ML14188B773

Engineering Evaluation of Hb Robinson RCS Leak on 810129

ML14188B773
Person / Time
Site:	Robinson
Issue date:	03/23/1981
From:	Lanning W NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
To:
Shared Package
ML14188B772	List: ML14188B771 ML14188B773
References
TASK-AE, TASK-C102 AEOD-C102, NUDOCS 8104150060
Download: ML14188B773 (60)

Text

ENGINEERING EVALUATION OF THE H. B. ROBINSON REACTOR COOLANT SYSTEM LEAK ON JANUARY 29, 1981 by the Office for Analysis and Evaluation of Operational Data March 23, 1981 Prepared by: Wayne D. Lanning Lead Reactor Systems Engineer NOTE:

This report documents results of studies completed to date by the Office for Analysis and Evaluation of Operational Data with regard to a particular operating event. The findings and recommendations contained in this report are provided in support of other ongoing NRC activities concerning this event.

Since the studies are ongoing, the report is not necessarily final, and the findings and recommendations do not represent the position or requirements of the responsible program office of the Nuclear Regulatory Commission.

8041:5 0

TABLE OF CONTENTS Page

1. EVENT DESCRIPTION.

2. EVALUATION OF THE EVENT.......

3 2.1 Operator Actions.....

3 2.2 Charging Flow Termination...... **....

4 2.3, Safety Injection Actuation.

5 2.4 Pressurizer Spray..

7 2.5 Relief Valve Bellow Failure*.

7 2.6 Letdown Isolation Valves..

2.7 Leakage Inside Containment.

10 2.8 Drain Valve and Pipe Cap.

10 2.9 Failure of Fire Protection Isolation Valve.......

11

3. CONCLUSIONS.

4. REFERENCES..

o.13 Table Number 1

Sequence of Events.

14 APPENDIX A -

Information Provided by Licensee at Meeting on February 20, 1981

1. Draft Plant Operating Experience Report

2. Operator's Log

3. Shift Foreman Log

4. Strip Charts

5. Figure 1 - CVCS Diagram (excerpt)

6. Figure 2 - Containment Sump Volume

1. EVENT DESCRIPTION A sequence of events is contained in Table 1. Problems with both oil pumps in the turbine electro-hydraulic (E-H) system forced the plant to initiate a plant shutdown. During the process a safety injection signal was generated by a high steam flow coincident with low RCS average temperature. The high steam flow signal was generated by the governor valves spiking open, believed to be caused by the erratic operation of the turbine E-H system. The low average temperature was the result of overcooling the RCS by excessive injection of boric acid solution. The safety injection (SI) signal tripped the reactor.

The reactor power had been reduced from 100% to approximately 6% at the time of trip. The duration of the high steam flow/low average temperature signal was apparently not of sufficient duration to latch the "A" train nor close the main steam line isolation valves.

Both were manually actuated. A containment fire alarm was received shortly after the SI.

After having determined that a spurious SI had occurred, the operators initiated actions (e.g., reset SI, feedwater isolation, restore letdown) to continue to hot standby condition. During the automatic isolation of the CVCS letdown line due to the spurious SI, it is believed that the outermost isolation valves (see Figure 1, valves 204A&B) closed faster than the two open orifice isolation valves (CVC-200B and C), or that leakage past the orifice isolation valves resulted in the opening of the relief valve and the rupturing of the bellows on the relief valve (CVC-RV-203).

In addition, a pressure surge due to the isolation valves closing caused a drain cap to be blown off.

Unaware of these two failures, letdown flow was reestablished. Subsequently, containment pressure and dew point increased.

The containment pressure and humidity increases attached additional significance to the already decreasing

-2 RCS pressure. Letdown was secured (valves closed and sequence unknown) about 15 minutes after letdown was reestablished. A containment entry was made.

A leak was identified in the letdown system area but no fire existed. The heat sensitive fire alarm detected the steam from the leak in the letdown system, which implies that this leak occurred in the CVCS during the first SI.

Approximately 3,000 gallons was estimated to be in the containment sump based on level indication in the control room.

After the letdown was thought to be isolated, the pressurizer pressure continued to decrease and the level to increase. A second safety injection occurred on low pressurizer pressure. Both trains of safeguards equipment actuated.

The level increase was the result of continued charging flow and heatup of the primary system.(the MSIVs had been closed to recover average temperature earlier). The cause for the depressurization could not be identified positively.

Four hours after the first entry, a second containment entry was made and the leak was identified to be from a drain line which was still leaking.

The drain line is located upstream of the orifice isolation valves (see Figure 1).

The cap on the drain pipe was missing and valve (CVC-200E) was manually closed. Water in the containment sump had now increased to approximately 4,500-6,000 gallons. Evidently, the two level control valves (CVC-LCV-460A&B) were leaking at five to seven gallons per minute between 0650 and 1120. After the drain valve was closed during the second containment entry, the RCS pressure continued to decrease.

Many steps were taken to determine the cause of the decreasing RCS pressure after letdown had been isolated; e.g., isolating charging line auxiliary spray, checking pressurizer relief and safety valve leakage, and increasing pressurizer heater output. The cause was identified when the operators

-3 stopped two of the three reactor coolant pumps in the loops with the pressurizer spray scoops and the pressure began to increase. One of the two pressurizer spray valves was not fully closed. Positive identification of spray valve RC-455B as the leaking valve was made later. The spray valve position is indicated by demand, not stem position, which delayed identification of the cause for depressurization.

During this event, steam generator samples indicated a primary-to-secondary leak of approximately 0.5 gpm based on activity of 10-4PC/ml. Steam generator "B" was isolated on the secondary side. Subsequent samples indicated decreasing activity and no leak. The licensee has concluded that the increased activity was the result of "crud" being agitated during isolation of the steam generators during the event.

Repairs were made to the spray valve and the relief valve bellows. The cap was replaced on the drain line and all drain valves were verified closed.

The unit was back online on February 1,1981.

2. EVALUATION OF THE EVENT 2.1 Operator Actions Operators responded to the events in a systematic and timely fashion.

Data entered into the logs were detailed and accurate.

After the plant was stabilized, the license contacted Westinghouse to ensure that their diagnoses were correct and no other unforeseen problems existed.

One shortcoming identified was the lack of a procedure for recovery from a spurious safety injection actuation. Guidelines should be available to the operators to differentiate between a real and spurious SI actuation. The licensee indicated that a procedure will be written for recovery from a spurious SI (identification criteria not included).

For this event, resetting the SI

-4 had no consequences. However, pressurizer pressure, level and average temperature had all been decreasing prior to the SI and had "stablized" for only about two minutes before resetting SI.

In retrospect, there was still a small reactor coolant leak and the spray valve was open. However, SI had been initiated on signals indicative of a steam line break and since secondary system conditions were stable and the governor valve position recorder indicated spurious valve opening, the operators correctly diagnosed the SI signal as spurious for this event.

An area of improvement would have been to test the safety injection actuation and main streamline isolation signals since one train of SI failed to latch and the MSIVs failed to close. Both were manually actuated. Although both SI trains actuated on the second safety injection signal, this was not adequate verification of operability on high steam flow/low average temperature actuation before returning to power. These tests could have helped substantiate that the signal was not of sufficient duration to latch the SI relay and close the MSIVs.

2.2 Charging Flow Termination Although SI actuation occurred twice, no boric acid was injected into the RCS based on samples of the boric acid injection tank. This was because the RCS system pressure exceeded the shutoff head (1,500 psig) of the SI pumps at the times of actuation. Hence, the charging pumps were making up lost RCS inventory during the event.

During the attempts to identify the cause for the depressurization and recognizing that pressurizer spray could cause the depressurization, the charging flow was isolated (closed valve CVC-HCV-121) to terminate a possible leak from the auxiliary spray valve (Figure 1).

This operator action did not terminate

-5 all makeup flow to the RCS. The flow path was maintained to the RCP seals which would provide makeup flow (approximately 60 gpm).

RCS conditions (approximately) at this time were:

Pressure = 1,600 psig; Tavg = 548*F; pressurizer level = 56% and increasing; normal steam generator level for the condition; and margin to saturation was approximately 50-55*F.

The charging line was isolated from 0726 to sometime after 1932 (shift foreman's log).

No consequences resulted from isolating the normal charging flow for this event although SI flow was not available due to the pump head limits.

However, it is suggested that NRR determine whether isolating the charging flow is advisable for small loss-of-coolant accidents or when the system pressure is above the shutoff head of the SI pumps. Westinghouse has indicated that no credit was taken for charging flow for the ECCS analyses. The emergency procedure for depressurization (EI-1) does not include criteria for terminating charging flow. The charging pumps are a part of the CVCS and not considered a part of the safety injection system at Robinson. However, the charging pumps provide high pressure makeup flow when the RCS pressure exceeds the shutoff head of the SI pumps. Ensuring that charging flow is not interrupted for the systems employing low/medium heat SI pumps may be desirable to enhance safety.

2.3 Safety Injection Actuation The first safety injection actuation occurred on a "high steam line flow/low Tavg" signal.

The licensee's review of the event indicated that the momentary spike-opening of the turbine governor valves caused the steam flow, in at least two steam lines, to exceed the steam flow set point for a period of about 25 msec. The combination of high steam flow in 2/3 steam lines and the existing low average temperature of the reactor coolant generated a main steam isolation valve (MSIV) closure signal and a SI actuation signal.

-6 However, only train B of safeguards equipment responded - the other train of safeguards equipment and all the MSIVs did not actuate. Licensee's observations are that the MSIVs require a signal duration of one second to close and that the SI actuation relays, including SI logic train latching relay, require a signal duration greater than 25 msec to actuate. Since the SI signal was of less than 25 msec duration, only the train B latching relay actuated. Reactor trip, emergency diesel start, feedwater isolation and other safeguards equipment actuations for train B occurred as a consequence of SI train B actuation.

Reviewing the logic diagram of.Robinson's Safeguard Actuation Signals (Dwg CP 300-5379-2759 sh 8, rev 6) it is seen that the reactor trip signal is initiated on SI actuation along with emergency diesel start, feedwater isolation and safeguards sequence actuation.

A review of a later Westinghouse logic diagram (typical) shows that the reactor trip signal is derived separately from the SI actuation signal; i.e., the reactor trip signal is taken off "upstream" of the SI actuation signal, similar to the MSIV closure signal on Robinson.

This could mean that on certain spurious SI actuation events of short signal duration, SI, feedwater isolation and auxiliary feedwater system actuations may occur with no simultaneous reactor trip occurring.

The comparison of logic diagrams also shows that the P-4 interlock (reactor trip breaker position) in the the Reset/Block feature of SI logic of later Westinghouse units is not provided in the Robinson design.

Additional analyses would be needed to ascertain the significance of different reactor trip looics for Westinghouse plants.

The need to provide a direct reactor trip on spurious safety injection actuation is referred to NRR for review.

-7 2.4 Pressurizer Spray The open spray valve could not be identified due to lack of spray flow indication or actual spray valve position. The failure of the valve to close evidently did not affect the capability of the valve to open as evidenced by subsequent testing. The Licensee is evaluating the possibility of relocating and replacing the spray valves during the next refueling outage. Previous problems have been experienced with the spray valves and their location in containment reduces their accessibility for maintenance.

2.5 Relief Valve Bellows Failure The licensee has experienced previous failures of this Crosby relief valve (number JB-36, Type B, shop drawing number H51380). Basic information about the valve and the discharge piping configuration were obtained from CP&L and Crosby Valve Company and are as follows:

Relief Valve 2" diameter inlet, 3" diameter outlet Set pressure, 600 psig System pressure, 300 psiq (approximate)

Dynamic backpressure, 25 psig (specified)

Bellows tested to 150 psig Piping A horizontal run exits from the relief valve before turning vertically up for at least 12 feet to the pressurizer relief tank.

-8 CP&L indicated that the bellows fails every time the relief valve lifts.

Since the bellows has been tested to 150 psig, it would appear that the system is operating quite differently from the anticipated mode. The dynamic backpressure probably exceeds 150 psig (six times the specified 25 psig).

A mechanism that could cause the high pressure might be stagnate water from either steam condensation or valve leakage in the line from the relief valve to the pressurizer relief tank. Boric acid crystal formations may also be a possibility. When the valve opens, water or other debris in this line could restrict steam flow and cause a high dynamic backpressure until the line is cleared. Also, if the line is filled with stagnate borated water, the bellows may be susceptible to corrosion attack, but.corrosion has not been identified from previous failures and replacements. From an operational viewpoint, the failure mode for the bellows should be identified and changes necessary to prevent additional failures should be implemented. The operation of the CVCS isolation-valves may be a major contributor to the bellows failures and is discussed in Section 2.6.

2.6 Letdown Isolation Valves The isolation valves played a dominant role in the sequence of events at Robinson. The failure of the bellows on the relief valve was attributed to the closing of the out board valves (CVC-204A&B) before the closing of the orifice isolation valves (CVC 200B&C) upstream of the relief valve.

Consequently, the set point (600 psig) of the relief valve was reached since this part of the CVCS was pressurized by the reactor coolant system which was at approximately.1,800 psi.

The design pressure downstream of the valves (CVC-200 series) is 600 psig. The sequential operation of the isolation valves is evidently causing this part of the CVCS to be pressurized to at least the setpoint of the relief valve, as evidenced by the opening of the relief valve whenever the CVCS is isolated.

-9.

In addition to the isolation valves, valves LCV-460 A and B (Figure 1) were closed in an attempt to isolate the leaking drain valve/pipe. Both of these valves leaked which permitted an additional 3,000 gallons (approximately) to leak into the containment after the letdown system was thought to be isolated.

The licensee did not perform any maintenance on these valves to ensure their operation before returning to power since these are not containment isolation valves.

These valves are part of the reactor coolant pressure boundary and are designed to close on low pressurizer level to conserve RCS inventory.

The design and operation of this part of the CVCS raises two concerns:

first, the potential for overpressurizing the system to 2,200 psig assuming the downstream isolation valves (CVC-204A&B) are closed; and secondly, the capability to isolate a potential break downstream of valves LCV-460A&B.

The licensee has indicated that the relief valve is designed to prevent overpressurization of the CVCS. The failure of the bellows does not appear to affect the pressure relieving function of the relief valve. In addition, the flow control valves (CVC-LCV-460A&B) have been designed to isolate a break downstream of these valves for the maximum size break and RCS conditions.

The functional and testing requirements for the flow control valves are not clear. These valves should be ASME Class 1 since there are no valves upstream and the valves downstream are classified as ASME Class 2. However, these flow control valves are not identified in the Robinson Inservice Inspection and Testing Program (Reference 4).

Since these valves are on the RCS pressure boundary and are designed to isolate the RCS on low pressurizer level, it is not clear why maintenance on the valves was not required after they were known to leak and before returning to power.

10 Both of these concerns could lead to a small loss-of-coolant event inside containment. This postulated event is within the scope of an analyzed small break loss-of-coolant accident and not a new safety concern. However, from an operational consideration, overpressurizing the CVCS could be prevented, provided the orifice isolation valves were closed before the outboard isolation valves.

Correcting the valve closing sequence for isolation would also reduce the challenge to the relief valve.

2.7 Leakage Inside Containment The licensee has acknowledged that the quantity of water that leaked into containment can only he approximated. The estimated 6,000 gallons (corresponding to approximately 15" in the sump) is a small fraction of the range of indication in a 65,000-gallon capacity sump (See Figure 2). A mass balance was not possible since neither charging flow nor volume control tank level are recorded.

The major leak was after letdown flow had been reestablished between 0635 and 0650. This could account for approximately one half of the 3,000 gallons indicated at 0650. The drain valve could have also been leaking at an unknown reduced rate from the initial SI until letdown was restored (approximately ten minutes). The ruptured bellows on the relief valve also contributed some amount to the inventory in the sump. These sources in combination with the inaccuracy of the sump measurements can lead to the conclusion that all the leak sources had been identified.

2.A Drain Valve and Pipe Cap The leaking valve was CVC-200E (see Figure 1) not CVC-204C as reported by IE (Reference 1). This helps to understand the leak rates and quantity of water reported in the LER (Reference 2) and the IE evaluation.

The licensee's explanation for the missing cap on the pipe was that when

0 0

the orifice isolation valves closed, a pressure pulse was applied to the valve and cap. Since the valve was partially open and the cap not tightly secured, the cap was blown off. The licensee believed that vibration in the CVCS (induced by the charging pumps) caused movement of the valve and cap. The valve position was last verified on October 11, 1980 during a refueling outage. Since the drain pipe is located close to the pressure reducing orifices, the flow instabilities at these orifices could also induce vibration in the CVCS.

All drain pipes with valves have been verified closed. Most valves have been chained and locked.

2.9 Failure of Fire Protection Isolation Valve When a Phase A isolation signal was generated by the safety injection actuation, one (FP-248) of the four containment isolation valves failed to close due to a tripped breaker. Since the other isolation valve in the line closed, containment isolation was achieved. This failure had no bearing on the leak and was.a separate reportable event.

3. CONCLUSIONS The event at H. B. Robinson involved four separate, somewhat unrelated failures:

(1) pump failures in the turbine EHC system; (2) two separate leaks in the CVCS (related failures); (3) an undetected open pressurizer spray valve; and (4) leaking valves in the CVCS. The event did not appear to include any safety concerns.

The following areas of review concering this event are referred to NRR for consideration:

12

a. Whether a requirement should be placed upon operating plants to establish a procedure for identification and recovery from a spurious safety injection actuation (if such a procedure is not already in place).

b. Whether criteria for terminating SI should include provisions for isolating charging since charging flow could be considered high pressure safety injection for very small breaks.

c. Whether there is a need for a direct reactor trip on a spurious safety injection actuation at other Westinghouse plants which do not have a direct trip.

d. Whether operation of the isolation valves in the CVCS at Robinson is causing the system to be operated in a manner which is contrary to its design bases. The closing sequence for the isolation valves appears to cause part of the CVCS to be pressurized to the setpoint of the relief valve and may be contributing to the failure of the relief valve bellows whenever the system is isolated.

AEOD did not find any basis for a need to study this event further. A formal response from NRR is not requested.

This event and the operator's response provide a good example of an operating experience which should be disseminated to other licensees for information and training purposes.

13

4. REFERENCES (1) Memorandum, H. Woods to E. Jordan,

Subject:

H.B. Robinson Event on January 29, 1981, dated February 12, 1981.

(2) Licensee Event Report 81-005, H.B. Robinson Steam Electric Plant, Unit 2, Docket 50-261, dated February 12, 1981.

(3) Meeting with Carolina Power and Light Company in Bethesda on February 20, 1981.

(4) Letter, E. E. Utley, CP&L to S. Varga,

Subject:

H. B. Robinson Steam Electric Plant Unit No. 2, Inservice Inspection and Testing Program, dated March 10, 1981.

0 Table 1 SEOUENCE OF EVENTS January 29, 1981 Plant at 100%

Primary to secondary leak of approximately 0.3 gpm.

0500 "A" EHC oil pump seal leak, "B" EHC pump already out of service due to vibration.

0541 Started load reduction.

0542 Added boric acid to RCS.

0543 Started "C" charging pump, "B" charoing pump running, "A" charging pump inoperable.

Opened CVC-200B orifice isolation valve, CVC-200C already open.

0549 0609 Continued to add boric acid.

0613 Stopped "1" feedwater pump and condensate pump due to erratic FWP hehavior.

0620 Tavq reached low Tavq setpoint (543 0F) alarm.

0623 Generator output breaker opened.

Turbine governor valves spike open.

SI signal and MSIV closure signal on high steam flow/low Tavq.

SI train "B" automatically started.

Phase A isolation; safeguard B emergency equipment started.

Reactor trip on SI signal.

Tavg = 532 0F.

Pzr pressure = 2210 psig.

Pzr level 13%.

0625 Fire alarm in containment.

Pressurizer relief tank level alarm due to opening of CVC-RV-203 relief valve.

Bellows probably ruptured and drain cap was blown off.

MSIVs closed manually.

SI train "A" started manually. Started "A" DG, AFWP, RHR, manually.

Letdown valves CCV-460A&B manually closed (should have automatically closed on PZR level of 13%).

0627 Reset SI and feedwater isolation.

0634 Attempted to restore letdown flow hut CVC-200A would not open (instrument air system isolated on Phase A isolation).

Restored letdown flow after resetting isolation signals.

Pressurizer pressure started decreasing sharply (~2000 psig).

Containment dew point and pressure started increasing.

0637 Received condensate collection alarm from the coolers.

Diesel generators A and B stopped manually.

15 0645 Isolated'letdown flow. (Isolation valves closed from control room.)

Containment dew point and pressure decreased.

Pressurizer pressure still decreasing ( 1840 psig).

Tavg increasing.

Pressurizer pressure increasing.

Notified NRC by ENS.

0650 Containment sump level indicated approximately 3000 gallons.

0700 First containment entry to check for leak and fire.

0705 Second SI actuation on low pressurizer pressure.

Both trains and all equipment started.

Pressurizer pressure = 1715 psig.

Pressurizer level = 50%.

0705 0727 Operators attempting to determine cause of depressurization.

0722 Steam dumps opened manually to control pressurizer level.

0727 Reactor coolant pumps B and C stopped and charging line isolated to eliminate possibility of leaking auxiliary spray.valves.

Increased pressurizer heater output to maximum.

Pressurizer pressure started increasing.

0729 Continued cooldown using steam dumps.

0735 Pressurizer pressure increasing (= 1720).

Tavg constant = 540.

Pressurizer level = 50%.

0738 Stopped diesel generators A&B.

0741 Stopped "B" RHR pump.

0745 Opened breakers on containment sump pumps.

0825 Secured SI pumps.

1000 Continued plant cooldown.

Sample on "B" steam generator indicated 0.5 gpm primary to secondary leak.

Isolated "B" steam aenerator.

Second sample showed decreased leakage (0.25 gpm).

1120 Second containment entry. Found CVC-200E open and cap missing.

Found bellows on relief valve CVC-203 ruptured.

Contacted Westinghouse.

1218 Blocked low pressure SI.

1230 Closed CVC-200E.

Isolated letdown by closing CVC-309D.

Containment sump level was 4,500-6,000 gallons.

1445 "B" charging pump out of service due to leaking relief valve 1830 Aligned "A" charging pump for operation after completing surveillance tests.

(late entry) Tested pressurizer spray valves.

1913 Started "B" RCP.

1932 Started "C" RCP.

(Later) Placed charging line and CVCS letdown in service. Removed excess letdown line from service.

2315 Spray valve RCS-455B identified as leaking spray valve No additional primary to secondary leak identified.

January 30, 1981 at 1700.plant on-line

APPENDIX A INFORMATION PROVIDED BY LICENSEE AT MEETING ON FEBRUARY 20, 1981 Contents:

1. Draft Plant Operating Experience Report

2. Operators Log

3. Shift Foreman Log

4. Strip Charts S. Figure 1 - CVCS Diagram (excerpt)

6. Figure 2 - Containment Sump Volume

-DRAFT P.ANT O.IIAIING EXE1 M

)NCE, E1POIT 1.Event Date January 29, 1981

2.

Identification of Occurrence A)

A spurious safety injection signal initiated by a "High Steam Line Flow/Low T

" signal.

avg B)

Reactor Coolant System leak through letdown line drain valve CVC-200E.

C)

Primary plant depressurization leading to a second safety injection signal initiated by a "Low Pressurizer Pressure" signal.

3.

Conditions Prior to Occurrence A plant shutdown to hot standby was in progress to repair a secondary plant problem. The unit had been operating at 100% reactor power (725 M e) with normal Reactor Coolant System pressure and temperature.

4.

Description of Occurrence (All Times Are Approximate)

A)

At 0624 hours0.00722 days <br />0.173 hours <br />0.00103 weeks <br />2.37432e-4 months <br /> on January 29, 1981, a safety injection signal initiated

"," train of safeguards. "A" train equipment was manually started at 0625 hours0.00723 days <br />0.174 hours <br />0.00103 weeks <br />2.378125e-4 months <br />.

B)

At 0635 hours0.00735 days <br />0.176 hours <br />0.00105 weeks <br />2.416175e-4 months <br /> on January 29, 1981, the chemical and volume control letdown system was restored and system pressure began decreasing with an increasing containment pressure end dew point. Letdown was secured at 0650 hours0.00752 days <br />0.181 hours <br />0.00107 weeks <br />2.47325e-4 months <br />.

C)

At 0705 hours0.00816 days <br />0.196 hours <br />0.00117 weeks <br />2.682525e-4 months <br /> on January 29, 1981, a safety injection signal initiated both trains of safeguards.

-DRAFT to1

-D R A FT

5. Designation of Apparent Cause of Occurrence At approximately 0400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br />, "A" turbine electro hydraulic (E-H) oil pump developed a seal leak.

"B" E-H oil pump had been taken out of service earlier due to high vibrations.

At 0541 hours0.00626 days <br />0.15 hours <br />8.945106e-4 weeks <br />2.058505e-4 months <br />, the decision was made to shut down to hot standby before receiving a trip signal due to the loss of E-H oil. Attachment No. 1 contains additional information on the failure of the E-H Oil System.

At 0624 hours0.00722 days <br />0.173 hours <br />0.00103 weeks <br />2.37432e-4 months <br />, immediately following opening the generator output breakers, the reactor tripped and a safety injectior was initiated by a "High Steam Line Flow/Low T

" signal. Only "B" train of.the safeguards was activated.

"A" train equipment was manually started at 0625 hours0.00723 days <br />0.174 hours <br />0.00103 weeks <br />2.378125e-4 months <br />.

It was determined that the erratic operation of the E-H Oil System and the fact that the operators were switching from "A" E-H oil pump to "B" E-H oil pump caused the governor valves to spike open. The resultant steam flow spike was high enough to cause a "High Steam Line Flow/Low T "g signal but it was of insufficient duration to fully latch.the "A" safeguards train seal-in relay. The seal-in relays in the safeguard trains are latching relays that require a finite period of time in the energized mode to mechanically latch them into the closed position. Attachment No. 2 contains additional information on the partial safety injection.

The steam line isolation signal that was generated from the "High Steam Line Flow/Low T signal was of insufficient duration to allow the main steam avg isolation valves to go shut. The open signal was reinstated so quickly

-DRAFT tm 2

-DRAFT

5.

Designation of Apparent Cause of Occurrence (Continued) after the isolation signal that the valves were unable to travel far enough to isolate the steam flow. The main steam isolation valves were manually shut to reduce the secondary steam demand following the reactor trip, thereby promoting the.return of T to the no load setpoint.

avg At 0627 hours0.00726 days <br />0.174 hours <br />0.00104 weeks <br />2.385735e-4 months <br /> it was determined that safety injection conditions did not exist and that the initiation was spurious. The safety injection and feedwater isolation signals were reset. The chemical and volume control letdown system was restored at 0635 hours0.00735 days <br />0.176 hours <br />0.00105 weeks <br />2.416175e-4 months <br />. The Reactor Coolant System pressure had been slowly decreasing, but when letdown was returned to service, the containment pressure and dew point began increasing. Another indication of abnormal containment conditions was a fire alarm from the area of the containment operating deck which was received at approximately 0624. Letdown was secured at 0650 hours0.00752 days <br />0.181 hours <br />0.00107 weeks <br />2.47325e-4 months <br /> with Reactor Coolant System pressure at 1850 psig. The initial containment entry made at 0700 hours0.0081 days <br />0.194 hours <br />0.00116 weeks <br />2.6635e-4 months <br /> to investigate the abnormal conditions confirmed that the RCS leakage was from the letdown line and that no fire existed. A subsequent containment entry at 1120 hours0.013 days <br />0.311 hours <br />0.00185 weeks <br />4.2616e-4 months <br /> further identified-the source of the leak as valve CVC-200E, a drain valve on the letdown line, which was found open and the pipe cap missing. The leak that resulted from the open drain valve was approximately 5 to 7 gpm with the letdown air operated valves closed and approximately 100 gpm with letdown flow established. The leak was com pletely stopped by shutting valve CVC-200E. The letdown flow was not restored until after the condition was found and repaired. Additional information regarding the RCS leak and containment fire alarm can be found in Attachment No. 3.

tm

-DRAFT-

0-DRAFT-0

5.

Designation of Apparent Cause of Occurrence (Continued)

However, even with the letdown control valves closed, the pressurizer pressure continued to decrease, leading to the second safety injection initiation at 0705 hours0.00816 days <br />0.196 hours <br />0.00117 weeks <br />2.682525e-4 months <br /> from a "Low Pressurizer Pressure". Both trains of the safeguards equipment functioned as designed. At 0727 hours0.00841 days <br />0.202 hours <br />0.0012 weeks <br />2.766235e-4 months <br />, charging was isolated (except reactor coolant pump.seal injection) to eliminate auxiliary spray and "B" and "C" reactor coolant pumps were secured to prevent the pressurizer spray valves from circulating cooler water from the Reactor Coolant System into the pressurizer through the spray valves, decreasing the pressure. It was subsequently discovered that the pressurizer spray valve from "C" reactor coolant loop had prob ably opened and not fully reseated. The pressurizer pressure immediately started to increase. The reactor coolant system was stabilized at approximately 2050 psig and 5350F with pressure controlled by the pres surizer heaters and temperature controlled by the secondary steam dump.

Attachment No. 4 contains additional information on the reactor coolant system pressure transient caused by the spray valve malfunction.

Coincidental with the decreasing pressurizer pressure, pressurizer level was increasing. This was caused by two factors. 1) The charging flow from two charging pumps was maintaining or increasing the system volume, including the system losses through CVC-200E. The slightly open pressurizer spray valve was causing the pressure to decrease. 2) The density changes in the reactor coolant due to the slowly increasing RCS temperatures and the heat up of the relatively cold water added by the charging system caused the system to expand. These factors combined to cause an increasing pressurizer level. The margin to subcooling remained tm 4

-DRAFT-

5.

Designation of Apparent Cause of occurrence (Continued) greater than 550F throughout the entire transient. The minimum subcooling margin occurred at 0720 hours0.00833 days <br />0.2 hours <br />0.00119 weeks <br />2.7396e-4 months <br />, with reactor coolant system pressure at 1620 psig and temperature at 551o The relief valve on the letdown line, CVC-RV-203, lifted following the

-first safety injection initiation. This was apparently due to the isola tion valves, CVC-204A and CVC-204B, closing slightly faster than the orifice isolations, CVC-200A, CVC-200B and CVC-200C, or leakage past one or more of the orifice isolation valves. This caused the pressure between the valves.to increase above the set pressure for CVC-RV-203 (600 psig).

The valve rcset after the letdown isolations closed, but the bellows had ruptured. Attachment No. 3 also contains additional information regarding valve CVC-RV-203.

6.

Analysis of Occurrence Several problems with the turbine E-H Oil System had occurred within approxi mately one week preceeding the reactor trip and safety injection on January 29, 1981 which could have contributed to the initiation of the event.

These problems are summarized as follows:

1)

The E-H oil had become contaminated with water due to a ruptured E-H oil cooler approximately one week prior to this event. However, the E-H.oil had been purified (replaced) and restored to specificaticn prior to this event. It is not felt that this contributed to the following problems.

2)

On January 28, 1981 "B" E-H pump unloader developed a fatigue crack in its discharge nipple.

While replacing this nipple, air was introduced into the "B" E-H oil pump portion of the system.

When

-DRAFT to 5

-DRAFT

6. Analysis of qqcurrence (Continued)

"B" E-H oil pump was restarted, it caused excessive vibrations throughout the E-H Oil System. "A" E-H oil pump was restarted and "B" E-H oil pump was secured after a brief period of operation.

3)

The seal leak which developed on "A" E-H oil pump on January 29, 1981 which necessitated the turbine shutdown is felt to have been caused by either age or the excessive system vibration.

4)

As the seal leak on "A" E-H oil pump became larger during the re maining moments of the turbine shutdown, the operators decided to run "B" E-H oil pump despite the vibration problem in order to allow the leak to be isolated so a normal turbine shutdown could be completed.

Coincidentally, "B" E-H oil pump was started as the generator output breakers were opened.

When the generator output breakers are opened the turbine switches from load control to speed control.

One, or some combination, of the above probably caused the turbine governor valves to spike open. The exact cause cannot be determined. This caused the first safety injection initiated on a low reactor coolant system average temperature coincident with high steam line flow. The high steam flow was of a very short duration, thus only "B" safeguards train was activated and the main steam isolation valves remained open.

Letdown line drain valve CVC-200E had vibrated open since it had last been verified shut on October 11, 1980.

It is postulated that the pressure transient caused by the letdown line isolation caused the pipe cap to blow off. Thus, a Reactor Coolant System leak existed.

-DRAFT tm 6

-DRAFT-

6. Analysis of Occurrence (Continued)

The continued decrease in pressurizer pressure was caused by the failure of the pressurizer spray valve from "C" reactor coolant system loop (RCS-455B) to fully shut after opening during the transient. The event identification was complicated by the letdown relief line lifting to the pressurizer relief tank which indicated that there were two separate leaks.

The Reactor Coolant System pressure decrease was stopped when "B" and "C" reactor coolant pumps were secured and the charging line was isolated to eliminate auxiliary spray. With the pressure decrease stopped, operator control of the Reactor Coolant System was re-established and normal hot shutdown conditions were established.

Following the first safety injection at 0624 hours0.00722 days <br />0.173 hours <br />0.00103 weeks <br />2.37432e-4 months <br />, the fire protection containment isolation valve FP-248 did not shut automatically and had to be manually closed. Attachment No. 5 contains additional information on the performance of the fire protection containment isolation valve.

A summary of the P250 computer output for this event is provided as Attachment No. 6.

7. Corrective Action A)

The E-H oil was completely replaced with new oil.

B)

"A" E-H oil pump and unloader were replaced.

C)

The unloader and discharge nipple on "B" E-H oil pump were replaced.

D)

The valve stem on RCS-455B was lubricated, stroked and valve posi tioner was adjusted to ensure the valve will fully close. RCS-455A was also checked for proper operation.

-DRAFT tm 7

-DRAFT

7.

Corrective Action (Continued)

E)

CVC-200E was locked closed and the pipe cap was replaced. Similar valves in the letdown and charging lines were also lpcked closed or otherwise verified to be secured.

7)

The breaker over current trip setpoints on the four Fire Protection System containment isolation valves have been adjusted and checked to insure proper valve'performance.

G)

The event was fully analyzed by the plant staff and Westinghouse, and the results discussed with the NRC, Region II, to ensure that all safety concerns were identified and resolved prior to returning the unit to operation.

Unit 2 Operat Sfp Isr Manager -

perati Dn and Maintenance General. Anager

-DRAFT tm 8

W-DRAFT SEQUENCE OF EVENTS 0541 Unit shutdown was initiated due to E-H System trouble.

0620 Tavg reached the.low Tavg setpoint (5430F) during plant shutdown.

0624 Generator output breaker is opened removing unit from system.

Load on unit is "\\6%.

Turbine governor valve(s) spike open (see Attachment No. 1).

High Steam Flow/Low Tavg signal generated.

MSIV's closure signal (see Attachment No. 2).

SI signal, train "B" actuates (see Attachment No..2).

CV isolation valve FP-248 fails to close (see Attachment No. 5).

Minimum Tavg 532 0F (based on incore thermocouple).

PZR pressure 2100 poig.

PZR level = 13Z.

0625 Fire alarm at CV operating deck (see Attachment No. 3).

Pressurizer relief tank level alarms from CVC-203 discharge (see Attachment No. 3).

-DRAFT-

0625 (Contd.)

Primary pressure begins to decrease (see Attachment No. 4).

MSIVs manually closed.

SI train "A" equipment manually started.

Letdown valves 460A & B manually shut.

0627 Manually reset SI.

0635 Restored letdown.

Containment dew point and pressure begin to increase.

0650 Isolated letdown (suspected leak in letdown system).

0656 Tavg reaches maximum value of 552 0F and holds steady.

PZR pressure 1750 psig.

PZR level 50%.

0700 Containment entry to check for leak and fire (see Attachment No. 3).

0705 Second SI signal due to low PZR pressure, 1715 psig.

Both "A" and "B" trains activated.

0705-0727 Operators attempt to determine cause of depressurization. The following equipment was checked:

-DRAFT-

-DRAFT-.

0705-0727 a)

PZR safety valves flow indicators.

(Contd.)

b)

PZR PORV discharge line temperature.

c)

PZR block valve position.

d)

PZR relief tank level.

e)

PZR relief tank pressure.

f)

PZR spray valve position (the valves indicated closed but since this indication is demand indication the valve controllers were-again manually closed).

0722 The RCS.temperature was lowered slightly using the secondary steam dumps to help control the increasing pressurizer level.

Tavg 549 0F.

PZR pressure 1620 psig.

PZR level 62%.

0727 The charging line was isolated to eliminate the possibility of auxiliary spray causing the depressurization. RCP "B" and "C" were stopped to eliminate the possibility of main spray flow causing the depressurization.

Pressurizer pressure begins to rise.

0735 Tavg 543 0F.

PZR pressure 1715 psig.

PZU level. =-A.

-DRAFT-

0820 PZR pressure stabilized.

Tavg 535o0.

PZR pressure 2050 psig.

PZR level 45%.

1120 Made second containment entry and isolated CVC-200E at 1230 hours0.0142 days <br />0.342 hours <br />0.00203 weeks <br />4.68015e-4 months <br />.

1120 (1-29-81) to 1700 (2-1-81)

Review and analysis of transient with Westinghouse.

Din assions of trannient with NRC Region 11.

2315 (1-29-81) RCS-455B positively identified as. leaking spray valve.

1700 (2-1-81)

Plant on-line.

-DRAFT-

-DRAFT ATTACHMENT NO.

1 E-H SYSTEM FAILURE The E-H System had experienced several problems prior to the transient on 1-29-81.

During the previous week the E-H fluid had become contaminated with water.

(This contamination was restored to within specification.)

On Wednes day morning, 1-28-81, a stainless steel nipple on the E-H System unloader on "B" pump cracked.

This caused a loss of approximately 30 gallons of E-H fluid.

The fluid and nipple were replaced and "B" pump.restarted.

However, the pump was immediately stopped due to noise and vibration.

Several attempts were made to troubleshoot the problem but no definite cause was found. The system was left operating satisfactorily with one pump in service. At 0500 on 1-29-81 the second E-H pump, "A", developed a seal leak which caused E-H fluid to leak out of the system. At 0541 the operators began to take the unit off line to repair the E-H System. At 0624 while the unit was being separated from the system, the E-H System generated a pressure surge to the governor valves which resulted in the valves momentarily opening. Three factors could have contributed to the pressure surge. The turbine control was switching to speed control.

The opera tors were trying to start "B" E-H oil pump to supply E-H oil during the final moments of the turbine shutdown. The E-H System had been contaminated by water during the previous week.

This caused a momentary high steam flow to be sensed on.at least 2 steam lines. The spike shows up on all three steam flow charts.

The effect of this flow spike is described in Attachment No. 2.

The failure of "A" pump seal on the E-11 System was due to age and transferred vibration from "B" pump. Subsequent to these pump failures, the unloader of pump "B" has been replaced and pump "A" was replaced in its entirety. The complete system was restored to service and is operating satisfactorily.

-DRAFT tm 1

-DRAFT ATTACHMENT NO.

2 PARTIAL SI AT 0624 HOURS On January 29, 1981 at 0541 a unit shutdown was commenced to do repair work on the turbine E-H System. At approximately 0620 hours0.00718 days <br />0.172 hours <br />0.00103 weeks <br />2.3591e-4 months <br /> Tavg dropped below the low Tavg setpoint of 543 F due to an inadvertent overshoot during plant shutdown.

At 0624 with the unit ateu6% power the generator output breakers were opened disconnecting the unit from the system. At this time the turbine E-H control system switched to speed control and due to pressure instabilities in the E-H control system the turbine governor valves spiked open. A review of the event indicates that the spike caused an indicated steam flow in at least two steam lines to exceed the steam flow setpoint for a time period less than 25 msec.

This indicated high steam flow in 2/3 steam lines combined with the low Tavg mentioned earlier generated a.main steam isolation valve closure signal and a SI signal. The duration of this signal would be the same as the steam flow spike. It has been observed during periodic tests that the MSIVs require a signal duration of approximately 1 sec. to close and so none of the MSIVs closed on the momentary high flow/low Tavg signal.

(The MSIVs were manually closed immediately by the operators in order to stabilize RCS temperature.)

The SI signal is divided into 2 trains "A" and "B".

Each of these trains con tains several relays including a mechanical latching relay (Westinghouse Type MG6) which is used to lock in the SI train until manually reset. A signal duration greater than 25 msec. is required to insure that all relays close and the latching relays lock in. Since the SI signal was less than 25 msec. only the latching relay for train "B" fully engaged. The operators immediately noticed that train "A" had not engaged and so they manually started the train "A" equipment. Containment isolation Phase A was initiated by train "B".

No SI water was injected into the system since RCS pressure wasv2100 psig and the

-DRAFT to1

-DRAFT ATTACHMENT NO. 2 (Continued) shut off head of the SI pumps is 1500 psig.

SI was manually reset at 0627 since the SI initiation was identified as spurious.

Once train "A" was manually initiated the SI System performed as expected, with the exception of CV isolation valve FP-248 (see Attachment No. 5).

The actua tion of the SI System did not effect the physical course of events during the transient, however it did obscure the cause of the RCS depressurization (stuck pressurizer spray valve). No repairs to the SI logic or components are considered necessary.

-DRAFT to 2

TTA AENT NO.

3 (Continued) at 0624 it apparently caused a heat sensitive fire detector to go off in containment. The detector was located above the drain valve on the operating deck. Since the operators had ihdidation of RCS leakage and a fire in the containment, an individual using respiratory protection was sent into the containment to investigate. This individual confirmed -the leakage and identified the source as the letdown line but was unable to identify the exact leak point because his air supply was low. During, the inspection no evidence of fire was found.

To prevent future occurrences the CVC-200E pipe threads were dressed and a new end cap installed. CVC-200E and several other valve/pipe cap arrangements which could be exposed to the same condition were inspected and physically locked or verified secured in the closed position.

-DRAFT tm 2

ATTACHMENT NO. 4 PRIMARY SYSTEM DEPRESSURIZATION The mainconcern during the transient of 1-29-81 was an unexplained decrease in RCS pressure. The pressure dropped from 2200 psig to 1620 psig in approxi mately one hour. Many steps were taken during the first hour of the transient to determine what was causing the depressurization. The pressurizer (Pzr) safety valves were checked by looking at the accoustic flow indicators down stream of the valves. No flow was indicated. The Pzr PORVs were checked by looking at the pipe temperature downstream of the valves.

Again, no flow was indicated. The Pzr block valves were checked to. verify that they were shut.

The Pzr relief tank level and pressure were also checked to verify that they were not increasing. The main Pzr spray valves were then switched to manual control and closed by the operator. The indication on the RTGB showed the valve to be closed, however, since this indication is only of demand position, the operator tried to insure that the valves had closed by manually closing them. The charging line was then isolated to see if the auxiliary spray valve, CVC-311, was leaking. Additionally, RCP "B" and "C" were stopped so that flow through the main spray valves 455A & B was not possible.

Pzr pressure began increasing. Later that night (2315 hours0.0268 days <br />0.643 hours <br />0.00383 weeks <br />8.808575e-4 months <br />) spray valve 455B was positively identified as the leaking valve.

An inspection of ithe valve showed that the sten was binding on the valve pack ing. One reason the binding problem was not identified earlier is that the spray valves do not move much during power operation. RCS pressure control is accomplished by varying the Pzr heaters with the spray valve partially opened.

The valve was rdpaired by lubricating the stem. The valve was then tested four times to insure proper operation. In addition, the electro-mechanical positioner zero setpoint was discovered to be slightly off and therefore was reset.

-DRAFT tm

ATIACHMENT NO. 5 CONTAINMENT ISOLATION VALVE FAILURE (FP-248)

At 0624 on 1-29-81 a signal generated a Phase A containment isolation. As part of this isolation the newly installed fire protection containment isolation valves FP-248, FP-249, FP-256, FP-258 were signaled to shut. FP-248 did not shut. The valve was then manually shut. The cause of failure was a tripped breaker which would not allow power to the motor operator. Subsequent review indicated that the trip point on the magnetic overload breaker was not set high enough to insure proper operation.

The breakers had been tested successfully upon installation, however, the current demand of the valve motors can change with time and so if the trip point is not set with enough margin the breaker can pass a test and yet fail at a later time.

The setpoints on all four valves have been readjusted to compensate for the above problem and tested. This should correct any future problems with these valves.

DRAFT tm 1

ATTACHMENT NO.

6

SUMMARY

OF P250 COMPUTER OUTPUT Time Event 0620 Alarm -

Low Tavg Permissive Set 0620 Alarm -

Low Tavg 541.2 (setpoint is 543.0) 0623 ORR - Control Rod Bank C Inserted (reactor trip) 0624 Alarm -

RHR Pump "B" BKR -Closed (SI signal) 0624 Alarm -

Low Tavg 532.7 (minimum'Tavg) 0625 INCR -

Hi PZR Relief Tank 75.2% (Valve CVC-203 lifts) 0627 RETRN -

RHR Pump "B" BKR Open (SI reset) 0705 Alarm -

PZR Low P & L SI (Second SI signal) 0705 Alarm -

RHR Pump "B" BKR Closed 0705 Alarm -

RHR Pump "A" BKR Closed 0726 Alarm -

RCLB Lo Flow (.RCP "B" stopped) 0727 Alarm - RCLC Lo Flow (RCP "C" stopped)

-DRAFT to

__JAN 291981 72J%~~L/~

~ ef/

/~,------------

4/

QL

(

p 'tC-.;

cc~

~

C-~~7 gKO J/ 7/r____

c 5

. 1 C

02 e

l4.-~

i5.

?A O227e

~~

~

el d-,"/A uI.i vo,.

[,e'X

~~~~~'~~C

--_-- vo~c~ --

oo O-P t

-I--

1'real 3 2:5

-o Dae~ 5~cz~-~-Tme-L~

i~L(

per.~~

BankPosition_(Steps)

--A.iiCL2:

B2 -

C zaiL_

D Tave_-- 7 -- F---rsue2->.~G Boron-Concentration-PPM

• Reactor Power Level

,----Amps 0

• ~~_g

__c;

?

~~ I.Cl--4 e def-A..

0

~

4 v4 &

~

_Pool_

/3,,.3 os7 e.~~d'Ge Ar4 S

4 ~ J

. w~

~77ct

~

• ~

tp~

~

A n

~

N

/~r~

(uI~j~~,-C

'~

W c~~.AlvW alal ik C--

2-Vf~

oii&,

/77,3' Ac j

(*17

-/

x~

r'A 1

(~~~~7

~-

A 5Z(4-2 ~/C M-(

/01

~ ~ r,?re-~

/~5

~

(A

1-4 1-4

-7~---

22

- try rs A-I (7-0

~/,

) ~

o.

- c~ -~ / ~ r -e71

'e~~1~~A7

__'A

_ - - - ( - -'

~ ' c -'

A _

?

6 E 7(L A

'~

-~

9 2

~

~

7 7 7 J

DA~ K TIM N.H.

R O B IN S O NN

?7 3

z-If(-5),5X

.A!Z

.4, U-A' f'G ~

~Jd4lA p4 j1

/e-J600, ;

/i C,

24

___________________A l o*

'0**1.

~~.

~H. B. ROBINSON'N 7 3 Two,11:, '

UNIT 2?

7 3 fo.e M l -

/

I-- jt'4Z IfL dn. i 01X Ci&.L Z 661 6-&

r.

mf

~

~

~

b S

~~

elxf ti7 XIN C

~~s r.~

LJ~Atse cQ Ll/4 1~~ &06 (5X f

4 AIL k

~

~~ L u L

Set fi 7,A ec_

ta'

,A*%*

1-jfGo T

.26

e. Y cP A

c-

H. B. ROBINBONN?

73 lasF%4 STUNIT~~

K1, 4--7

-11 64(2 P

T-

/ ~z-00, aa P-x

F4~

1 il H;I I

1,1 il H1' 8H H

I if!

Is

• 'I[

H ti l ' l I n I I! I I I H I i.i

ii I

- 0 I!,

1l l

'1 I

i'l I: I 1:

0 til li l 11 1 lli l 1 41 I

11 W l I

ll I !I I

II

[I~i WI dill 11 Hillii I~6" 6

II H

I 11 Il Il 11111 IHi HlI IHilI il I il I

ill I M ;:!I T

1! 1 111 H I 1*

.1111 11 H l H ll~ I, 71 f t l H6J 11 1!

H I III I Ill H

I i fQ II Ii~ 1166 Iii fll fl I1 I

L

.Q 7V

'II Ai II I~ 1 k40 I t 0

w;O

!f H id

II Ella I

I.

i I IT I I I

I I

IM I

L 1

I id

.1 I4iI it

i.

ol1 1 1 i

tj.

.1 I

I 1 7171 1 11 lH 1

1.1 op 01

.1 II.

~

.~.fill

j II I I

Ii I1AJjjw A it i

if AT[

I-ME'

.3 i-6 0

zP O..:

I mi I

I I

f i

I

_T 1 -7_!7

--t

-7v 74

AN ~

~

~.

.ldl.OW~dO

.1ttO

.IdW

.0.

2900 25004 I

0 F 004111 70 10 20 302 0

I 70 1900

!ooj 23

.... 0 1

100 0

0010020230 20 I~00

'100 2

0I 100 1

701900 2S1 Hi 4q~j'+/-~~ I;Tjw Ih 10

vvI!IflhIIII 17000 1IIN~6IVlo~ 1210

~12306H 2SO 0

2 L..4 0apA~

1 - l i

• • IIj

-i I.

1 IIt iti

-T, T

i It 1,

i 7,

v -

.~ -17

-~ I I

id flhi-Ji! I

S-T t

SOO*@S@

~~~IO 0

AW l Smm9~@

Er-E~

49

t-QT\\,/AQ 4g~

~~

00' e0

• 0@S 0 o0 000 9

• 0 0

0S

@ 0

~~AOA MBNOVWdfl NOI+/-W

7 AM-4-w---,

'I 6M.

Sp-&

4 AM CI

Ir U3 lit iji I

ti I

IIII I

iti t

t i 1'1 i

1,1d 11II::,! 1 1

1I I

LiI~

I

-L..

. J L

-TI FrIT!

TI A

4 -I I'I n-TM

.1 Ji'

_7 H,

4 1 5

-hl Hil 7-7-

FIGURE 1 C VCS DIAGRAM1 (excer~t)UNA;5E 2470N 0~

C~

e34e 0

ID 2-C.~ v -W-Z

1.

0L

-bo 20B 122-r.

CAT2 205\\

LINE.

1A1f!

C~25O30 A

~

- ~.C'~Z~C-A3-CM jr

<A~~~~~~c wsQ&

co*.

7 j

FIGURE 2 CONTAINMENT SUMP VOLUME 0J3Q L9 ~

GO

~Top of Sump f1 z

/ ( Z

!4 iI S

20.

27t 4Z4 Z'

3 34-3r&