RCS voidin CR0 Log 0112 VCT (RCS) makeup of 19,747.3 gallon to This filled the RCS based on RVLIS 0640 reading of 100%. CR0 Log 0230 Gassed PRT with Nitrogen and cycled PJRVs while continuing to raise PZR leve CR0 Log 0610 Vent installed between Rx head and PZ SR0/CR0 Log 0623 Rx head vented; Pressurizer level dropped approximately 10% to an indicated level of 36% (cold cal). SR0/CR0 Log 0723 VCT (RCS) makeup of 419.8 gallons CR0 Log to 0732 1354 Added 600 grams LI0H to primary CR0 Log 1425 Vented Rx head to PZR via head ven PZR level decreased approximately 1%

on cold cal level. Noted 1% increase in RVLIS full rang SR0/CR0 Log

.

___._____._--______w

,

10- .

-l 1128 3 VCT (RCS) makeups 2340.6 gallons CR0 Log to 1425 1800 Vented Rx head; PZR level dropped approximately 1%; RVLIS went-offscale high to approximately 102%. CR0 Log 1858 Closed HCV-1303A to stop seal leakoff flow from "B" & "C" RCP from leaking out through "A" RCP sea CR0 Log 1828 3 VCT (RCS) makeups of 1275 gallons CR0 Log to 2242 Notes:

1) Prior to 6/17/87, reactor had been shutdown for 60 days for a j refueling outag I 2) Unit remained in Mode 5 during entire even ) RHR remained in service during entire even ! Potential Safety Issues Associated With the Event i The potential for the loss of RHR during the event was reviewed.' The actual vessel level decrease. by-RVLIS indication at the lowest point was still 5.3 feet above the nozzle centerline. However, if the vessel level had dropped. below the top of the nozzle, the surge line for the pressurizer would become uncovered and most likely would have equalized or burped the remaining water in the pressurizer into the RCS. The level reduction in the pressurizer would have initiated several alarms in the control room associated with pressurizer level, alerting the operator to the RCS inventory problem. The operator would have adequate time to initiate make up to prevent cavitation of the RHR pump As discussed in Section 3, the only time at which the RHR pump was in jeopardy was when the operators took action to pump down the pressurizer level on June 20. This pump down also reduced the RCS inventory and if ,

the conditions had been right and the pump down had continued RCS level 1

'

-may have been reduced enough to cause cavitation of the RHR pum The inspector had previously reviewed the licensee's action taken in response to a loss of RH Based on these planned actions, the ,

availability of the LHSI pumps to back up the loss of RHR per 1-AP-11 and l the very low decay heat in the core, it appears that the operators had adequate time and equipment necessary to keep the core cooled.

l

!

.

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

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

'

The potential for the core to beccme uncovered was reviewe The actual vessel level decrease by RVLIS at it's lowest point was'still 8.8 feet above the top of the core. If vessel level had been allowed to drop below the top of the loop nozzle, not only would the water in the pressurizer and the remaining water !n the steam generator tubes most likely burp and flood the core, but it would also provide the operator with indications of a problem with RCS inventory in plenty.of time to initiate the necessary makeu Also if no operator action had taken l place once the RHR pump cavitated and failed, letdown would become greatly {

reduced if not secured, and make up would continue to the RCS refilling  ;

the vessel Therefore, there does not appear to be any potential for the

,

core to become uncovered due to the actions taken place during this even Based on the plant conditions, Mode 5, RCS temperature 110 degrees F and RCS pressure of 15-20 psig, the Hcensee could have purposely drained the RCS to the levels achieved during the event. This would not have been considered an unsafe condition. However, the problem with this event, <

even though the safety significance appears to be minor, is the fact that the RCS inventory was allowed to slowly decrease over a three day period without the knowledge of the licensee or the operators. and, therefore, should be considered a significant event. The licensee is currently reviewing and performing a human performance evaluation of the event and will be modifying procedures and operator training based on the results of their review. The inspectors will monitor the-licensee's actions taken as a result of the lessons learned from the even . . Contributing Factors or Factors Which Complicated the Event

,

There were a number of factors which contributed to or complicated this event. The following will be a discussion of the ones identified by the inspectors:

The initial problem associated with this event was the lack of procedural instruction and licensee understanding of the condition being placed on the unit. 1-0P-3.4 which established the condition for the maintenance war inadequate because the procedure did not describe the conditions that should be established in the pressurizer prior to allowing maintenance to j be performed. This will be addressed later in the Enforcement Section 9 of this repor The licensee also performed several evolutions on the unit without procedures such as the forced reduction in pressurizer level on June 20 and the attempt to place the pressurizer in a vacuum on June 21. These evolutions were not well thought out and did not follow any procedures which might have at least provided initial conditions and precautions to be observed prior to performance. This will also be l discussed in the Enforcement Section 9 of this repor There were no requirements to monitor any additional indicators reflecting ,

RCS inventory such as RVLIS or mass inventory balance between the make up I to and losses from the RC The operators relied entirely upon pressurizer level which because of the vacuum being created in the pressurizer was being artificially maintaine .

_______m____

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

.

'

The RVLIS instrumentation would have been the most direct method of telling the operators that a problem existed with the RCS inventory. The operators did not monitor or believe the RVLIS indication for several reason The RVLIS system had just been modified during the present outage and several work request stickers still appeared on the control room display. A system malfunction alarm continued to be annunciated in

'

the control room. Neither of these problems affected the operability of the RVLIS ' instrumentation but this had not been properly communicated to the operators. Additionally, RVLIS is capable of displaying several different displays of informatio The majority if not all of the time the heatup and cooldown curves were displayed instead of the RVLIS dat The strip chart associated with RVLIS had the wrong scale paper on it and the ' vessel level decrease was so gradCal it would be impossible to detect a trend in the viewing area of the trend recorder. Finally the operators ,

are not required to monitor RVLIS because Technical Specifications do not require RVLIS to_be operable in Modes 4, 5, and Since the operators had a positive pressurizer level indication and in most cases that. is the best indication of RCS inventory, the operators did not feel the need to monitor RVLIS. If RVLIS were monitored as demonstrated by the performance of 1-PT-44.7 on June 19, where an operator logged a RVLIS reading of 89% vessel level, it most likely would not have been believed over the constant pressurizer level. Another point of confusion, the performance of 1-PT-44.7 was considered unsatisfactory for the upper range because the readings indicated greater than 3% difference, one channel indicated 91% and the other indicated 86%, therefore, the operator declared the RVLIS inoperable as required. The full range instruments were both indicating 89L Logging the cooldown data per 1-0P-3.4 was terminated on June 17, even though-the 71 ant and pressurizer continued to cooldown up until June 2 If the operstor had been required to monitor the cooldown, especially of the pressurizer by logging the data every 30 minutes, the operators might have been alerted to the change in the pressurizer condition Pressurizer vapor space temperature is not required to be monitored or logged in Mode 5. This temperature reduction could have also alerted the

, operator to the loss of the bubble in pressurizer.

l There is no method of monitoring pressurizer pressure in the pressure ranges associated with this event. Therefore, the operator could not use a pressure instrument to verify a positive pressure in the pressurize A review of the R0 and SR0 logs revealed no mention of the evolutions that had taken place on the pressurizer on June 17 and 18. The logs made no mention that the pressurizer contained a steam bubble or that the PORV ,

was closed. Additionally the procedure which the operators were using to

'

place the pressurizer in that condition would not have indicated the final condition of the pressurizer. Therefore, there was no documentation which indicated that operators other than the ones involved in the pressurizer drain were made aware of the actual condition of the pressurize '

.

- - _ . __ _

i

'

.

'i The operation staff did not understand the origin of the leakage by the

"A" RCP shaft. This resulted in the performance of two unnecessary evolutions ~which could have had adverse affects on the uni . Reactor Vessel Level Instrumentation System (RVLIS) Training The design change installing the original RVLIS was completed in 1982 and TMI action item II.F.2.3.8, Implement Reactor Vessel Level Instruments, i for both units, was closed out in Inspection Report 50-338/85-12 and 50-339/85-12 issued in' mid-1985. However, the original RVLIS was not installed on the training simulator until late-198 The installation of RVLIS on the simulator was also never validated to ensure it accurately reflected the operation of the RVLIS in the control room. Therefore, ifcensee training staff members involved with simulator training stated that the RVLIS on the simulator could not be relied upon for accurate dat The licensee also told the inspectors that during training on Revision 1, Emergency Procedures (EP), conducted early-1987, operators were - told that the system was providing unreliable data. Because instructors considered the data provided from RVLIS on the simulator as unreliable and the fact that it was not installed, therefore, not used in

' training until late-1986, there was the potential for conditioning the operators not to rely on RVLIS. This conditioning of the operators could have contributed to the lack of response and use of RVLIS during the RCS vacuum proble Classroom training on the newly installed RVLIS in the control room was given by Westinghouse in mid-1987 to the operations personnel. The

-

inspectors reviewed the lesson material that was given to the students and determined that it appeared to adequately cover the operation and installation of the new system. The inspectors also reviewec, class -

attendance records to determine if all operation personnel had attended the training. The inspectors determined that three shift and several non-shift licensed personnel had not received the classroom trainin Due to the nature of the training and in light of current problems with the use of RVLIS at the facility, this training should be provided promptly to all operational personne The inspectors interviewed training personnel associated with improvements to the simulator and was informed that the modification of RVLIS on the simulator is currently in progress and will be completed in the near future. Completion of this modification to the simulator should include the validation of the instrumentation and then the full utilization in future simulator training sessions. This will improve the confidence of operations personnel in the use of the new RVLI . Enforcement The most significant aspects of this event involve allowing the unit to be placed into a condition that was not fully evaluated or understood by the

,

licensee and that voiding in the vessel head and steam generator tubes

! occurred over a three-day period without the knowledge of the operator This existed due to poor planning and inadequate procedures provided to

'

the operators and mechanic .

~ . .

.

'

Technical Specification 6.8.1.a requires written procedures be established, implemented and maintained covering the areas associated with filling, draining, venting, startup,- shutdown and changing modes of the RCS and maintenance relating to repairs of the RCP Operating Procedure (0P) 1-0P-3.4, Unit Shutdown from Cold Shutdown (Mode 5) <200 degrees Fahrenheit to Cold Shutdown (Mode 5) <140 degrees Fahrenheif and K 0.95, was the procedure used by the operators to place the unit Tbft7<le condition which eventually resulted in the inadvertent voiding of the vessel head and steam generator tubes. This procedure was inadequate in that it did not provide detailed instructions on how to depressurize the pressurizer or the final conditions of the pressurizer. The step covering this evolution simply stated "If the RCS is to be depressurized but not opened, then perform steps 4.24 thru 4.32 of this procedure, then go to 1-0P-1.2 when maintenance is completed. N/A all remaining steps of this procedure". Step 4.27 of 1-0P-3.4 states

" Secure the pressurizer heaters". Step 4.28 of 1-0P-3.4 states

"Depressurize the reactor coolant system by raising and lowering pressurizer lev.el, by use of auxiliary spray and cycling of the pressurizer power operated relief valves (PORVS)."

Steps 4.27 and 4.28 are the only steps which provide the operators instructions on how to depressurize the pressurizer and what condition to leave it i These steps imply that the pressurizer should be depressurized without a steam bubble but do not specifically state tha Instead, on June 17, the operators using step 4.28 actually drained the pressurizer with a steam bubble still remaining (i.e. not fully s depressurized) and unvente This appears to be contrary to the intent of the procedure 1-0P-3.4, but since the procedure was not specific, it is considered inadequate. The inadequacy of 1-0P-3.4 is identified as the first example of violation (338/87-21-01).

An additional confusing point involves 1-0P-1.2 which was supposed to be performed following the maintenance as stated above. This OP has an initial condition of a steam bubble in the pressurizer implying that upon exiting 1-0P-3.4 there would still be a steam bubble. The licensee has reviewed 1-0P-1.2 and changed the procedure to no longer require a steam 1 bubble to be an initial condition.

( As discussed earlier, on June 20 and 21, the licensee attempted to perform i two evolutions on unit 1 in an attempt to reduce RCS leakage out of the

"A" RCP. Since the conditions placed on the unit were not familiar to the operators, otherwise routine evolutions such as lowering pressurizer level can not be considered as routine evolutions and therefore they should require procedures. However both of these evolutions were performed without procedures and because the origin of the RCS leakage

.

_ . - _ _ - . - _ _ __s

- - _ _ _ _ _

.

'

was. not understood by.the licensee, neither would have had any affect on leakage reduction. ' Additionally because these evolutions, reducing pressurizer level and drawing a vacuum in. the pressurizer, were not reviewed properly which would have happened if they had been

, proceduralized,. both evolutions had the potential for adversely

'

' affecting the unit without accomplishing their their intended goa .The performance of the reduction in pressurizer level on June 20 without a procedure, which resulted in additional voiding of the vessel i identified as the second example of violation (338/87-21-01). The intended performance of drawing a vacuum in the pressurizer on June 21 without a . procedure; is identified as the third example of violation

.(338/87-21-01).

Maintenance procedure MMP-C-RC-28, Reactor Coolant Pump Coupling Disassembly / Reassembly, used during the repairs to the "A" RCP in June 1987 appears to be inadeqtate for the following reasons: There'is no action' statement in the procedure directing the operators to maintain the RCS at 15 PSIG to minimize the amount of leakage from a disassembled RCP. The operators are required to not exceed 15 PSIG by step 5.4.a while lowering the shaft to minimize RCS leakage, however, no guidance is given to the operator on maintaining.15 PSIG while the' pump is disconnected. A caution preceding step 7.1 that the maximum expected leak rate from a backseated RCP is 1 GPM with the RCS at 15 PSIG, however, this.is caution and implies no action on the operators part to maintain those condition There is no guidance provided for the use of seal injection when the procedure is used without the RCS being drained. Because this condition can occur and the effect of seal injection on a 1 disassembled RCP directly effects the amount of leakage from the RCP, 1 guidance should be included in the procedure, The steps for the alignment of the RCP and motor are not included in i

'

the procedure The procedure is used to disassemble and reassemble the RCP coupling and prior to reassembling alignment of the coupling .

is needed, those alignment steps should be included in the procedur ! There is no statement concerning the minimizing of RCS leakage when ,

the pump internals are raised from their backseated position. If the procedure states that the operators shall not exceed 15 PSIG when lowering the pump internals, the same direction should be required when the pump is being raised as an increase in the expected ,

leakage rate is just as likel j

.

These four exemples of the inadequacy of MMP-C-RC-28 listed above are identified as the fourth example of Violation (338/87-21-01).

10 CFR 50.59 states in part that the licensee may conduct tests or experiments not described in the Final Safety Analysis Report (FSAR)

without prior Commission approval if the licensee determines the test or experiment does not involve a change to the Technical Specification or

'

an unreviewed safety questio The licensee must maintain records of these tests or experiments including a copy of the written safety evaluation which provides the basis for the determination that the test or experiment does not involve an unreviewed safety questio Contrary to the above on June 21, 1987, the licensee conducted an evolution with the intent of drawing a vacuum in the pressurizer, in an attempt to the reduce RCS leakage. This evolution is not described in the FSAR and is clearly not a routine evolution. Because the purpose of the evolution was an attempt to change a plant parameter with a nonroutine plant manipulation, it is considered a test or experiment. Based on the above and the fact the licensee did not perform a 10 CFR 50.59 safety evaluation this is identified as a violation (338/87-21-02).

2 Conclusions As shown in Figure 1, the reactor vessel level as indicated by RVLIS started to decrease following the "C" RCP being secured. Figure 2 shows that RCS temperature also dropped rapidly at the same time indicating that the initial vessel level decrease to be caused by a shrink in RCS inventory due to the reduction in temperature and reduction in pressure in the vessel head which would also allow gases to come out of solutio Several hours following the "C" RCP being secured as shown on Figure 2, the pressurizer level was reduced. During this level reduction, Figures 1 and 2 show that RVLIS continues to drop and the slope of the vessel level decrease does not change until the pressurizer level decrease is secure This indicates that the forced reduction of pressurizer level using the RHR pump was also reducing vessel level. This was probably as a result of the pressure reduction in the pressurizer as compared to the pressures in the vessel head and top of the steam generator tube After the securing of the reduction in pressurizer level on June 18, RVLIS indicates that the vessel level slowly decreased until June 20 when pressurizer level was again decreased. This slow level decrease was due to RCS make up not being equal to the letdown plus the leakage out of the RC Also, during this timeframe, the pressurizer liquid space temperature was dropping as indicated by Figure 4. This temperature reduction in an unvented pressurizer indicates a low pressure or vacuum condition was being formed in the pressurizer. This pressure decrease in the pressurizer continued to develop a Delta P between the pressurizer and the vessel head and steam generator tubes allowing the pressurizer level to remain constant as indicated by Figure 2 and the level in the vessel and steam generator tubes to decreas As the pressures equalized (i.e. reduced) in the vessel head and steam generator tube area, low pressure vapor was probably formed as well as gases were allowed to come out of solution filling the voids being create . __ - _ _ __ _ _ _______ _ _ - - - _ ___ _ _ - _ _ _ _ _

.

'

On June 20, when the pressurizer level was again reduced using the RHR pump, RVLIS shows a greater reduction in reactor vessel level. This indicates that the forced reduction in pressurizer level also reduced the levels in the vessel head and steam generator tubes. Figure 5 shows the relative elevations of the pressurizer, vessel and steam generator RVLIS did not indicate an increase until the pressurizer level began to increase on June 21, 1987. The licensee, once recognizing the possible void formation in the RCS, continued to makeup to the RCS until RVLIS indicated 100L The root cause of this situation was the inadequate evaluation of the plant condition established for the maintenance of the "A" RC Contributing to the problem were inadequate procedures, the failure of the operators to use indications other than pressurizer level to verify RCS inventory, failure of the ifcensee to quantify and understand the source of the RCS leakage and the failure of the operators to understand and evaluate the condition of the pressurizer.

i ll

!

l

.

_ . . . _ . _

-

-

-

.

- i i

i i

-

i 8 i

i i

i 7 E i M

i

%

i I

T i i 6

. 7 s8 v/2 1 i

i w

o n

E6 / i

d t

G78 -

' i 5 h u

s N/ P A7 / 1

~

i R

C R6 i

'

C i 4 '

R 1 i

e c

ETI PN i'

i i

n s

U i s

P S i 3 0 r u

UP A i H

o S N i

I L i V i 2

R i i

i

-

i 0

i i

g i

3 /

- - - - - - - - - - - - -

0 5 2 0 8 6 4 2 0 8 6 4 2 0 8 6 4 N /ng 0 0 9 9 9 9 9 8 8 8 8 (

1 1 8 7 7 7 n

C $ UuliC_ et

_

f e$3 Eytr

-

. 1

i l

9

-- i

- i

~/

i i

i 0

i 1 i T

I i

N i 7

U ' i

- ~

~

i i

N __

~

i

O I

' s i

i 6 T ~

i A

i 7 i T8 / i 5

S1 2 - ~

i R6 /

i

' i E- i i 0 W8 7

/ ~

i

O7 1

~

i .

i P/ 6 ~ - i

A ~ i 3 i

N i N i A i i 0

H -

-

i T o* 0 i R

i

\ i O \{

i 0 1 N N\g i

i i

i t'

-

l'

i\ lI rI i

1 !' !iIje fI ii 4 - - - jM H _- - ,i bI

- f- ' - - - - _

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

.

.

_

_

_

.

_

_

_

_

-_

_

__

_

_

_

_

.

_

_

. 1

_

i

_

.

_

.

i

_

_

_

i i 12 i

i i

-

1 i

T i

i I

N i

- i 0 U -

i i

- _ i i 9 N i O

I i

T ~

i 8 A 7

~ -

i T8 /

i S8 1 i 7 i

R6 /

i i

E- i 6

'

i W7 8

/

. - i O7 i "

1 i 5 P/ 6 i

i A i i 4 N i N i A

i g' i 3 H i T i i 2 R i O i N

i

' i

,

x i b

s' M ' i I i l'

O i

- - - q - - - " j - - - - - O 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 O 4 2 0 8 6 4 2 0 6 6 4 2 0 8 6 4 2 3 3 3 2 2 2 2 2 1 1 1 1 1

.

. _ _ - - - _ --

% ..

20 ,

. . ... -

.-

l j . ,

..b , , .

. . ~ i l .

. . . .

~~ 1,

.

_....- _ _ f b k ' *. . UM S.PM_ K E ~~

I

._ ._ _

.,.'__ ,

.

7__...

._

.m . ._ ,

{ .

_ . _ . . .

, ,.

1~- _ _ _. . . ( .. .. - ( . . . . _

-

_ j . . . -.

2 _ . . . _ . . __

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

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

. ...

.._.._9

. . . ..

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

. 1

.-

t --

.

---

T . _ . _ . -

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

,

. . _ .

. . . . .. . . _ _ .

I

_ . . _

_ .._ , . . _ _ . .  ;

_ . .

UC0

_._,

___. _ .

, ___

._ . _ _ . .

.. . . - _ . .

. ..

. .. _ , . . . _ . _.. .. .__

. . . _ .

_ . . . .

v.. n

.

- . . ...

I .

. . . . .j .j _

j __ . . _ _ . . . _ . . . .._, .

. ..,. .. .. {

___).

.

.. . . . . . _ _ . .. '

_ ~ . . ,

..

.. . . . .

._. . . j ..-

. . _ . . . _ . . .

. . . .. . . . .;. _ . . . . . _ .

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

. _ . _ . .

__

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

- _ _ _

, ,

_

,

__

_ p._ _ _. .._._;._..___. . . . _ . 1

..

..._y_.-

_ . . .

, .._

q_. _ _. ..._..p.._ . . . _ . _ j

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

____.9..._ g

. . _ _

.f...__

_

c. _ _ _ . . . J . .. _ .___... .;

. _ . _ . . -

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

. ..

. . .

. . . . _ _

._.~

~~~

__ZZ_ _.[ ._

] . _ . . , _.J ~ " ~ ~~: 7 '- ~.Z. .

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

- . _ .

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

.

. _ . . _ . . .____7..._

)

.

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

~

_,..{.. .

_ . _ . .

._ . . . .. -..

_

__- . . . . . . .

.._._:. . . . _ . . . . _ _

_ _ _ . . .

..._- ..

. . . . . _ _

..y_ ._ . _ . . - -

. . _ _ ..

4 _ . _ . _ _ _ _. _ . . . . . _ . .

?

_,

r .. p  : ._. . ._ _ .

_______.L. i . . .

. . _ _ . _ . ._L._._..._

.._}- ,_...__ ,

. . _ . _ . -

.._

.4,

. .- .

._

_

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

. . - . . . _ . . .

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

.j M _a f

- 4 _ _ _ . .. . . . . _ _ . _ _ -

'

/

3 __ .j . . . . 4 _ . . _ _ . .. _ _ . . ._ .._..._ -

l  ! ._ - . . _ . . . _ . . - _ . .

_1- . h

., } .. .

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

}

_

.{ . .

- 4. . N _.___4._. _ . . . . j . .. . . _ __ . _ . . _ _

.-

. _ . . . _

..,.

.

.

. + . . . , ._

..h. ,. . .__

-h-

..

[- ,I ._._..j

- . .

.. ..~. l... . . _

. . . . _ .

.. ._ . 4- .} . ._.__

. . _ . . , _ _ . . . _ . :I q

..

t __ __ . . ,

- . . _ _ . _ - . . .

._

.

,

,

j

. . _ .

.

N .__ . _. __ t.,___... . .

,

..

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

, ..

_ _ _ . _ _ . ___ .. _ . ..

. . g ___

_ _ _ . _ . _ . _ y.. ... _

,

, i

, . _

. . . _ - _ . _ _

__.__.L__ _ . . . ! . _ _ . . - . _ _ . . {. . .

_ _ _ . _ .

. . . . _ _ .

__

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

,

ICO

. _ . . . . . _ . . .

- .

_...y

'

_

.. ..

. . _

. . _ .

. .. .

..} .. _ _ . . . . .

_ - . . .

. . . _ .

.

,

.__.q_..-..._ l . _ .

4 .-  ; . - __ .

l - .. ..

4 . _ __  ; ,

.. . h _ ... _. _ . ,.

._ .

.. .

_ . . . . . .

, .

_ . . ... _ . _ . _ . _ _ _.. i

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

. _ . . _ .

.

. . _ . ._ .-.4.,._. .

._ _ _ _ . .' ,

, . _ _ .

.

,

L.._ z

. . . .~ q . . .._ i , , i k l . . . . . _ _

__

.. , .

,

, - . _ _ j. - .

__...._.}. ,

.._.__3 .

_ . _ . _ . ,. .

. _ - -

..

_..y___- q l _ .. a __ _ . .. ._ j

,. ..j .

._;_..._. .-

. . .

.lg . . .. .

.1

._ . .

. ,

j j ._ . _ .f - , .

.

. _ . _

.

_ . . . . _ _ . .

. q

_ . . . . . . . _ . .. .. , _ .{ <

&QC iDX (W M ((((

4Jg g FIGURE 4 l2DQ CCCD g a

.

_ E L T L- a L

_ 2_ a a r n E.

_

_

_

O

_

.

_ m C

_

- n . t

_

r - s .

e - O

. T

_

. -

. .

_ M . c

.

_ . A - a

.

- R - r T . n -

.- S -

. _ . .

-

-

,

,

N I

- ,

-

.- , L .

-

, E - -

. -

_

_ T -

- -

.

, 2 1 .-

-

-

1 - - -

r X . - -

-.- .

1

.

-

-

-

- . -

- . - -

- .

._ - -

R O .- ,

, i

-

-

a MT . sf s ,

AA -

1 s

.j:

s c::j: .. .:sx

-:

.

' :

s 5 -

RR .-.

TE . ::q::f v t f i: a Ps s.ss x s e

c:s

gt-'

v- x. -: :.

.. :. :-

2:-cx

-- ::

-: :ys%

.

sC

.N'

SW

-

j  : J

-  :: .

x'-p 1 ~ s ,,

W .-. s

-

-

. . ,_ *

.- -

. * .- .

.

.

.- * .- -

-

. *

.- -- --

. P -

-

A P ** .- - A* -

. . -

-

T*

_

_

L .-

E L

• .- .

-

.-

.-

-

.

_

.

V .- E" - -

E V~ -

-

L E ~~ .- -

. L~ . -

~ - - -

. .

.- .

-

.

.- . " . -

-

. . - -

N-

. - - -

N**"*~~

. . . - -

-

. "

.

- .-

-

~

- -

. O*

- -

- -

- - .-

.-

.

~

P* - -

-

.

• -

- .- - -

- .

- * - -

• - .- -

.- , l

- -

,

-

- .

iu, . -

. '

. ii Ij

..

n i

,

- -

. -

-

-

% .

L

!

.

. - .

.-

.-

-

.-

- - -

E - -

V .- .. .

-

-

E .- .

.

L -

-

.- n

._ -

.-

-

-

1r -

r -

-

- . -

.- ._ .- -

-

N o

-

.-

- .

- .

.

.

.

P E

-

-

. -- -

-

. -

- -

-

- 9 - -

-

-

- 2 - - -

- . - .

- -

.

. -

-

-

.-

s lJ -

-

. -

. s

.

-

, s t f

. 0 6 . : .s :- T 4x:: :c ara

.

>

,

s- s-x :- .-

-:e

_

-

-

m

.

.- .- .

- -

. .

q%-

-

-

- . - .

.- .-

.-

- . -

.

- . -

.- - -

- .-

-

. -

3 g

L

2 g s

m 6 V 1 f

e .

R ~

-