ML20236E725

Mode Switch Team Rept Root Cause Investigation of Inadvertent MSIV Closure

ML20236E725
Person / Time
Site:	Pilgrim
Issue date:	05/08/1986
From:	BOSTON EDISON CO.
To:
Shared Package
ML20236E691	List: BECO-86-062, Responds to NRC Confirmatory Action Ltr 86-10 for Restart of Plant Re 860404,11-12 Events at Plant ML20236E712 ML20236E725
References
NUDOCS 8710290329
Download: ML20236E725 (32)

Text

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

e- <

1 l

l l

l 4

1 i

-I

.i ATTACHMENT (3)

MODE SWITCH TEAM REPORT I

1 gAo2%888: Sfossip

f ,, _

~

4

~

/.

Pi L t .l.'.-

,3 c'

~dk ' '

,;yg ,

if .

1' g 4 - . .

I

..(l e .t :' i .j

.} s '- ,,

.. u .,.  :

S ,

5

,; y.,

i n-1

. , i e

, 'M00'E SWITCH TEAM REPORT :q

.c ,

l Root Cause Investigation- H of Inadvertent MSIV Closure _,

s 5 .

.'. T.. ,

i 4

a

]

.; ']

Attachments: -

1 l

.J

, Typi. cal PCIS' Logic .

. Report on:Cause of Scram of April'4,-1986

' Decision Tree _ Logic. Diagram -

1

'l 4

i t

i t

i_ b i

L '

1

.l May 8, 1986 1

1 l

('

t l

'l

(-'

L1 = _ d

1.0

SUMMARY

OF EVENTS On April 12, 1986, power was being decreased for an orderly shut _down to examine possible RHR system valve leakage. The turbine generator was-tripped off line and the power descent continued. Before the APRM downscale alarms and main steam line low pressure alarms were reached, the Reactor Mode Switch was moved from the "Run" to the "Startup" position. The orderly shutdown continued, the main steam line low

, pressure (less than 880 psi in Run) alarm Channel "B" was received and l a short time-later the main steam line low pressure (less than 880 psi l

in Run) alarm channel A" was received. Approximately 30 to 40 seconds after the Channel "A" alarm was received, the Main Steam Isolation

~ Valves went closed causing a Reactor Scram . This event is similar to <

j an event on April 4, 1986.

Following the MSIV clostre and the subsequent reactor Scram', and after the PCIS logic had been reset, the control room operator tried to open the outboard MSIV's as called for by station procedure. The outboard MSIV's could not be opened. After continued effort the operator decided to try to open the inboard MSIV's. Those valves did open when ' --

the control switches were operated. The HPCI and RCIC Systems were operated for the purpose of depressurizing the reactor vessel. Reactor pressure was reduced to approximately 320 psi before the outboard MSIV's were able to be opened. This event was also similar to the .__.

April 4, 1986 event. .

The reactor depressurization continued until the cold shutdown condition was reached.

2.0 ASSESSMENT OF RELATIONSHIP OF 4/4/86 - 4/12/86 EVENTS TO PREVIOUS EVENTS An evaluation of the similarity of the subject events to previous events was conducted by,the Mode Switch evaluation team. The evaluation showed that 4 previous MSIV closures have occurred with the plant in other than the "run" mode. The dates of those events were

-10/3/72, 7/27/75, 4/9/78 and 3/22/86 .

On October 3, 1972, the mode switch was placed in the "Startup" - -

position approximately 20 minutes prior to the scram. As reactor pressure was reduced, the " main steamline low pressure" alarms, Channels A & B were received. The MSIV's then went closed and caused the subsequent scram. On July 27, 1975, again with the mode switch in the "startup" position, the MSIV's closed with reactor pressure approximately 900 psig, resulting in a scram.

On April 9, 1978, the mode switch was placed in the "startup" position 4 approximately 9 minutes prior to the scram. The " main steamline low pressure" alarms had been in for approximately 4 minutes when the MSI'/'s closed and caused the scram.

F .

w ,

e ,

~

.3 l0n March'22,;1986,:the station was-inJcoid shutdo'wn-and the mode: switch-  ;

Jx , - cas in:then" shutdown" position when.a Group 1;. isolation ~wasl received. ,

p' No alarms had come in prior to.or at.the time of the.' isolation which '

were related to the event.- <

+

~ '

.As described,ow1th-the exception.of'thkJ30Yt'o:40 second time delay, the April;4.71986' event:was' identical to'the_.Aprjl-- 12, 1986 event. A "

-detailed evaluation of the 4/4/86 event was performeo by pla'nt

~

! personnel (Jeference. Report:of Cause'_of Scramg of Apr.il-4, 1986 attached).; The other four events discussed were similar to the April ,

12, 1986 event in that an inadvertent MSIV closure' occurred'with the-1 mode switch out of.the'"Run" position'and plant conditions..such that' the,isolationLshould not have occurred. Since.these events were spread

'out;over a'.long period'of time with many contrc)ied shutdowns having I

-been performed during the. interim periods, thew events areLconsidered to be isolated events with a potential intermittent mode switch malfunction, spurious reactor high water level signals or loose ground connections as. common factors.

.3.0 ROOT CAUSE METHODOLOGY AND DETERMINATIONS 1 l_ . H The Kepner-Tregoe_(K-T) change analysis process was used'to identify a t first-cut. list of potential root causes.

-Change analysis techniques lwere originally developed at the. Rand ,

' Corporation and improved by Kepner and_Tregoe. Change analysis is a j systematic approach to problem solving with a very high credibility- , j built into the processi The change analysis technique was selected because in_-the initial stage of a root-cause analysis the causal- .

factors are ill ' defined _. Therefore',-it is'important.that all changes and differences-are identified whether they appear to make-any D difference'or:not. Change and difference: analysis efficiently identifies theiobscure and helps prevent' wasteful and ineffective

. actions. The method helps to reveal-factors which are not obvious. '

l Typical. steps in:the change analysis process are shown below.

i f

f f

CHANGE ANALYSIS SCHEMATIC Incident Situation A

Compare m Set-Down  %, Analyze n .<-

Differences Differences for Effect on Incident V

Comparable Incident-V Free Situation Integrate Information Into Incident -

Investigation Process With the " incident" being defined as an inadvertent MSIV closure, the following MSIV operational and functional information was used as a basis for the K-T change analysis, o MSIV closure occurs as a result of;

a. Reactor Low-Low Hater Level, a potential indication that s reactor coolant is being lost through a breach in the Nuclear system process barrier or that the normal supply of reactor feedwater has been lost,

b. Main Steam Line High Radiation , a potential indication of a -

gross release of fission products from the fuel,

c. Main Steam Line Space High Temperature, indication of a '

breach in a main steam line.

d. Main Steam Line High Flow, indication of a breach in a main steam line.

e. Main Steam Line Low Pressure (in the "Run" mode), potential indication of a malfunction of the nuclear system pressure regulation in which the turbine control valves or turbine bypass valves open fully.

f. Reactor High Water Level (in other than the "Run" mode),

potential indication of a malfunction of the nuclear system pressure regulation in which the turbine control valves or turbine bypass valves open fully.

o To' initiate MSIV closure, the activator logics of both trip systems must be tripped. The overall logic of the system is one-out-of-two taken twice.

i o Plant response to an MSIV closure event is a reactor. scram signal initiated by isolation valve position switches before the valves have traveled more than 10% from their "open" position.

o Instrument air is used to open the MSIV's, o Instrument air, assisted by spring pressures, shuts the MSIV. The valves will shut on a loss of instrument air.

o Each valve is piloted by two, three-way solenoid-operated pilot -

valves, one powered by AC, the other by DC.

o 'A test switch is provided to allow manual testing of each isolation valve from the control room. -

o A hand switch is provided to manually open and close each isolation valve from the control room.

o The MSIV control switches must'be placed in the closed positicn prior to resetting a full PCIS logic trip.

The first-cut list of potential root causes which was identified by the_K-T change analyris was used to develop a decision tree logic diagram (see /.it ziment 3). This logic diagram consisted of the previously listed MSIV opercH o ul and functional information itemized as potential root causes along with methocs for testing, inspecting and further analyzing the items and their associated circulty as a basis for eliminating or identifying the root cause(s) for the subject event.

The decision tree logic diagram s presented below in an outline format with an explanation of the tests, inspections and analyses performed and the results thereof.

l J

-s-L ,

L - )

l

_e !A. -PCIS LOGIC CIRCUITS l Primary Containment Isolation'is initiated from a "1 out of,?

taken twice" logic scheme. A trip of'any one-input from two redundant sub-channels together with a trip from any one input 1 from-a similar channel is required to initiate the isolation.

There are six trip signals per subchannel for the Primary Containment Group I Isolation. They are Reactor Low-Low Water Level, Main Steam Line High Radiation, Steam Tunnel High Temperature, Main Steam Line High Flow, Main Steam Line Pressure when in the'"Run" mode and Reactor High Water Level when not in the "Run" mode.

A failure in each of the channels could have-caused Primary Containment isolation due to any one of the following:

.o Circuit Operability i

A~ circuit failure was investigated as the root cause for the PCIS Group I Isolation. The four Primary Containment

-Isolation logic circuits were tested in accordance with Pilgrim's Station Procedure No. 8.M.2-1.5.3. This test-includes removing fuses and noting proper operation of the_  :

logic relays and control room annunciators. Five out of the i six relays associated with the Containment Isolation circuits were tested by this Procedure. The remaining relay, Reactor .

Hi Water Level, was tested by Pilgrim Station Procedure 8.M.1-19. All actions required to take place during the performance of the test were completed without deviation. '

The tests concluded that the logic circuits for the Primary -

Containment Isolation System was not the root cause.

o Relays Relay failure could have caused the Primary Containment Isolation. The relays were visually inspected in addition to the.above circuit checks. Inspection included examination of the relay contacts and coils. The visual inspection showed no evidence of failure, i.e., contact pitting, burning, etc.

Voltage checks of the relay contacts were also performed.

All relay contact voltage readings were taken while in service. No excessive voltage was detected across the contacts which indicates that contact resistance was acceptable.

All measurements exhibited the same electrical characteristics. Test results showed no inconsistencies and therefore relay failure was eliminated as & possible root cause.

o Wiring All PCIS Group I Isolation logic wiring was checked using Pilgrim Station Procedure 3.M.3-8. All wires were checked for integrity of termination including ground connections (i.e., loose screw, incorrect terminations, etc.). Many loose terminals, fuse clip and inappropriate growiairig terminations were found. These findings, aitnough significant, could not conclusively be assigned as the root cause of this event. (For further discussion, refer to Section 3.G.) '

c__________.______________

o Sensor Calibrations-Sensors which input into the PCIS Group I Isolation logic were functionally checked and calibrated except for the Main Steam Line High Flow and Steam Tunnel High Temperature sensors. All sensors that were calibrated and functionally tested proved acceptable. The main steamline high flow and steam tunnel high temperature sensors were not calibrated because the Steam Flow and Steam Tunnel temperatures would not be near setpoints for the level of power Pilgrim was at during the events of 4/4/86 and 4/12/86. The Main Steam Line High Radiation, Main Steam Line High Flow and Main Steam Line High Temp. Relays are alarmed in the plant computer alarm logger. Review of the alarm logger showed no evidence of one of these alarms prior to or during the event. The  !

sensors and corresponding Pilgrim Station Procedures used for l the Function test / calibration are:

SENSOR PROCEDURE Low Steam Line Pressure 8.M 2-1.4.4 Reactor Water Level 8.M.1-19 Main Steam Line Radiation 8.H.1-10 -

The sensors calibration and instrument channel function was therefore eliminated as a possible root cause. -

o Mode Switch Operability The Mode Switch b'ypasses the Main Steam Line Low Pressure Trip in all positions except "Run". Failure of this switch to bypass could cause a PCIS Group I isolation. A thorough test of the Mode Switch was performed using Pilgrim Station Temporary Procedure No. TP86-59. This test involveo monitoring the bypass contacts while moving the switch through its positions with the mode switch key either left inserted or removed following a switch transfer. The test indicated the need to remove the mode switch key after transfers to ensure proper contact alignment. Additionally this test and subsequent contact resistance checks revealed no evidence of mode switch contact degradation.

An extensive amount of data relating to Mode Switch maintenance and failures was reviewed for both Pilgrim Station and the entire industry. This review data consisted of an IE Information Notice, General Electric SIL's, General Electric Instruction Manuals, Industry Surveys and Pilgrim Station Failure and Malfunction Reports. The evaluation of this data indicates that the type of Mode Switch used at Pilgrim is a suspect for numerous intermittent failures. In rddition, GE in SIL No. 397 recommends replacement of this switch with a later "SB-9" model.

y< ,

1 Also. General' Electric indicated'that.the Mode Switch has~a 5 ' suggested life of.2000. operations. We have-estimated that LPNPS's; switch was= repositioned approximately 2500 times; therefore'the switch was not totally eliminated as a root:

cause to the PCIS. Group I Isolation.

o RPS Interface- 1

.The' logic'sub-channels which contribute to a Group.I ,

, Lisolation contain only one-signal which is interfaced.

directly with'the RPS. This signal is . supplied .by the. four Main Steam Line High Radiation Monitors. . One instrument channel contributes to each' logic sub-channel. Neither the y Main Steam Line Hi Radiation Recorder nor the Computer Alarm typer showed;any evidence of a Hi Rad trip. The Main Steam ,

Line Hi Radiation Honitor'RPS to PCIS interface is therefore 1 eliminated as a possible root ~cause. 1 B.. LOSS'0F' INSTRUMENT AIR  :

a

.At:the time of the event both the. inboard and outboard MSIV's were being. supplied by the-station essential instrument air supply. .;

The air supply normally operates at 110-100 psi. Closure of the MSIV's could-have resulted by essential instrument air supply low  ;

pressure, or localized air supply failure. Atwood Morril, the .

i MSIV manufacturer, states that 46 psi air pressure would be'the~

minimum air pressure neededito open the MSIV's and holdLthem ,

open. The following control room annunciators monitor the air supply system: -

92 psi. decreasing - Standby compressor running 65 psi decrea. sing - Service Air Header Isolation  !

- Instrument Air Header Low Pressure 55' psi decreasing'- Non-essential Air Header Isolation 5 psid - Air Dryer High Differential Pressure.

It is concluded that the loss of the essential instrument air supply would have been instantly and easily recognized in the control room-and it would have been noted on the Scram Reports.

Also loss of the air supply would have caused the MSIV's to close at various times. The alarm logger shows that the MSIV's closed ,

within fractions of a second of each other indicating the necessity of a logic actuated trip.

The analysis of the events shows that failure of the air supply to the MSIV's could not have contributed to MSIV closure and is therefore eliminated as a possible root cause.

C. LOSS OF AC/DC SOLEN 0 IDS OR LOSS OF AC/DC POWER

, Each MSIV is controlled by three (3) solenoid valves, an AC, a DC L and an AC test solenoid. Both the AC and DC solenoid valves are energized during normal operation (valves open). The AC test solenoid only energizes when the MSIV test pushbutton is depressed

_a_ 1

__.____.u-___m._--_.- _ _ _ . -

" at the Main Control Room. To close the MS!V's both the AC and DC solenoids must be de-energized at the same time. Beyond this, the t inboard MSIV's are supplied by the "A" division of the 125-V DC l System and'the 120 V AC Safeguard Power Supply, the Outboard MSIV's are supplied by the "B" divisions of both.

To close all eight MSIV's, the following scenarios or combinations would have to occur:

- Loss of both 120V AC Safeguards power supplies and Loss of both 125V'DC power supplies at exactly the same time.

Failure of both the AC and DC solenoids in all MSIV's.

- Failure of AC and DC supply fuses.

After the event, tests were performed (3.M.3-8) to verify the condition of the AC and DC solenoids, none were found to have failed. The AC and DC power supply fuses were also' tested and e found to be functional. These tests verify the power supply and solenoid status following the event. Power supply failures at the time of the event would have been easily noticed in the Control Room because of the large number of other equipment failures and alarms. Failure of fuses would not have gone undetected because the fuses would have had to have been manually replaced to enable -

reopening of the valves. Besides this, It is highly unlikely ell fuses would blow at the same time. Failure of the AC and DC Solenoids would have been a decisive failure, which would have been recognized after the event by not being able to open the ~

respective valves.

The testing and subsequent analysis of events shows that failure of the power suppl.ies, fuses or solenoids could not have contributed to the event and are therefore eliminated as a possible root cause.

D. MSIV TEST PUSHBUTTON FAILURE OR INADVERTENT ACTUATION I i

Each MSIV has a test pushbutton associated with it. The purpose j of the test pushbutton is to prove MSIV closure by slowly closing ]

the valve. It is used in testing the 10% closure scram trip  !

circuit. The pushbutton energizes an AC solenoid that is powered I from the same 120V AC safeguards power supply that supplies the AC trip solenoid. This causes the air to slowly bleed off the MSIV causing the valve to go closed. It takes the valves up to 30 seconds to get to the 90% open position using the test switch. No two valves have the same closing speed when the test pushbutton is used.

To cause a scram due to MSIV closure at least three MSIV's must go less than 90% open. Electrical failure of the test solenoids would result in not being able to close the MSIV's by the test pushbuttons. Mechanical failure of the test solenoids could cause .

the individual MSIV to slowly close. The inadvertent use or l operation of the MSIV test pushbuttons is discounted as a possible contributor to the event because an operator would have had to physically operate at least three pushbuttons at once, this is

n

) .

m . physically impossible. Also,thealarm-typerbouldnothaveshown l thelMSIV closure trip as occurring at'the same. instant. Failure k

'ofcthe. test solenoids'are' discounted because:1t would have been isolated to each MSIV.

The analysis of events shows that failureJof the MSIV test-

'7' solenoids and associated circuitry could'not have contributed to the event and.are therefore eliminated as a possible root cause.

L E. INADVERTENT MOVEMENT OF MSIV HANDSWITCH The.handswitches are G.E. type SBM with pistol grip handle and~

maintained contacts. These handswitches are located at the-C904 Panel. 'They must be manually operated.and it takes multiple I

operation of the switches in order to close.all MSIV's '.

simultaneously or a minimum of three'to Scram.

J y Because of the. type'of: switch'used, the location of each switch and the number of switches that must.be operated to.cause a. Scram it:is concluded that inadvertent-handswitch movement did not cause the scrams-of 4/4/86 and 4/12/86--and therefore it is eliminated as a root.cause.

.F. OPERATOR INFLUENCE A

Single operator ~e'rror-'has been ruled out as the cause of-these .

events.- Since there was no requirement to do so, the mode. switch key was'not removed from the mode switch subsequent to changing the switch's position prior-to the.4/4/86 event. Test Number -

4

.TP86-S9, entitled'" Mode Switch Test for Steam Line Low Pressure i Bypass", performed'on April 18, 1986, verified the need to remove the~ key,from the' mode switch after changing its position to assure that the switch wa,s in'the correct position. This was done prior .i I

to the 4/12/86 event. Analysis indicates that no single inadvertent operator action could have resulted in these events. ,

Interviews with operation's. personnel were conducted which

-verified that procedural steps were adhered to during the controlled shutdown' leading to both events.

G. . INADVERTENT RELAY ACTUATION A concern exists where the PCIS initiation relay may inadvertently 4 actuate independently of the logic circuit inputs.

o' Physical Relay Fai' lures The relay failure modes considered were contact failure and coil failure. All associated relays were checked to be in the proper state for current plant conditions and that the contact fingers were physically together. Further testing was' performed.to prove that clos >.d contacts had continuity with no excessive-contact resistance. No anomalies were  ;

l- found during this testing. l

o Relay Actuation Due to Outside Influence The high water level instrumentation which contributes to the actuation of the isolation circuitry was checked for

= _ --- - _ _ -

_ excessive sensitivity while the 16A-K7's (Group I isolation relays) coil voltage and ground-loop voltages were monitored. The level instrumentation was found to be ~

sensitive but not excessively. 4 The initial test performed to verify sensitivity did, however, Indicate some unexpected effects on the three of the four 16A-K7 relays and ground circuits. Subsequent testing was performed to verify these findings and implement techniques to attempt to localize the problem. These_ tests did not generate the symptoms detected on the first test and therefore no concrete conclusions could be drawn. The initial symptoms did, however, emphasize the need to inspect for loose connections in the logic and particularly the neutral connections of the PCIS circuits.

These inspections revealed that approximately 25 percent (12 of 44) of the connections were found with varying degrees of looseness. Al o in PCIS Panels C915 & C917 most neutral,

" compression type" connections were found to be oversized and some of the neutral wires within the compression connector had their insulating jackets removed for only a depth of one-eighth (1/8) inch. For all neutral cables in these panels, proper size connections were installed and all cables were cleaned, lugged and properly torqued. These actions .

will eliminate neutral terminations as a contributing cause for any future event of a similar nature. Review of past events appears to indicate an increased occasion of inadvertent initiations since the end of Refuel VI. During -

Refuel VI the HFA, PCIS logic circuit relays were replaced with the new Century series HFA's which are more voltage sensitive than the previous HFA'S. The loose connections revealed during the inspections coupled with the new HFA relays could certainly have contributed to events experienced during plant operation but because the events experienced do not occur when the reactor is in the "run" mode, these factors are unlikely to be the sole root cause of these events.

As a more comprehensive check of the relays contributing to the Group I logic, a temporary test procedure was performed which checked the bypass circuits around the 880 psig pressure permissive and associated logic.

The 880 psig pressure permissive switches were pressurized and held at 1000 psig. The mode switch was then moved into the "RUN" position. These two conditions were maintained for approximately 45 minutes. This amount of time was chosen to allow the pressure switches and the mode switch to stabilize at their normal pressures and positions. At the end of the 45 minutes, the mode switch was moved to the startup position and the 880 psig pressure switches were slowly reduced in pressure until they just tripped. During this time the 16A-K7A thru D relay coils (each coil represents 1/4 of the isolation logic) and associated logic were monitored for changes.

The 16A-K7A thru D relays and associated logic did not change state and no significant voltage spikes appeared. This is what is expected to be seen if all circuitry performs

I correctly. With'the 880's just at trip point, a wait of approximately 2 minutes was maintained for observation.

Seeing no logic change, each switch was reduced further in pressure by another 20 psig. Another 2 minute wait was  !

maintained while the monitored logic was observed. No logic j chsnges occurred during this wait. The switches were then reduced to zero psig.and again no logic changes occurred. No relay failures were found to exist, therefore that issue is dismissed as a candidate for root cause.

H. CHANGE ANALYSIS Plant modifications since 1/1/86 were reviewed to determine if any could have affected associated relays. Reactor Water Level Modifications were identified as possibly affecting l Reactor Water Level PCIS initiation signals.

On March 15, 1986, the unit was shutdown due to high drywell leakage. Inspections revealed the source of the leakage to be a failed socket weld in the connection between the 2-inch nozzle (N16A) and the 1-inch pipe which constitute part of the lower (" active") leg of the "A" Yarway temperature equalizer column.

All major tasks performed during that outage which relate to -

water' level instrumentation were reviewed to (a) provide a  ;

summary of all changes , and tests of the instrumentation during that period and (b) assess whether or not any of the outage activities might have affected signals from this -

instrumentation. Particular attention was given to the protective measures and restoration activities associated with the instruments.

It was conclu'ded that (1) appropriate protective measures were taken and proper restorations were performed in conjunction with the N16A repair work and (2) there is no reason to suspect that any of the outage work would have affected the instrument signals during subsequent start-ups or shut-downs.

The possibility of PCIS initiation due to flashing caused by replacement of missing insulation on Yarway Variable and Reference leg was also analyzed.

During the N16A nozzle outage, insulation was found to be missing from portions of both the Variable and Reference legs of the Yarway Temperature-Compensated Hater Level System.

The concern is that hydrodynamic or thermodynamic effects due to replacing this missing insulation may have caused spurious initiation of the PCIS Group I isolation.

Effects cf Replacing Upper Leg Insulation Insulation was found missing from the portion of the upper sensing line between the reactor vessel and the condensing 3 chamber. This insulation was to maintain high temperature in this portion of piping to allow steam to conaense in the condensing chamber in order to

f maintain a reference water' level for the Yartays.

Hithout the insulation in place, the potential exists l for condensing tle steam before it reaches the

! condensing chamber in which case the condensate would run back into tht vessel due to the slope of the pipe. This could result in inaccuracies in'the water level system during start-up conditions when the~need for make-up water in the condensing chamber is the greatest. This situation would make all instruments indicate a higher reactor water level than actually exists. Replacing the missing insulation would eliminate concern in this area.

Since all instrumentation off of this leg would be i affected in the same way by any temperature effect, an " inadvertent" trip of PCIS is not likely as all indications would show that the Reactor Water Level is at the PCIS trip point. It is therefore concluded, that replacing the missing insulation on the reference leg could not cause inadvertent PCIS

. group I isolation. i Effect of Replacing Variable Leg Insulation Insulation was also found missing from.the portion of '

the vLriable leg between the reactor vessel and the Yarway Temperature Compensation Column. This insulation was part of the original design. Pilgrim Station operated with it in the past without this problem. The insulation is installed to reduce thermal stresses at Nozzles N16A and N16B caused by l a large thermal gradient. It is also to reduce heat input into the primary containment from the variabie  !

leg. The concern.is that the temperature in the vertical portion of the variable leg would be higher due to the presence of the new insulation and increase the probability of variable leg flashing.

The Recorder Charts from LR1001-604A & B show no indiction of flashing during last shutdown. This is, however, inconclusive as the recorder may not have been able to capture a high speed event like momentary flashing of the variable leg. A flashing event could not be proven unless the conditions existing at the time of event were duplicated and the system was monitored with high speed sequence of  ;

events equipment. '

L VIBRATION SENSITIVITY Vibrational effects on relays in the Control Room Panels C915 and C917 and on Instrument Racks 2205 and 2206 were also assessed. The vibration found at Panels 915 and 917 were minute in comparison with seismic qualification vibrations witnessed at test laboratories for various safety related equipment.

Instrument racks 2205 and 2206 which contain sensitive water level and reactor pressure instrumentation were monitored in January 1986 to determine if the possibility of vibration induced trips were possible.

It is unlikely that vibrations significant enough to contribute to pressure switch sensitivity could have caused a scram.

.3 Work activiti_esithat could have caused' vibrations in the vicinity of

, :the' instrument racks (2205 or-1206)'at the time of any mode. switch-(-  : actuation.during.1986 were also investigated. All available-paperwork

( showed:that no work was.being performed around the instrumentation-L 3 racks at the time:the'isolations occurred.

~4.0 ' CORRECTIVE' ACTIONS ,

c JAs can be seen by Section 3, exhaustive testing and evaluation of the I potential root causes has been performed. All tests resulted-in the instrumentation and' associated circuitry being shown to operate properly. Physical inspection and.walkdowns of the relays, contacts and wiring ~showed grounding-inadequacies and _ instrument sensitivity which could contribute to the subject event. Other items which could not be eliminated as potential root cause(s) are the reactor. water level instrumentation and the Reactor-Mode Switch. It is recommended

'that the mode switch be changed because of its possible-contribution to "

the MSIV closure.

The Reactor High-Hater Level Isolation signal, with the reactor in other than."RUN", continues-to be of concern because it is in the

. bypass circuit of'the Main. Steam Line Low Pressure Isolation Signal.

-All data shows'that High Water Level did not occur during the shutdown

.sequerce..and initial calibration following the event proved the instruments'to be. calibrated.and functional. Testing was performed to .

Identify any excessively sensitive water level instruments, none were considered sensitive enough to contribute to spurious isolation signals. Work was performed on the Reactor Vessel Hater level sensing l

lines during the outage previous to the 4/4/86 event. Analyses are -

inconclusive in determining whether this work contributed to the occurrence of the spurious PCIS isolations.

Neutral Bus connections as stated earlier have been repaired and eliminated as contribut'ing factors of any future event.

As a result of this event and subsequent trouble-shooting, several Failure'and Malfunction Reports were generated. Two Failure and Malfunction Reports (F&MR),were written as a result of the 4/12/86 event. This initial event was captured in F&MR #86-088 and 86-089. <

Their solutions have yet to be conclusively identified for reasons covered in this report.  ;

An F&MR (#86-90) was wr.itten to identify a problem with the resetting of a High Water Level trip that occurred during the controlling of Reactor Water Level, abeic one (1) hour after the MSIV closure. The associated circuitry has been tested numerous times during this investigation and the eent could not be reproduced. Therefore, no conclusive root cause cos.ld be identified; however, an intermittent ground connection could cause an effect of this type.

I F&MR (#86-99) was generated because an MSIV closure occurred during a ,

sensitivity test of the Reactor Water Level instrumentation. The MSIV closure is attributed to jarring of the instrument rack and is not

-likely to be related to the initial events. F&MR #86-095 identifies a Group I isolation that occurred during a surveillance by I&C technicians. The isolation could not be reproduced but could be related to the 4/12/86 MSIV closure.

L L" The. Team' recommends monitoring the: system during an actual plant shutdown as a final check that actions taken have eliminated the cause of these events. Under the limiting. condition of the' current plant status..1t is-impossible to duplicate the conditions that would' lead to; '

possible' excess vibration, or water level transients which can only be duplicated..during plant operation. Performance of such a test would indicate, whether water level perturbations or instrument rack .

' sensitivity to vibration were contributors to the spurious isolation.

signals. Plant personnel should monitor.the PCIS logic and other associated variables,. including the water level instrumentation, during a controlled ascent and descent through the-problem area. This monitoring ~will. prove there are no inadvertent isolations that could have been corrected through the mode switch replacement; or by  !

tightening of. loose wires and fuse holders. The test will also identify the source of the PCIS initiation if it has not been corrected. General Electric has been contracted to supply the Transient Analysis-Recording System (GETARS) for this startup test.

During this test twenty-eight individual digital signals will be monitored in panels C915 and C917. These signals will be one of each of the followlog from the four PCIS sub-channels for a total of 28 digital inputs. '

o . Reactor low Low Hater Level o -Steam Tunnel High Temperature .

o Steam Line.High Flow o Main Steam Line. Low Pressure o PCIS Group I Isolation Relay o Reactor High Water Level -

o' Main Steam Line High Radiation These signals will be extracted from unused " dry contacts".on the associated relays, wired out to terminal blocks and monitored by the GETARS.

Additionally, eight resistive voltage divider circuits, two per PCIS sub-channel, will monitor mode switch bypass circuitry, i.e. mode switch and high water level contacts.

Finally 'ive analog (Non-Q) signals will be used to. monitor Reactor Pressure and Reactor _ Water Level. The analog signals will be read by high quality, high input impedance amplifiers and will not interfere with Control Room indication. One spare contact will be monitored to flag mode switch movement. This will be accomplished by monitoring a dry contact on the SA-K27A relay (APRM Set Dovm.) which is deenergized when the mode switch is not in run.

t i

This testing will include bringing the plant up to approximately 20 percent of rated power and allowing the plant to stabilize for .

approximately eight hours. An orderly shutdown will be entered and monitored with the GETARS. If the event occurs during this shutdown, this system will capture it and an applicable corrective action plan will be developed. Barring a MSIV closure and Scram, a normal ascent to full power operation should take place. Monitoring of the PCIS system will continue until the installation of the EPIC computer system is' completed, at which time the temporary monitoring systern (GETARS) will be removed. The EPIC system will monitor the RPS & PCIS logic circuits in the future to make events of this type more identifiable.

In conclusion, no clear-cut failures contributing to the spurious isolations have been positively identified through the efforts of the evaluation team. As outlined in previous sections of this report, all potential sources of failure have been considered and tested to assess the PCIS system operation. The mode switch remains a prime suspect as the root cause of failures, both at PNPS and elsewhere in the industry. It is for this reason that this team recommends its replacement. '

5.0 BASIS FOR'RE-START The unexpected MSIV closure and subsequent reactor Scram has been -

exhaustively researched and tested by a multi-disciplined team in order to determine root cause. The results and methodology of this process have been reviewed by INPO representatives, who have concurred that all possible failure paths have been thoroughly tested or investigated to -

the extent possible, without duplicating the exact station operating conditions which existed at the time the event occurred. On-line i testing previously described is imperative to further pursue the root l cause(s) of this event.  !

Since this would require station startup, the safety concerns of starting up with the possibility of repeating this event are addressed as follows. Analysis performed to date indicates the only failure mode .

to be a spurious PCIS Group I isolation. An inadvertent or spurious l PCIS Group I isolation with,a resultant reactor Scram is of minimal i safety concern in that this sequence of events places the reactor i system in a shutdown condition which increases the margin of plant i safety and the transient, itself, is well bounded by the FSAR. It is I possible that the root cause(s) for spurious closure of the MSIV's will I have been corrected by replacement of the mode switch, the redressing of the water 1evel instrument cables, or the repairing of the neutral ,

connections. Should the operational testing not cause a duplication of '

this event, a normal increase to full power operation should take .

place. Continued improved methods of monitoring shall serve as a means l to analize any future transients of this kind.

-[ *MiCiyidhw ston Edison Company ,< RMG Control Numcer To: C. J.'Mathis

. Qr From: PD Sm1th/MT McLcugnitW'< .

Approved: J. A. Seery- Record Type A4.08 Date: April 8,-1986 Dept. Doc. TCH B6-115 (W1300)

Subject:

REPORT ON CAUSE OF SCRAM OF APRIL 4, 1986

References:

1. PNPS Procedure 1.3.37 (Scram Report' dated 4/04/86)

2. PNPS Procedure 3.M.3-8A

3. GE SIL #155 & #397

4. GE Test Report No. 70709-2

5. IE Information Notice 83-42

6. NED Memo dated 4/08/86

7. F&M 83-133, F&M 86-80

8. Drawing MIN Series ~

Distribution:

.. M. N. Brosee M. Maguire S. S. Wollman P E. Mastrangelo D. E. Sanford M. Pickett (4)

.; R. Sherry H. R. Balfour

~

EVENT DESCRIPTION On April 4, 1986, at approximately 2015 hours0.0233 days <br />0.56 hours <br />0.00333 weeks <br />7.667075e-4 months <br /> while shutting down the reactor, a reactor scram. occurred at reactor pressure of approximately 898 psig. At .

the time of the scram the Reactor Mode Switch was in the startup mode.

CAUSE Investigation revealed that the reactor scram was initiated by actuation of

/ the Reactor Protection System (RPS). The RPS System was actuated by closure

-'- of the Main Steam Isolation Valves (MSIV's). Automatic closure of the HSIV's was actuated by the opening of the Reactor Low Pressure Switches 261-30A-0 when reactor pressure decreased to approximately 880 psig without these same reactor low pressure switches being bypassed by contacts of the Reactor Mode Switch when it was transferred to the "startup" position. The initiating action of the four (4) Reactor Pressure Switches 261-30A-D should have been inhibited by the previously performed operator action of transferring the Reactor Mode Switch, GE Hodel 58-1, to the STARTUP mode from the RUN mode.

Although other indications generated by transferring the mode switch {

demonstrated a successful transfer from "Run," as evidenced by SRM rod blocks j and short period alarm being received, subsequent analysis showed that at least two (2) of the four (4) contacts of the Mode Switch did not provide the intended bypass function. The mode switch was transferred by an operator-in-training under direct supervision of the Nuclear Watch Engineer.

The. Nuclear Hatch Supervisor wiggled the mode switch to " feel" that it was in the right position. The mode switch key was not removed from the mode switch following the transfer.

)

1 l

l t

i l

l \

L

'c: C. :. "atn!* _ Date:- W !' 5. l?86 Op; Suejecti REPCRT ON CAUSE OF SCRAM'0F APRI 4; 1985 l: tCDE SWITCH EXAMINATION & DISCUSSION ls

[ ' Reactor mode switch malfunctions of.a similar nature have occurred at several l L 'other nuclear. plants and on the subject of I.E. Information Notice 83-43, attached, Pilgrim last experienced a problem with the mode switch in 1983 . .

l- '(Reference ORC Meeting Minutes84-104 and failure & Malfunction Reoort 83-133).

General Electric Information Letters (SIL) Number 155 & supplements 1 & 2; and SIL 397 discuss instances of failure of "SB" model switches and recommend actions to be taken. Following this event and in accordance with SIL 155, an inspection of the Reactor Mode Switch was conducted on April 5, 1986. No indication of cracking,_ broken contacts, or of any other adverse condition was observed. Examination did indicate that proper preloading of the switch contacts existed. (A copy of.these SIL's and inspection report are included as Attachments A & B respectively to this report.)

'Previously, in accordance with SIL 397, during Refuel Outage VI, in order to

,- facilitate. replacement of the reactor mode switch if problems were encountered with it during the post refuel startup, testing of an SB-9 model mode switch was= performed. This SB-9 mode switch, which was furnished by General Electric Company, is a used unit rebuilt by GE. (This unit was obtained through GE who had replaced two SB-9's during startup of another unit. These two replaced.,

SB-9's were used to construct the rebuilt one.) Testing was also performed on a 58-1 model switch. This testing indicated that both-switches functioned properly and e M b had a definite ' feel' during a transfer operation. The 58-1 required a specific technique be used to ensure proper alignment of contacts while the SB-9 operated in a stiff and hard manner.

Following this testing, some Operations personnel visited the test site and familiarized themselves with the feel and technique used to properly transfer the existing 58-1 switch. .

This familiarization reduced concern for the proper operation of the SB-1 Mode Switch. -This; experience coupled with the knowledge of problems encountered with the.new SB-9 switches and the extensive time required to change out and post-work test the replacement contributed to a subsequent' decision to continue operation with the existing 58-1 Mode Switch.

The operator who had transferred the mode switch in the Control Room prior to the scram had not been trained dn the SB-1 Model Switch at PNPS and had no previous experience with it. In retrospect, formal training of operators on the SB-1 Mode Switch should be routine, required training to continue operation with this model.

ROOT CAUSE The root cause of this event is that at least two of the Reactor Mode Switch contacts (10, 26, 42 & 58) did not close or remain closed after the mode switch was transferred from "RUN" to "STARTUP." Contributing factors to this event were insufficient development of a prescribed technique for transferring the mode switch and of a training program to instruct Operations personnel in its application.

. c ,- .

.%. . :, u . .2. .

' l C. 2. Mathisi .

Cate: ' A c.r t : 5. 366 sa ge :

Suoject* AEPORT ON:CAUSE OF SCRAM CF AFR:. J. 1956 CONSE0VENCES. -

A. review of'the designed operation of the Reactor Mode Switch (RMS) as applied

~

ito the Reactor Protection System and the Primary Containment Isolation System has reached the following conclusions:

1. RMS contact arrangement in'the RPS and PCIS automatic scram circuits is-the same as other contact input signals to these systems,'i.e.,

"one-out-of-two taken twice." Hith this arrangement no single contact failure can cause.a reactor scram, nor can any single failure prevent a reactor scram when one is necessary. He find no circuit where the RMS contact arragement degrades the logic of operation of these two systems.

In the case of the RPS man #hl scram circuits "two-out-of-two" logic is '- y.7,gg g employed. RMS contacts.9-9C (RPS ch. A) and 49-49C (RP_S ch_. 8).are designed to cause a reactor scram when the RMS is moved from the Refuel

.toward the Shutdown position. Should.one of these contacts fail to open

-(single failure), a half-scram will be initiated. This situation is obvious to the operator who is expecting a full scram at this time.

' Operator action-is then required to manually scram the reactor by'using the manual scram push buttons on panel 905. This is not an unsafe condition. .

2. Failure of two or more. contacts to actuate as required in either '

intermediate or actual switch positions could produce a scram (or a " half scram") if other parameters so dictate, as in the events of 4/04/86, but .

such'actilons are in the safe directions for RPS and PCIS. -

RECOMMENDATIONS Short Term ,

1. Develop a technique based on experience garnered during operator familiarization in Refuel Outage VI, which will ensure proper positioning of the Reactor Mode Switch.

(a) Suggestions indicate that removal of the key switch after transferring to Startup will improve the probability of a successful transfer and should be included as part of this technique. This point is not an established fact.

2. Provide operator training on the documented technique which is selected to ensure proper RMS positioning, t Long Term 4 I

1. Replace the existing SB-1 switch during RF0 #7 with a new mode switch.

Re-analyze and investigate any design improvements or experiences with SB-1 or SB-9 replacements throughout the industry in deciding the replacement option.

i

{-

I.

5... s.e . , vw- - a....-

'c; Z. '. Natris Date: Acrii 8, 1986 oage 1 Sucject: REPORT _0E_CAUSE OF SCRAM OF APRIL 4, 1986-

2. Consider whether spare contacts or tne existing or new moae swilch should be wired to indicating lights or the new EPIC computer to more positively indicate mode switch position. If feasible, prepare conceptual PDC for N00 review.

3. Replicate the exact mode switch at the Chiltonville simulator and include mode-switch experiences in modeling.

Attachments MTH:mp O

e O

u 4

1 i

_ _ _ _ _ _ -__ ___a

!' L .i' W Ns ?L.,1F,. l

q b '= = h 9  :

, , y '

_ - . i i- - i

.. .. a . l

\ iI

$ flm f,

-), lg .

• j i I g j, E

~$ N '

m

~

a t_. ,

C .

j.

g.,

se i-k l IjI~h i

~~

y- ) l il -

i, I.

l' 3 .I

' l N f h gR d

'h5 h. . ! l ll $ l

i s. 1'i !p! !.

_ Na 1 >t ,

r. o. ot gt

_ __ ) b: U.w T:4I)_ _ l ..___

6a .

• b

,e ,

l, ,

e

--p<)

i

,..

• g i .

/. . , ,

^

., , 3, _- m s. .a .

. I, I e , s._Q,i. l  !! _1-- *{'

w =

-l i

1.. 1ga1_ g ga F 4 b 1 L('_;l,'h}! U It , i1

' ~

.p. 5 L_&, 'l ll 5

,?

- i

, i. t i - g g .

4 e

.qi 3 .i Lei 5 1-a! - r m E)a d r._  : ,: u N

i

-j ~, {

= i I r l{ s 3

9 g r e ), N 5

in s , e _'j ; ( ; a;h ,

i 2

l

.w'

• [ dl j ly .a

, i - i, e 8 . d 1 l __, e_ L_ L _ _t 'l g

. !! 5--.5

-2 %l -r- ]

,.1 -

I I FS, --

I, ql lei

_ pll io- '

j l ,k c j(s' 3

I i

c d l g .. _. .

a : .

L . *i } .. _ l l _ . _ -

.ur 5 -

. . L i.1. ll ! _:i.'lht!, '-*3 Itii .

)- i i,.!_ . .

g; . I< .li. 9

, ,_ . . . . . . -- I ~~ j

_.lk -

f.--~9_ ,,

7 g -- g_1. ,' ,a _ .

~ v ., ..

ule. I ,

.; /, ,

4 ;aje; }i , . .. ..

l, n

p

. nj h

!-

tm

,pt pp I l

,e ..

I ..

V ,Y 1. ,

.l.l, L _._J l_ ._ . _J _.._ i Ie I*

t,,. ge i

.a _

1i !lfd, ij# _..

i i j t a .

.7 11,:f

.q , ,

8% .e-

,, . . - . i - i , i

, , 1

$ i $ k .

4-9__ E 7,_") M em.--- --be **' **

y

__l ee  !'

, l!II ' ,,, y e

( lyJl-- '4 ... ki i '*

l

.ly8l

,;  ;& a

. o ~' 40 j

W g _c: 6 (*.yv. # i pg l .

,, }a  ; E

_ _ _ _ _ _ _ _ _. u. _ 2 p, y ,

- llt--- pt

, _. i ... It ?

8 je a 4 8 o

y -

! !,7,, ! .k97 -

, 1

.. n. 4 gg: e g! l  ! l d

^l

_ ::  : . :; .'. :: .. . :: . . . l wl 14 -

l lk )

T  !! ( _h li! II

.is,s s

n ___I._y ,

l

__ - r<. Ii Is; .

li' . ',

1  ; .

g ,

I'E  !!$ Nj=u_ $ $$ l3 ll! ki$,!i_!i!!)_l$ , $-l Ill y

.eu,e a, t x8 t

.i e T 8 1a g .

g_. .

e g ,. g  :. -

I 3

' Im a p'

h .

sf 8 f

l u; ..!!$==8 === .

2

$l .

l,, i  ! t s 9 t! C  ! r;

-)l8l

~

gy. - =- _ _. ._

1

$ l n 4 ,e s - "+ s = . ,.. e ll

'ld 4

11 .

x I

s a; s 4.0 a i r  !

• . l;' . s' [' rt "i g

1 li /

1 e I PI d I s:l!!!

LS

8 35 .

E Si

$ E5 m --~-

ATTACHMENT (4)

BECo DISPOSITIONING OF RHR TEAM RECOMMENDATIONS L

Page 1 of 6 L' -- -- - . _ _ _ _ _ _ _ _ _ _ _ _

J p -

u t _

r a

.t -

E S M

A o R t -

F ._

r _

E o

% M i r

I T P _.

1 1 1 A A A e s e s e s nna nna nna oo d oo d oo' d did e dide did e e ter ter t er N n eat r eat r eatr O o bd ce) bd ce) bd ce)

I d nefh nefh nefh T t' el ec tel ec tel ec omera I e omera nmspo omera nmspo nmspo S .b O o r o r o r P l l csep l csep l csep S .l l eah p l eah p l eah p I i iR wta iRwta iR wt a D W W( W( W( .

4 6 m -

T f N e e' o E v t M l t s 2 H a s y -

C A

f v o

e l

l ei cf r'

.s p 'A e

g T i a T

A eh st u

w.y el n

ee dh o'

o l e P

ae gb i t h ct am v es Rt H -

a ke eu as i Ro hl es r t _

ta l a o8 v s f8 'A ns ee si 6 '

o n td 8 _

i) i 8e ei m2 B 6v ht r( ' e8 l ta e r2 1 a nn d

td u- 0V -

en s1 0 ie m

d a n0 1 .

e0 .e gm

)t 1 enc no ,

1 n o0 vor ic re'

( e tM li u rr r ato or Bf vas u tu R 9o R l R ce c O 2 O kl e O cs _

8 u cir oe -

8 g 1 e ecu h 2n 0i h ss et i 0l cos b 1 t 1 e yn d

0a n tdr 0e ri cnp aa mp 1 s . i ea 0 s am pp p e ne x .

Mec eg s gh o

ui nnt _

eqt r ii i o vrs 3 dao tm tt a

l oi d3 atr n ner dl aa v e a asf rn tt n

_ enc ea ern go ei

_ l ea lh l eo btr bt b pi dt N

O msa eih m

es mot erc rol p .

I ssc ss spu a l smr se l se T sn ii w A aot al ait D sca s s s m N i ne ie ifb it s E Dis Db Doo SI a M

M O 4 C 1 2 3 E

R A A A A

l i1il 1i!))i p

u _

t r

t a -

E S M -

A o R t -

F r -

E o M i I r -

T P -

s l t l i i m 1 nwedi a l nl B hba e s tca iitss

.mh pto e nna oo dide d -

ebweih ter N nu cmt eatr O osnci bdce)

I d oal r nefh T li o tel ec I

S el teef bi adg omera nmspo O

P l wuuas llk e o

l csep r

S l Dacas l eah p I iE vnea iR wt a D WNeilb W( 6 -

f o

3 r s o i e f e s g o g y a t n a l P o k a s i a n di . s seea e t i elh ld n c n tg - -

l o u .t e iJ n oh osnd l fi rt muu xor e

te aroA fi nm o dTa om fdne onon cd e . eii e l een kptg apan at lhh w s otto . eAl E geer ns l l ngvto i gl nte

,k r u

ran got a wuao osvcnab c d ec i

nf s nf l rr o kdii l Asd u gl ree a snt el s avRd y . paaedii po e r

r wht bmyc m .h de s ebioem arli et fth en h ol gsei tt e nrb ayvcersh yet a k sl eguyw l ent a apaf s R lb os eevsk eO a ce -

lh ased ne -

tk aeeh a ivee e c l vt m gagr hte) t i xch D r.wvzb laee ih a r

osabo k en f e d o i ee i o ennR rn mnldt ogakH ua eitl c tt r R sc tl uua c aCa s s oo eeh f en ye cn f gc ,

f asBo woro pi ssr eof a

eki l g t shffh ad ,l noc iT oc ee Aait a h dk u hl R0 t t nas

- t H0dael .ae

. fR1l ega nd ld N eo usan i es e

- O I

T t

ata gn n - o 1

weai h et k o l ngt sl rnr eaeio A i un0stli vtck p

- D N

toe 0i i mp1hffd d l sncp anoau E

M M

O Mao (TI oa Vicps C 1 2 E

R B B Il!l ll -

E e e o

_ M t t i

. I f f r T A A P b.

l e)

.hh Ct c e a nnso d ooar p e

di i td p e d e N eaea n u n O bdt o t o I ncd d s d T teee I oml r e e e S nmer b b b O ose P l c f l l l S l ese l l l I iR ar i i i D W( wp W W W t

e . n n r eoe o u gti i s d a b t s nt em a e ausa c r nd on i p oe 8 on f . i s 20 pi id r ta 1 s re o i b n - ee er f s e6 rs vu oe e8 a

/ c s . pb w ne te 7 tL or ss '8 ed eApc e 60 vl bC un ts F i u i i nR to md g o i h  : oed ng ti n ss o

mrse ef av in ti s pd t b r al ed n sp e eo t e ea yues so t d stbb c ka i m r o p ai ve nal un et ot i tl r tw li ri sio ro n p s w ad ei m eo m .

tt g or gtsrg su ae t e u ean h tb t a r gl i es u s goaap r Oo eg i g i pn g

t vn er rpd r ep e R l o eh O ns al v

u s - l se eum l w i i b l a . s l st aB es k ) e8 aes t ui csn p r2t ry sps pd fh eio en eT h i f n ca R cst l oihR c / - ia a gH aA r nh u nd yiR 8 ey gTl orl h n6 pt i( a s v i

t oemo aw f a - i r0 o0 1 el ci nb d

e ens

.e ib ne dt d uesr t e f1 oap rod aoh yu

/i e t st eh dt ma rc et e can l d l nf rr a

N i o o ac aaoee h p O vb f p li a t I o ro pd s ,et m T rf er enn n9 sie A Po Pd Ri o I 2U et D

N E

M M a.

O a. b. c.

C 2 E l. l. l.

R C C C C

r r E o o o e e M i i i t t I r r r f f T P P P A A

~

e e e e e N n n n n n O o o o o o I d d d d d T .

I e e e e e S b b b b b O .

P l l l l l S l l l l l I i i i i i D

W W -

W W W -

s w r e o e r t l t i y n l r u c4t e a arq n8 x r . uee e/e t 'y r oe qpr u1 nb u t g q1 s e c a r 'e t e vl n et eci rnr o o nu pn f ei er c i o oe - e tt l e s ewq an . -- 0 cou ct e uoe n1 nta uer qcg g o - o c db e a n i6 te e d ot i t8 ssb R si at u r c ee e

.o o eL jA vtm r -i m gd t l no0 i nC af e st u ih1 n i vot e dt - o m k yg v2 c ue6 m R o l 6 e l m8 H r ccn . a c e Rf enom vd n d L' r u np h el a cu qsg r nk r e

i nA aC t ou ooe

- a t B eio i rt

-m r d r r r tt m o e ea df np asg asr p l t n o l rtf m l s aeiT oeo gn e a gtS srl oip t t o AncI ie rou D D st n

8 aa w ppt N m N 6h cse d 9 h r A e A i r 2B gga t . o h i v 9 nit se ei t h o &2 o ih s yg gr o Tr p

t sa up b n 8 nl o t a w p 28 c eat ru g- so .a 8 vc oo nd e f2 or f eh et ge o-ml o 0 rt puah 1 e i s1 ua ohtt e0h t rr n - sp su k0Th sop a6 s m oo o1 r

yf e8 et ad t s - m rn rl g gone t

n L pe s

_ pos eA v ocig t8 oit dC e r l n ss/S N l tl i va P nea se2 O eau vh o iuh et

_ I vcs ot mr eo T f c c

- T eoe ri Sf e xf n E oa A Dl r P w Rf Ii ra

_ D N

E

_ M M b. c. a. b.

_ O a.

C 3 3 3 4 4

_ E R C C C C C

t M i I r f A

T P e

std eras t a mn ap a c el i sb p d a e n l c e n iRl i N n o R i f O

I d o* d i sN wi c ,6 T e h ose I e b t t t p S b rs O l

,d o P l l d ef d S l i

etf n I i st ea D W W ui ebt t mes t

n r u an e as re t - r e d pp am s as rr "s n eo tl h n uu t oc s nasl n a ' n o e t .en et ee i

i d , v

s nmo r t ch t e n e n u oei o ant o cp od nomirt f r u t a oi o oi tii t f t l tf it) aud s a , mei o t ctt qd k r vd aarnea c nmh e .

eedd z oer o ort u t d a e iepm o l ma n m N r v euet r el si genr u rrd n e d a et t n nbI o s,i t t ui t o scmsl n or ao l l f

n eea nc y i t occ i ' l .i rreine e d k dl -

i )

prt n

i k v ec nl rwa tsr eo woii o rcR ol o l a el w st o H pl t nit f dR ui v iv re y r i "m amc on wd n g et on t

rl b e ep up a .t i r nt a l 7 nn a . s t s s ot1 eo s s T n n r n di ,n p e i ,

goni r eh s tnmgei eenris mo 0 eh a ni a .

ot l t d os apvat 0 wt ts h l cP n e csp eel c 4 ez sd ay " a e( d raa e ep c mdi rs f o ct n s s ril os3 s een n s ee es a f a irh o ut i y aeh u n o 'n T e n ri d i o t r l t t q s s g wl n nd o l e f oaa oef a t i yf r si nC l n i c o u ed oR e.

l o2 t o l s el of ) rE L N T ai c r ea i vh p o t s apc v h rg we . x n t ec ne uuec y oin s oint eo x h e td rs ee e t et D c rv ai d eisrd n goel l es n each n rn h o p y r nT oua a nh rl i t B eee it ib ueo pc s c y l

l t aed h eh dd c s y n t t k t ei . eor i e y i ed rl vi ea cvmer et o

vt xh n ue h hf os ooot ri E t a s r Wt ot a rrremh a ul ppf l rt g pa i si eonml gt b  :

d a E N st ooaos ni T N O i i c l e O I vdt e v el O I T enc eh on re N T A R aAbT ti T r (

A D '

C N -

I E F M I M 3 R O . A 6

C L E S. .

C C C R

l

I l.

L

\L l

i ATTACHMENT (5) k l

BECo DISPOSITIONING OF  ;

MSIV TEAM RECOMMENDATIONS i

l ll l

Page 1 of ?  !

i l

l l

1i ,

h' t

p if u

wc t nn r cc a ii t tt E S ca M nt8 A o un2 je -

R t F

r nm6 oe8 E o cl pC M i I r nmD T P iip e

N n O o I d T e I e n S b o O d P l S l s I i a D W W 5 2 T f N o E

M 2 H

C e A g T a T P _

A 8 2

6 8

C D y P l b

r m e ye p l s bs s ma e

'V st I se ap S

M - p to d sp l o pt _

i u o b) tl es ii mp rl i i d a l e nt h ae ot _

,d t f

tr yo co l ef bs p mt st er nr sa sp io p a

,e n eR Vo l I bm Sg N

O ma ee Mn i

I sl en T s gi

- A ae nh D se ac N is h a _

E D( Cm

- M 4

0 _

C E .

R _

l l llllllL

i j

l 4

l i

i q

l ATTACHMENT (6) I l

i BECo DISPOSITIONING OF I MODE SWITCH TEAM RECOMMENDATIONS i

'l l

I l Page 1 of 2 I-f L.

l

\ >

F g r n r E o e i

M i r t I r u f T P D A e e e N n n n O o o o I d d d T

I e e e S b b b O

P l l l S l l l I i i i D W W W 6

T N

E l M e H d C o A m e T h e T d r t r A e e o d

n h

t rm o

e r t e a

r u ip m f . ny o ) ot c f s m e il s r i oi ea th

_ - nt t i e d md nf r ao E. er p G t o xs ef et

- d d n e t d e v or nv o t o ae r p p t e me m nR ek .

i e t ae h

cm sa smr y u t ae sot i T t u w r r f ee es h

c we os t me ueh t p( pt t i ms oyn w d r S eu csi t c e i c CS e d

o i mo I I l PCb M l s EP a N e i O e ti l df I c cl l ni T a ua A l d m ta a t.r D p no sS e N e on nPd E R C a I Ri M

M O

C E

R l