ML20086A635

From kanterella
Jump to navigation Jump to search
Special Rept,Forwarding Problem Investigation Rept 1-C91-0304.On 910725,boron Concentration in RWST Below Required TS Limits.Cause Not Determined.Sample Loop Validity Restored & Sampling Frequency Increased
ML20086A635
Person / Time
Site: Catawba Duke Energy icon.png
Issue date: 11/12/1991
From: Mccollum W
DUKE POWER CO.
To:
NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
1-C91-0304, 1-C91-304, NUDOCS 9111190272
Download: ML20086A635 (16)


Text

11 l

.tiure nm er rompany pop nsui>i>

(i:latrba N;u k.if Slaltort 70 lins 2%

Chtr. S(':9710 DUKE POWER November 12, 1991 Document Control Desk U. S. Nuclear Regulatory Commission Washington, D.

C.

20555

Subject:

Catawba Nuclear Station Docket No. 50-413 IIR C91-080-1; PIR l-C91-0304 Gentlemen:

1 Attached is our Problem Investigation Report 1-C91-0304, submitted concerning REFUELING WATER STORAGE TANK SAMPLE DILUTION /BORATION.

This incident has been determined to be non-reportable but has been investigated and is being documented as a Special Report to ensure industry awareness of this event.

The health and safety of the public were not affected by this incident.

Very truly yours, h.l}_ /Y

,/

W. R. McCollum

/

Station Manager kon: REPORT.SP xc:

Mr. S. D. Ebneter M & M Nuclear Consultanta Regional Administrator, Region II 1221 Avenues of the Americas U. S. Nuclear Regulatory Commission New York, NY 10020 101 Mariett a Street, NW, Suite 2900 l

Atlanta, GA 30323 1

R. E. Martin INPO Records Center U. S. Nuclear Regulatory Commission Suite 1500 Office of Nuclear Reactor Regulation 1100 circle 75 Parkway Washington, D. C.

205b5 Atlanta, GA 30339 Mr. W. T. Orders NRC Resident inspector o $ 1 t hfy2 Y @ Ti 1 Y '"

p ppa f4 DOCK 05000413 l O y t! p?

PDR 4 r r

D t

R 3

N*

DUKE POWER COMPANY CATAWBA NUCLEAR STATION PROBLEM INVESTIGATION REPORT NO.:

1-C91-0304 REFUELING WATER STORAGE TANK SAMPLE DILUTION /B0 RATION ABSTRACT on-July 25, 1991, Catawba Unit I was operating in Mode 1 (power Operation) at 100% power.. Makeup to_the Refueling Water Storago Tank (FWST) had been completed at 0130 hours0.0015 days <br />0.0361 hours <br />2.149471e-4 weeks <br />4.9465e-5 months <br />. -After-recirculating the tank's-contents as required, a sample was taken and analyzed with a resultant boron concentration of 1809 ppm, below the 2000 ppm required by Technical Specifications. Confirmatory samples indicated-concentration of 1865 and 1900 ppm boron.

Based on the wide fluctuations in indicated concentrations and the volume of water added, it was concluded that these results.were not representative of the FWST and that the

[

samples had'been diluted. Alternate samples consistently indicated an FWST l

concentration of 2057 and 2065 ppm boron. Ongoing sampling of the FWST via the normal sample: loop continued to produce widely varying results, from 1958 to 3035 ppm boron. A unit. shutdown had been initiated but was stopped when FWST operability was confirmed. A Waiver of Compliance rocoived from NRC was not

used as the FWST was shown to have the proper boron concentration.

An investigation into the cause(s).of the FWST samplo dilution /boration was initiated and continued for several weeks.' No singular, definitive cause of "this problem was found. Corrective actions include restoration of FWST sample loop validity, testing to determine the sample dilution /boration source (s), and sample procedure enhancements. This report is submitted as -a Special Report.

DUKE POWER COMPANY / CATAWBA NUCLF1.R STATION y

pIR l-C91-0304/Special Report

page.2 BACKGROUND The Refueling Water-(FW) System is designed to provide

a.

a-source of borated water at refueling water boron concentration for uso during refueling or a postulated Loss Of Coolant Accident; b.

recirculation of the Refueling Cavity and fuel transfer canal water for-cleanup during refueling, as necessary; c.

recirculation of water in the Refueling Water Storage Tank (FWST) for cleanup following refueling, as necessary;

-d.

borated makeup water to the fuel pool; and an assured source of borated water for makeup.of primary loop shrinkage and e.

leakage during an emergency cooldown in the event of a tornado-induced puncture of the Refueling Water Storage' Tank (FWST).

The FWST provides a source of borated water to be used during refueling, following a Loss of Coolant Accident (LOCA), as makeup to the spent fuel pool and for other miscellaneous purposes. The FWST nominal capacity of 395,000 gallons is sufficient to provido a useable volume exceeding 350,000 gallons.

The FWST is an insulatod, electrically heated tank.

It is located outdoors on a concrete pad surrounded by a tornado missile shield wall approximately 14 feet high.

This missile wall-is intended to assure retention of a sufficient volumo of borated water for makeup of primary coolant shrinkage and leakage during an emergency cooldown following a tornado induced puncture of the FWST.

A twelve inch vent and a six inch vent extend approximately two feet above the top of the tank and terminate in a 180 degrees bend with-a screen.

Both the vent piping and screens are heat traced to preclude ice blockage in freezing weather. The 24 inch discharge header from the FWST to the Auxiliary Building

~

needs to be maintained above approximately 70 degrees F.

The Refueling Water

-Recirculation Pumps A&B are used to recirculate this volume of water to the FWST as necessary during cold weather.

These pumps are operated manually from the Control Room.

The Refueling Water Pump is designed to transfer the water volume of the Refueling _ Cavity and transfer canal in approximately'24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The total dynamic head is: sufficient for the most stringent operating mode which is the transfer of water from the Refueling Cavity to the FWST via the Spent Fuel pool Cooling (KF) System demineralizer and filters. The pump is operated manually from the Control Room. The Nuclear Sampling (NM) System provides representative p

samples for laboratory analyses of chemistry and radiochemistry, which are used for guidance in the operation.of various nuclear related systems. Typical information obtained includes reactor coolant boron and chloride concentrations, fission product radioactivity level, hydrogen -and oxygen content, fission gas I '-

content, pH,Lcorrosion product concentration, and chemical additive concentration.

m

' DUKE POWER CuhpANy/CATASBA HUCLEAR STATION PIR l-C91-0304/Specia) Report--

,Page;3 A sample room is provided for each unit.

Samples may be obtained within the sample room of each; unit from locations inside containment as well as locations outside containment; including FWST discharge.

Local sample; connections are provided at various other locations outside the containment, but are not considered part of the Nuclear Sampling System.

This system has no emergency function.

Sample lines originating within the containment are provided with motor-operated containment isolation valves, locatedJoth inside and outside the containment, which are closed automatically by a co'.tainment isolation signal. These valves are also operable from the sample room to select the sample to be taken.

In addition, a manual isolation valve is located close to each_ sample' source.

Sample lines originating outside the containment are also provided with an isolation'valvo close to the source. The residual heat removal heat exchangers are provided with remote motor-operated isolation valves which are activated-from the sample room forJselecting the sample to be taken.

Within the sample room, sample heat exchangers are provided to cool samples from l

the reactor coolant loops, residual heat removal loops, pressurizer steam.and 11guld spaces, and the steam generator blowdown lines.

purge lines are provided through which sample fluids may flow until sufficient volume has passed to permit collection of a representative sample. Sufficient flow is directed to the sample sink to purge the remainder of the sample line prior to taking a sample. Remote air-operated purge line isolation valves are provided wherever radioactivity makes the use of manual isolation valves prohibitive.

Needle valves are provided in the sample purge lines for the reactor coolant loops, the pressurizer steam and liquid space, the residual heat removal-heat exchangers, and the seal injection filters.

All_ sample lines containing. reactor grade water are purged to the Volume Control Tank. ~All steam generator blowdown lines are routed to the Condensate Storage Tank, and flowLeontinuously. The volume control tank gas space is purged to the Waste Gas _ System, and all other sample lines are purged to the Waste Evaporator Feed Tank Sump A for Unit 1, and Waste Evaporator Food Tank Sump B for. Unit 2.

Two sample sinks are provided-in the sample room. One sink includes all the samples which require heat exchangers or sample vessels plus the accumulator sample lines. The second sink includes all the other sample lines that are a part of the syatem. The sample-sinks are provided with continuous ventilation discharges to-the unit vent.

Hunual valves, for sample line discharge isolation and flow control, are mounted in the appropriate sample sink.

Sample vessels are provided at the primary sampling _ sink to facilitate the collection of pressurized samples from the pressurizer, the reactor coolant loops, and the volume control tank gas space for laboratory analysis of hydrogen, oxygen, and fission gas concentration.

p 1-

DUKE p0WER COMPANY / CATAWBA NUCLEAR STATION pIR 1-C91-0304/Special Report page 4 Each sample room is arranged so that all sample heat exchangers and a large portion of the sample and purge line piping is located behind a shield wall within the room.

This reduces significantly an operators exposure to radiation while obtaining a sample.

Technical Specification a.5.4 requires the FWST boron concent ration to be maintained between 2000 and 2100 ppm of baron and verified at least once per 7 days.

Chemistry section procedures require sampling of the FMST every week and following overy addition to the tank.

Event Description On July 16, 1991, Unit I was operating in Mode 1 (power Operation) at 100%

power. A routine sample and analysis of the Refueling Water Storage Tank (FWST) indicated the boron concentration to be 2047 ppm baron.

The expected range of boron concentration is 2000 to 2100 ppm.

On July 22, MOVATS (Motor Operated Valve Actuator Test System) Lesting and subsequent maintenance on valve 1FW-49B was completed. This work included IWV (inservice valve) testing of the valve's performance per procedure PT/1/A/4200/20, Enclosure 13.5.

At 0955 hours0.0111 days <br />0.265 hours <br />0.00158 weeks <br />3.633775e-4 months <br /> on July 23, the FMST was placed in recirculation using the Refueling Water (FW) pump and sampled from the NM Sample Room.

The results of this sample wera 2017 ppm boron.

At 2154 hours0.0249 days <br />0.598 hours <br />0.00356 weeks <br />8.19597e-4 months <br /> on July 24, an addition to the FWST was begun (using procedure OP/1/A/6200/14, " Refueling Water System", Enclosure 4.4 FWST Makeup from Blender") to raise its level from 95.5% to 97.8%.

The addition was completed and level reached 97.8% at 0130 hours0.0015 days <br />0.0361 hours <br />2.149471e-4 weeks <br />4.9465e-5 months <br /> on July 25.

Catawba administrative controls require a boron concentration analysis following overy FWST addition.

By procedure, the final 1000 gallons of the solution directed to the FWST was pure domin water from the Reactor Makeup Water Stotage Tank. This is done to flush remaining boric scid from the blender piping. At 0330 hours0.00382 days <br />0.0917 hours <br />5.456349e-4 weeks <br />1.25565e-4 months <br /> the FMST was placed in recirculation using the FM pump. At 1130 hours0.0131 days <br />0.314 hours <br />0.00187 weeks <br />4.29965e-4 months <br /> the FMST was placed in recirculation using FWST Recirculating Pump 1A.

At 1210 hours0.014 days <br />0.336 hours <br />0.002 weeks <br />4.60405e-4 months <br />, an FWST sample was obtained from the Nuclear Sampling (NM) system and taken to the chemistry lab.

There, the Chemistry Technician analyzed quality control (QC) standards in preparation for the analysis.

A problem was encountered with the mechanical stirrer.

The problem was resolved and the QC standards were reanalyzed.

At 1330 hours0.0154 days <br />0.369 hours <br />0.0022 weeks <br />5.06065e-4 months <br /> results were obtained which indicated an FWST boron concentration of 1809 ppm.

This result was well below the T/S required minimum value of 2000 ppm and was a 208 ppm decrease from the results of the analysis two days before.

The low result caused the Chemistry Technician to question her analysis, especially ir. view of the problem with the stirror, which could have recurred.

Boron titration is automatic and does not require constant attention by the technician.

By 1420 hours0.0164 days <br />0.394 hours <br />0.00235 weeks <br />5.4031e-4 months <br /> the sample had been reanalyzed by two separate technicians with similar results.

DUKE POWER COMPANY / CATAWBA NUCLEAR STATION.

PIR 1-C91-0304/Special'Raport-Page 5 At 1430 hours0.0166 days <br />0.397 hours <br />0.00236 weeks <br />5.44115e-4 months <br /> a decision was made to obtain a confirmatory sample, as required _

by Chemistry management procedure and a technician departed for the NM room. At 1440 hours0.0167 days <br />0.4 hours <br />0.00238 weeks <br />5.4792e-4 months <br />, the technician began purging the sample line, as required by procedure. Due to the low flow rate from the sample line, 40 minutes was needed to obtain'the required 5.5 gallon purge volume.

Two samples were obtained, one at'1520 and a second at-1525 hours. By 1545 hours0.0179 days <br />0.429 hours <br />0.00255 weeks <br />5.878725e-4 months <br /> the results of the first confirmatory sample were obtained and indicated a concentration of 1865 ppm boron.

This represented a change of 56 ppm from the sample analyzed two hours previously. The results of the second confirmatory sample were 1900 ppm boron.

Such wide variations in sample results were clearly anomalous.

At 1520 hours0.0176 days <br />0.422 hours <br />0.00251 weeks <br />5.7836e-4 months <br /> the unit Operating Engineer had been notified that there was an indication of an inoperable FWST due to low boron concentration and that a confirmatory sample was forthcoming. Discussion ensued that involved the Shift Manager, Chemistry Section Manager, Operating Engineer, and Chemistry supervisors and technicians. Discussion focused on the operability of the FWST.

While'it seemed inappropriate to allow the FWST to remain in an " operable" status because two unsatisfactory samples had been taken and analyzed per station procedures, declaration of."inoperability" seemed inappropriate due to the impossible nature of the sample results. A conservative decision was recommended to declare the FWST inoperable. A Technical Specification Operability Notification Sheet (TSONS) was generated by Chemistry which declared the FWST inoperable due to low boron concentration. The TSONS was taken to Operations and signed at 1658 hours0.0192 days <br />0.461 hours <br />0.00274 weeks <br />6.30869e-4 months <br />,-one hour and thirteen minutes after the confirmatory sample results were obtained. T/S 3.1.2.6 and 3.5.4 allow one hour to restore the FWST to operable status and six additional hours to bring the unit to Mode 3, Hot Standby.

Sampling of the FWST from the NM rm continued, with times and results as follows. A continuous purge fre FWST sample line had been established to expedite sample collection.

Time Result 1625 hours0.0188 days <br />0.451 hours <br />0.00269 weeks <br />6.183125e-4 months <br /> 1937 ppm 1643 1927 1709 1958 1726 1974 1750 1977 1820' 1988 1931 19o3 1942 1963 1952 1957 An anticipatory alignment was made to add boric acid to the FWST. The addition

-was not made due to a concern that accurate and representative samples would-indicate that the FWST was~already above 2000 ppm boron and that the addition would cause the FWST to exceed the T/S upper limit boron concentration.of 2100-ppm.

i l

DUKE POWER COMPANY / CATAWBA NUCLEAR S1AT10N PIR 1-C91-03C4/Special Report

.. page.6 At 1900 hours0.022 days <br />0.528 hours <br />0.00314 weeks <br />7.2295e-4 months <br />, the Shift Manager initiated discussion with the duty Station Manager and the NRC Resident Inspector about the possibility of obtaining a waiver of T/S compliance for continued unit operation.

At 1930 hours0.0223 days <br />0.536 hours <br />0.00319 weeks <br />7.34365e-4 months <br />, the Shift Manager called the Nuclear Production Department duty engineer to request a j

review of system load and the dispatcher's needs. By 2045 hours0.0237 days <br />0.568 hours <br />0.00338 weeks <br />7.781225e-4 months <br /> the duty engineer had indicated that unit operation was needed to match vistem demand through the night.

I A decision was made to begin a power decrease in compliance with T/S requirements. At 2027 hours0.0235 days <br />0.563 hours <br />0.00335 weeks <br />7.712735e-4 months <br />, an addition of 325 gallons of concentrated boric acid was made to the reactor coolant system.

Due to the previous anticipatory alignment for adding boron to the FWST, it is believed that some amount of the boric acid was directed towards the FWST before the system was aligned to the raector coolant syste.

Since circumstances indicated that the NM samples were rat representative of the FWST contents, an alternate sample point was sought.

Chemistry procedure Op/p/A/6200/11 allows Chemistry management to authorize alternate sample points when the primary point is inoperable. A decision was made to sample the FWST via the containment Spray (NS) pump.

NS pump 1A was started at 1839 hours0.0213 days <br />0.511 hours <br />0.00304 weeks <br />6.997395e-4 months <br /> and stopped at 1931 hours0.0223 days <br />0.536 hours <br />0.00319 weeks <br />7.347455e-4 months <br />.

Samples were obtained from the NS pump suction at 2035 and 2036 hours0.0236 days <br />0.566 hours <br />0.00337 weeks <br />7.74698e-4 months <br />.

Results indicated boron concentrations of 2057 and 2062 ppm.

Tha Station Manager, duty station manager, and Compliance Mana;sr notified the NRC of the situation.'t. 2100 hours0.0243 days <br />0.583 hours <br />0.00347 weeks <br />7.9905e-4 months <br /> and requested a waiver of T/S compliance for continued unit opera n.

A decision was made at 2130 bours to hold the power decrease at 75-80% power due to the high level of confidecce that the FWST boron concer lon was actually acceptable. The walvor of compliance was received from N.

AR at 2145 hours0.0248 days <br />0.596 hours <br />0.00355 weeks <br />8.161725e-4 months <br /> and was to be in effect from 2358 on July 25 until 0758 on ly 26.

Once in the waiver, unit load was not to be increased and any indicatic of unexpected results from boron manipulations would require piacing the unit in Mode 3.

If the T/S action statement was exited prior to the waiver being in effect, the waiver was to be retracted.

i After discussion betUeen Operations and Chemistry personnel, Chemistry submitted a TSONS at 2200 houru indicating that the FWST was operable based on the results of the two samples obtained from the NS pamp suction (2057 and 2062 ppm).

Operations signed the TSONS at 2203 hours0.0255 days <br />0.612 hours <br />0.00364 weeks <br />8.382415e-4 months <br />.

The waiver of compliance was never entered. To further assure FW2T operability, the FW pump discharge vent was identifjcd as an. independent sample point.

Samples were taken at 2225 and 2226

' hours wath results of 2061 ppm boron.

This provided additional confirnation that the FWST was operable and that a dilution source was confined to the recirculation loop.

Unit 1 began a return to 100% power at 2335 hours0.027 days <br />0.649 hours <br />0.00386 weeks <br />8.884675e-4 months <br /> on J;ly 25, increasing power

.~

at 2.5% per hour.

Full power was reached at 0738 hours0.00854 days <br />0.205 hours <br />0.00122 weeks <br />2.80809e-4 months <br /> on July 26.

+

Operations and chemistry personnel had initiated an investigation of possible causes of sample dilution.

Evidence indicated that the FWST recirculation loop and the NM samples were not representative of the FWST.

In an effort to confirm this, samples were taken at the recirculation pump discharge vent (via 1FW-64) l and at the NM sample panel at about the same time.

2....

DU2 POWER COMPANY / CATAWBA NUCLEAR STATION PIR 1-C91-0304/Special Report Pags.7 Sample.pnel pump Vent Time Result Time Renuit 205% hours 1956 ppm 2105 hours0.0244 days <br />0.585 hours <br />0.00348 weeks <br />8.009525e-4 months <br /> 1974 ppm 205o 1961 2106 1974 These sample results clearly indicated that the entire tecirculation loop was not representative of the contents of the FWST.

During this investigation, it was noted that FW Recirculation loop flow indication was lost at 2005 hours0.0232 days <br />0.557 hours <br />0.00332 weeks <br />7.629025e-4 months <br />.

At thu _ time, the only potential ingress path that could be found originated at a domin water (YM) makeup connection to a loop seal on a line in the annulus drain header. A cross connection was thought to exist that might provide a flow path

-to the NH room if valves 1FW-49B or 1FW-33a were in the closed position. A test alignment was impicmented which isolated the source of YM to the loop seal makeup. The premise was that if the YM ingress stopped, samples in the NM room would return to values representative of the FWST.

It was thought that 1FW-49B was allowing enough flow to obtain a sample but not providing enough pressure to block the ingress of dilution water.

Ilourly sampling at the NM room was started.

Time.

Resulta -

2259 hours0.0261 days <br />0.628 hours <br />0.00374 weeks <br />8.595495e-4 months <br /> 2750 ppm 2300 2743 2355 3035 0220 2294 0405 2172 0607 2091 The high concentration results were totally unexpected. The boric acid blender was sampled in an effort to find the source of the high results.

t 0405 hours0.00469 days <br />0.113 hours <br />6.696429e-4 weeks <br />1.541025e-4 months <br /> 2833 ppm 0607 1574 At this time, it was suspected that the boric acid blender inlet to the FW system could be the source of dilution and boration ingress. 'The investigation into this series of events continued.

-An FWST sample taken from the NM sample room at-2325 hours on July 28 indicated a-boron concentration of 2141 ppm. Additional samples taken from the same source at 0050 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br /> on July 29 indicated 2136 and 2121 ppm boron. These values-were above the T/S upper limit of 2100 ppm.

Later FWST samples indicated Time Location Result-0335 hours NS Pump Drain 2048 ppm 0525 NS Pump Drain 2041 0805 2054

Dure POWER COMPANY /CATAt1BA NUCLEAR STATION PIR 1-C91-0304/Special Report

.page.8 During investigation, valve 1FW-49B was found in tne almost closed position and not opening electrically. This valve is a 2 inch Kerotest packless globe valve l

in the suction line of the FM Recirculation pumps; IFW-33A is an identical valve in series with IFM-49B. Operations initiated Work Request 557980PS at 1400 on July 30, and white tagged IFW-33A closed in order to maintain operability of the FWST.

IAE investigated the problem and then turned the problem over to Mechanical Maintenance (MM).

By 1720 hours0.0199 days <br />0.478 hours <br />0.00284 weeks <br />6.5446e-4 months <br />, MM and operations had determined that the problem with the valve was mechanical in nature.

By 0500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br /> on July 31, Operations had tagged 1FW-49B out of service and MM began to work on the valve.

j i

Valve 1FW-49B was found to have rust inside the bonnet and a crushed diaphragm.

The disc ansembly appeared to be deformed which' caused the valve to seize in the (nearly) closed position. Copies of previously completed work requests for 1FW-49B were reviewed and no maintenance or QA irregularities for faulty workmanship were noted. The failure of-this-valve was believed to be crucial in the explanation of_the source of_the FWST sample dilution /boration.

By 0520 hours0.00602 days <br />0.144 hours <br />8.597884e-4 weeks <br />1.9786e-4 months <br /> on. August 1,1FW-498 had been reassenbled and 1AE was to reattach and set up the actuator. The valve was operable at 2130 hours0.0247 days <br />0.592 hours <br />0.00352 weeks <br />8.10465e-4 months <br />.

The valve was MOVATS tested'and an IWV stroke time test was performed with a result of 4.4 seconds.

1 An FMST sample was taken at 2100 hours0.0243 days <br />0.583 hours <br />0.00347 weeks <br />7.9905e-4 months <br /> on August 3, during recirculation of the tank using the FW pump; results indicated a concentration of 2061 ppm boron.

It was discovered that the flow gauges for the FWST Recirculation pumps had failed at a reading of 40 gpm. Repair work was initiated by IAE and completed by 1800 hours0.0208 days <br />0.5 hours <br />0.00298 weeks <br />6.849e-4 months <br /> on August 5.

Evaluation of possible explanations of the sources of FWST sample dilution /boration continued. An Abnormal Plant Evunt (APE) meeting was held on August 5 Involving plant management and personnel involved in the ongoing investigation.- A number of action items resulted from this meeting, including:

1.

Operations and Performance were to test the FWST Recirculation pumps following completion of repairs to IFW-498.

possible backleakage through the pump discharge check valves was to be assessed.

2.

Chemistry, Operations, and Radiation Protection (RP) were to develop

' alternate sample points for this and other sampling situations critical to unit operation.

3.

Performance was to develop permissive values for procedures to use in verifying adequacy of flows, prassures, etc. In sample lines to ensure valid results.

4.

Operations was to evaluate the adequacy of instrumentation used in verifying recirculation flows, sample flows, discharge pressure, etc. and consider adding them to preventive maintenance programs or periodic calibration checks.

I i

l e--e n

U.

.-,,,,,m,..

DUKE p0WEa COMPANY /CATAWRA NUCLEAR STATION pIR l-C91-0304/Special Report

  • Page '9 5.

Following repairs, Chemistry was to compare samples from the NM system with local samples for consistency.

The Operations Manager's group was asked to see if the observed FWST recirculation loop' flow rates were reasonable. Operations tested both pumps at 0500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br /> on August 6; both apperred to produce about 30 gpm. The flow gauges read 10 gpm with_the pumps off anc 43 gpm with pump 1A running: 39 gpm with IB running. MES advised operations that 10 gpm flow indication with the pumps off may not necessarily mean that the reading to inaccurate with the pumps running.

Performance was requested to ina:all precision flow instrumentation and verify the readings.

At 1700 hours0.0197 days <br />0.472 hours <br />0.00281 weeks <br />6.4685e-4 months <br /> on August 7, Operations ran the FMST Recirculation pumps and Performance determined that ooth pumps were producing 40 gpm. These results also confirmed that the pump discharge check valves were seating properly at that time.= on August 24, a low FMST Recirculation boop Flow alarm (from flow gauges 1FWPG5080)-we. received following Auxiliary Safeguards testing.

Valve IFW-49B was again foand failed in the closed position and not passing flow.

During repair under W/R 4127PL1, MM found that the valve seat appeared to be OK but that the diaphragm.'as deformed / crushed, and the disc assembly was mushroomed causing the va;ve to become stuck in the closed position.

PIR 1-C91-0333 was written to ovnument this condition.

l A meeting-was held on September 5 involving Chemistry, Operations, Performance, L

and Safety Review Group personnel to discuss investigation results to dato and to plan additional action.. A plan was developed to test th? NM sample header in order to identify existing leakage sources that could have contaminated the FWST samples taken in the NM sample room. This testing was conducted during the following two weeks.

Another meeting was held on Septenber 24 to review investigation results and possible action. A plan was developed to recreate a previous recirculation loop alignment and determine the source (s) of water drained from the FW Recirculation pumps casing drain line, and in particular, possible leakage from the WL loop seal. This testing was performed the following week.

CONCLUSION The primary conclusion resulting from the investigation surrounding this event is that the. abnormal baron concentrations (less than 2000 ppm, greater than 2100 ppm boron) were NOT representative of true FWST boron concentrations.

Samples taken from the FW pump discharge vent-and the NS pump clearly indicated that FWST concentration was within Tech Spec requirements. Samples taken in the NM samp1w room and from the FW Recirculation Pumps discharge vent indicated that the recirculation loop was experiencing both sample dilution and boration.-

The FWST was declared inoperable at 1658 hours0.0192 days <br />0.461 hours <br />0.00274 weeks <br />6.30869e-4 months <br />, one hour and 13 minutes after the results of the confirmatory samples from the NM Sample Room were obtained..

During this' time, discussions took place involving the Shift Manager, Chemistry Section Manager,' Operating Engineer, and Chemistry staff.

While it was

-recognized that two samples indicated that the FWST was not operable, declaration of inoperability was considered inappropriate due to the impossible nature of the sample results.

I

DUPE POWER COMPANY / CATAWBA NUCLEAR STATION p1R 1-C91-0304/Special Report page,10 Discussion of the impossible nature of the sample results included the following facts.

1.

Even if pure demin water had been used to raise the FWST level from 95.5%

to 97.8% earlier in the day, the resultant boron concentration would have been approximately 1969 ppm, well above the 1809, 1865, and 1900 ppm sample results.

2.

The results of the two confirmatory samples indicated an increase of 35 ppm between 1520 and 1525 hours0.0177 days <br />0.424 hours <br />0.00252 weeks <br />5.802625e-4 months <br />, an increase beyond the capability of station boration equipment if it actually occurred in the FVST.

3.

Decreases in Boric Acid Tank level and increase in FMST level occurred at the expected times and indicated that the boric acid was in fact transferred to the FWST.

4.

FMST level indicated no other additions which could have diluted the concentration.

The cause(s) of the sample dilution /boration have not been conclusively proven.

Several possible causes have been postulated and investigated, thus far no single explanation has been found that fits all of the observed data.

It in believed that a conbination of factors were at work.

The condition of the FMST recirculation loop isolation valvo 1FW-498 has been a key element in the investigation.

Its position and ability to pass flow at critical times has been carefully investigated.

Unfortunately, conclusive

~

evidenco has not been found. What is known is as follows.

During investigation by MES personnel, it was discovered that the IAE valve test crew had accidentally stalled the Limitorquo motor during MOVATS testing after the July 31 repair.

The potential for motor stall exists during MOVATS testing since adjustment of the control switchen is often required to satisfy the test criteria. A stalled motor can produce over 30,000 lbs. of force.

According to the manufacturer (Kerotest), the valve stem bearings and disc assembly would experience damage first if the notor applied too much force.

Kerotest has stated that over 27,000 lbs. of force would be needed to damage the disc assenbly.

The MOVATS procedure has been revised to prompt reporting of stalled motor situations to Technical Support and Engineering for evaluation.

Appropriato training has been provided.

No report of a stalled motor was received following the maintenanco and testing of 1FW-49B on July 22.

The fact that the valve stroke time increased from 2.7 to 7.8 seconds is a result of changes in the limit switch setup method. This change was implemented a year ago; however, this was the first time that 1FV-49B was setup with the new method. Motor current traces from the July 22 testing were analyzed as normal with no indication of degraded actuator performance.

It should be emphasized that in the Kerotest valve, the valve actuator is not directly coupled to the disc assenbly. As a result, actuator movement is not conclusive indication of valve movement.

DUKE POWER COMPANY /CATAtfBA NUCLEAR STATION PIR.1-C91-0304/Special Report Page 11 It is also important to note that indication of FMST recirculation flow in g

excess of 27 gpm (per computer point D2602) was recorded when expected on July 25 until 2005 hours0.0232 days <br />0.557 hours <br />0.00332 weeks <br />7.629025e-4 months <br />; at that time indication of recirculation flow was lost and not regained.

Theory 1: YM Intrusion Into FW Recire pump Flow Path When the FWST low sample concentration event began on 7/25/91, it was initially felt that a source of water was being introduced to the EV recirculation loop, providing a diluting effect.

Demineralized Makeup Water (YM) is provided to a seal loop in the Liquid Radwaste system (WL) which ultimately ties back into the Refueling Water Storage System (}W).

This theory is thought to be one explanation for the low boron results prior to 2300 hours0.0266 days <br />0.639 hours <br />0.0038 weeks <br />8.7515e-4 months <br /> on July 25.

(It does not explain the high boron concentration results.)_ Two seal loop auto-level control valves in series (1WL981 & IWL982) must have failed or be faulty,to establish this flowpath to the FW System.

That is, 1WL981 must have failed to the open position, allowing YM to fill and overflow the WL piping seal loop. This water would then need to flow towards IWL982 and in reverse-fashion through this valve.

This is despite the fact that the path of least resistance would be towards to Floor Drain Tank sump, and not through the faulty valve (1WL982).

Testing performed during the week of September 30 determined that thin dilution mechanism could not be replicated.

Theory 21 NM System Cross-Contamination In order for cross-contamination of the FWST sample point to occur at the FWST sample point in the NM Sample Room, several conditions not previously experienced would have had to occur.

FWST results were found to be low ~after extended purges to the WEFT and to the saraple _ sink.

Review of Flow Diagrams CN-1572-1.0 thru 1.3 and of Chemistry area logs revealed that samples which might cross-contaminate the FWST sample (RMWST, NI Accumulator, NB Distribution, etc.) had not been purged to the WEFT 1A from the NM Sample Room while the FWST samples had indicated out of spec low boron concentrations.

In the event a sample of lower boron concentration had been in purgo during FWST sampling / purge, several conditions'would have to exist to dilute the FWST sample.

1)

The FWST' purge valve to WEFT A (INM129) out of position (Open) or leaking by while purging to NM sink.

INM133'(Check Valve) would have to bc leaking by and into sample loop.

2)

While FWST was purging to the WEFT with INM133 and 1NM129 open, a reverse flow against FWST pressure would have to exist.

(Two Way Flow)

DUVE POWER COMPANY / CATAWBA NUCLEAR STATION PIR 1-C91-0304/Special Report Page 12 The RMWST would be the nest likely source of contamination via cross-connect to the FWST Sample Loop.

The RMWST sample was last taken prior to thin event, on 7/23/91.

The RMWST sample / purge valves were found to be in the correct position.

In order for cross-contamination of the FWST Sample point to occur resulting in concentrations greater than 2100 ppm boron, only one sample point is believed possible to be the source:

the NV Blender Outlet Sample point.

The out of spec results reported on the FWST (2750 ppm Boron) were taken at 2259 on 7/25/91. This was prior to the alignment for sample of the NV Blender in the NM LAB at approximately 0400 (blender boron was 2833 ppm).

In order for cross-contamination to have occurred to the FWST Sample boop at 2259 on 7/25/91, the following valves would have to be leaking by or out of position; INV160 (open), INM110 (open), INM133 (Check Valve Leaking By), and 1NM129.

Testing of the NM sample header confirmed that this scenario is very unlikely.

In particular the FWST sample purge valves 1NM-129 and 1NM-133 were not found to be leaking.

The Bori: Acid Blender sample valve 1NV-160 was not found to be leaking sufficiently to provide a sample dilution /boration source.

s Theory 3:

NV System Blender Ef fluent Intrusion Into FM Recirc pump Flow Path Another explanation for the FWST sampling anomaly is one that describes a flowpath from the Chemical and Volume Control System (NV) to the FWST sample point. For this flowpath to exist, the following failures must have existed or have been present when FWST sampling occurred:

1)

A recent addition must have been made from the NV System Blender to the FWST.

2)

One of the FW Recirculation pump dis;aarge check valves must have been stuck in the open position or leaking by cignificantly.

3)

One of the FW Recirculation Pumps must have been not running or damaged to the point of not being able to pump wster or to develop a discharge head.

4)

One of the FW Recirculation pump suction auto-isolation valves (1FW33A or 1FM498) must have been failed in the closed position.

The heart of this theory lies in the fact that during the latter stages of the troubleshooting process on the evening of July 25, redundant samples were taken at the NM sample point with elevated boron concentration. These sample concentrations were in the range of 2700 ppm to 3000 ppm boron concentrations.

The only known source for such high concentrations is the Boric Acid Tank via the Boric Acid Blender.

The theory was that the 1A FW Recirculation pump that was running at the time (to facilitate the taking of the FWST sample) had been damaged to the point of not being able to develop a discharge head.

This was suspected due to the belief that the running FW Recirculation Pump had been in service for approximately three days with the suction isolated; suction valve IFW49B was

~.- - -

DUKE p0WER COMPANY / CATAWBA NUCLEAR STATION PIR 1-C91-0304/Special Report

, Page 13 found to be failed in the (nearly) closed position on July 30.

This theory makes an assumption that the damaged pump's discharge check valve allowed some backloakage to the NM sample point.

The presence of elevated boron concentrations in the FWST makeup / recirculation piping is believed to have occurred as a result of changing valve alignments following a decision to begin an orderly plant shutdown.

Control Room personnel attempted to add boric acid from the Boric Acid Tank via the blender to the VCT to begin the shutdown process. Valves from the blender to the FWST makeup line were open from the anticipatory alignment made earlier in the event for a possible boris acid addition to the tank, which was never undertaken. This alignment may have introduced a small quantity of boric acid of Boric Acid Tank concentration (approximately 7000 ppm) into the FMST makeup / recirculation line l

which mixed with and took.the path of least resistanca - towards the NM sample point. When control room operators determined that the valve alignment would allow boric acid to flow towards the FWST recirculation line in addition to the intended VCT destination, actions were initiated to isolate boric acid addition l

towards the FWST and to-redirect flow to the VCT for the shutdown process.

This evolution occurred on July 20, af ter FWST Recirculation pump B was secured, FWST Recirculation pump A was placed in service, and loop flow indication was lost.

The presence of-low boron concentratiens at the NM sample _ point which initiated the sampling event is unlikely to have occurred in the same manner. On the evening of July 24, when the initial FWST makeup was made, the final 1000 gallons of solution directed to the FWST was in the form of pure dilute water from the Reactor Makeup Water Storage Tank.

This is done by procedure to flush remaining acid from the blender piping. This evolution occurred during a time when FWST recirculation loop flow indication was operable and indicating flow greater than 27 gpm during sampling.

This theory was evaluated by performing testing of the Recirculation pumps in question and their respective check valve after completion of repairs to 1FW-490.. Results indicated that both FW Recirculation pumps were fully operable and delivering their designed head and flow. This testing also showed that both pump discharge check valves were working properly at the time of testing.

In summary, the only explanation of these events which corresponds to observations and data are A)

Prior to 2300 hoars on July 25, yM intrusion into the suction line occurred and d'. luted the FWST sample boron concentration.

(FWST recirculation loop flow was greater than 27 gpm during this time.)

IFW-49B was at least partially open during this time.

B)

. Af ter 2300 hours0.0266 days <br />0.639 hours <br />0.0038 weeks <br />8.7515e-4 months <br /> on July 25, YM had been isolated, FWST recirculation f)ow indication was lost, and Boric Acid Blender intrusion occurred into the sample line via reverse flow from the pump discharge header.

The' position of IFW-49B at this time is not known with certainty but it is believed to have oeon closed.

i l

i

~

_.m._ _ __

DUKE POWER COMPANY /CATAtfBA NUCLEAR STATION PIR 1-C91-0304/Special Report Page 14 As a result of this event and investigation, the seal loop intrusion source will be cut and capped off.

Chemistry procedures will be changed to require r

verification of FWST' recirculation flow, by one of three specific means, prior to each sample.

(OAC data point (preferred method), recirculation pump delta-pressure, or local flow gauge reading verified by momentary pump shutdown and restart are the methods to be used.)

Until these actions are fully implemented, local samples will continue to be used to verify FWST T/S i

requirements are met. -Chemistry procedures are being revised to indicate the FM l

pump as the alternate FWST sample point.

CORRECTIVE ACTION IMMEDIATE 1)

FWST sampling frequency was increased.

'T 2)

FWST samples were taken from alternate locations.

3)

Acceptable FWST boron concentration was verified.

4)

Unit shutdown was begun (and later stopped).

SUBSEQUENT 1)

FWST sampling was begun from alternate locations independent of the recirculation loop.

2)

InvestigatlOn was begun to determine the cause(s) and source (s) of FWST sample dilution /boration.

3)

Valve 1FW-49B was repaired, tested, and returned to operability.

4)

The performance of the FW Recirculation pumps and their associated discharge check valves were tested and verified to be acceptable (ape).

5)

The integrity of the NM sample system was investigated and determined not to be the source of the dilution /boration.

6)

The potential for YM intrusion into the NM system was investigated and

-could not be replicated.

7)

MOVATS procedure has been revised to prompt reporting of stalled motor situations to Techical Support and DE for evaluation, PLANNED 1)

The seal loop intrusion source will be cut off'and capped.

2)

Chemistry procedures will be revised to require verification of recirculation flow during EwST sampling ( APE).

. ~....

DUKE p0WER Cor4PANY/CATAWA NUCI. EAR STATION p1R 1-C91-0304/Special Report

, Page-15 t

l 3) local sampling of the PWST will continuo until the above actions are

{

completed.

4) _ Chemistry procedures will be reviewed and revised as.noeded to specify the location and method of collecting alternato sample (APE).

l 5)

Chemistry will comparo local samples with NM samplos to verjfy consistency (APE).

6)

Operations will evaluato the adoquacy of instrumentation used in verifying recirculation flows, samplo flows, diccharge pressure, etc.

and consider adding them to PH programs-or periodic calibration checks (APE).

SAFETY ANAINSIS t

The FWST boron concentration was shown to be acceptable at all times and capable.

of performing its intended function. Thus, the assumptions of the design basis safety analysis were maintained.. The health and safety of the public woro maintained.

t

)

b I

l f

i f'-

1

. =,.. -

...,.,_.,.,,,,.,,2-..,.-----.