ML20011F209

LER 89-026-00:on 891128,determined That Particulate Levels in Samples Taken from Lower Third of Tanks Exceeded Allowable Limits.Caused by Inadequate Sampling Technique. Tanks Cleaned & Filled W/Clean fuel.W/900227 Ltr

ML20011F209
Person / Time
Site:	Calvert Cliffs
Issue date:	02/27/1990
From:	Denton R, Muth D BALTIMORE GAS & ELECTRIC CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-89-026, LER-89-26, NUDOCS 9003020118
Download: ML20011F209 (6)

Text

y r m l-c 't -

j BALTIMORE ,

3- GAS AND t ELECTRIC l m

CHARLES CENTER

• P.D. BOX 1475

• BALTIMORE, MARYLAND 21203 1475 l

E CALVERT CLIFF 8 NUCLEAR POWER PLANT ptPAntutNT

[$$$$$"EE" *"' " f February 27, 1990 U. S. Nuclear Regulatory Commission Docket No. 50 317 p Document Control Desk License No. DPR 53 Washington, D. C. 20555

Dear Sirs:

The attached LER 89-26, Revision 0, is.being sent to you as required under 10 CFR 50.73 guidelines.

Should you have_any questions regarding this report, we would be pleased to discuss them with you. -

Very truly yours, W

['

R.k.Denton Manager -

1 DWM/1r cc: William T. Russell Director, Office of Management Information and Program Control Messrs: G.-C. Creel C. H. Cruse J. R. Lemons

[ L. B. Russell R. P. Heibol J o Ibo1W 9 /

,oo39penMTyc7 W

'/f PDR 5 i

l n

W Peren'gg

1. 83' W 8 NUC441A kl.ULatomy C0anniesi0ss i t PPR0vt0 0M8 NO $1tikelm LICENSEE EVENT REPORT (LER) ' " N 8 ' '8' "

l S&CILif V NAME H) , DOCK t1 NUM$t h (si FAM J Calvert Cliffs, Unit 1 -

o 9 5 ; o , o ; o ; 3; 1,7 titL4 tal i lorl 0 l5 4

Potential Emergency Diesel Generator Failure Due to High Particulate Level in Fuel tytNT Datt sei sta Nun. eta esi alp 0nt DAtt its OtMan f ACILivits wvskytD (8, MONTH DAY vlam vtAR NNN'6 [*fy*,$ MONtw Day vtAm '

• Cikit , have s 00Ck ti Nvwsta:Si )

Calvert Cliffs. Unit 2 0 l610;oIo;3l1 l8

~ ~

1l 1 2l8 8 9 8l9 0l 2l 6 0l0 l l l D isto ic io s I ;

)

,,,,,,,,,, tais s tP0nv is eveMirsto PunsuA=v 1o vut at0vintMENn 0, a Cea i ,c... ., , . ., ra. ,...-,, nu "o** #8 5 no ora,i no .m.i u n nnn . _

ntu i  :

g _

n asei.nuin no mi.im u n.nsn.i n.tu.i no, Ot 40 m assi.nuti,e u mi.nri

= n.nin .

T_ gt,utgnege7.<e n .ni n i., n.ni.nzna ni.nin. .nAi mA,

] so .oei.in ni.i 30 eeli. lit n.1

] u ni. nan.i 60 t34.llinnd

[ nni.nsn..wei 60 r3L.nnn.1 Voluntary Report tiCIN$tt CONTACT 80m THis 4tsi ne NiMt tittPuoht Nyveta ehtaCODI D. W. Muth - Licensing Engineer '

3 10 l1 21 610 l- 1 3l59] 1 COM* LITS ONE teNG 70m E ACH COMP 04 TNT S Aitutt OtsCaist0 IN twas atPont Hal SvttlM MA C a omia 48 '*

C?U$t COMPONINT Cav$t t vit t M COMPONENT N C yo pad I l l l l l 1 I I I I I I I l I l l l I l i l l ? l l 1

$UPPLEMINt At atP0mf E kPICilD H4i MCN?w Day vtan Sv9Yr5310%

vt 5 (H v.a r.mos.n iketCttO $90w$$ TON OATil NO l l l Aesn AC, su~, u. sex Ma., s . aan.mw., e.n . ,. u r. r.-n.. n. s,.w u s, On November 22, 1989, samples of fuel were taken from both Fuel 011 Storage Tanks (FOSTs) in accordance with Technical Specification 4.8.1.1.2.b. The samples were taken from the top third, middle third, lower third and bottom of each tank. The samples were analyzed and on November 28, 1989 it was determined that the particulate levels in the samples taken from the lower third of both tanks exceeded allowable limits. After careful consideration of the fuel filter design, the alarm setpoint, the operating history of the filter, and the other controls, we determined this item not to be reportable.

Both tanks were emptied, cleaned and filled with clean fuel oil.

) Fuel sampling of the FOSTs will include taking an extra sample from the bottom of the tanks in addition to the three presently taken. All sample analysis results, including trends, will be provided to the system engineer.

A review will be conducted to determine the optimal amount and frequency of biocide addition to prevent biological growth in the tanks.

. u.

7,,e,,,

1 L

l wne

. pea u: wuctsu c.t:vmony coMMesion  ;

• LICENSEE EVENT REPORT (LER) TEXT CENTINUATION maovse oue no mo-oio. l t oints.e si es j paCsLif y haut its DOCetti Wumeta GI Lin NUMeth (6p P&Of 13>

Calvert Cliffs, Unit 1 .... ..... . .

~~  :

"*"* l o l5 l0 lo lo l 3l1 l 7 8l 9 -

0 l 2l 6 -

0[0 0l2 0 F. 0 l5 i taxiIn . = = ._ we w.muvon I

I DISCUSSION On November 22, 1989, while Unit I was in Mode 5 and Unit 2 was defueled, i samples of fuel were taken from both Fuel Oil Storage Tanks (FOSTs) (EIIS DC TK) in accordance with Technical Specification 4.8.1.1.2.b. The samples were taken from the top third, middle ' third, lower third and bottom of each tank.

The samples were analyzed and on November 28, 1989 it was determined that the particulate levels in the samples taken from the lower third of both tanks i exceeded allowable limits. Table 1 of ASTM Standard ANSI / ASTM D974 81 lists the maximum water and sediment for No. 2 D Crade Diesel Fuel Oil as 0.05 percent by volume. The nonconforming sample from FOST Number 11 contained 3.0 percent particulates and the nonconforming sample from FOST Number 21 contained 2.8 percent particulates. No water was found in the samples.

The control room was notified at 1:30 p.m. on November 28, 1989. The Shift Supervisor immt diately commenced actions to verify containment integrity in accordance with Limiting Condition of Operation (LCO) 3.8.2.2. He ensured that the action requirements of LCOs 3.1.2.3 and 3.8.1.2, which require suspension of core alterations, were met. Upon notification of the condition, the Shift Supervisor initiated a reportability review in accordance with Calvert Cliffs Instruction C01 118. At 9:00 a.m. on November 29,1989 he determined that this condition required a 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> telephone report in accordance with Title 10 Code of Federal Regulations (10 CFR) paragraph 50.72 (b)(2)(iii). It was determined on December 6, 1989 that this condition was not reportable under 10 CFR 50.73. After further review, we have determined that a voluntary report is appropriate.

A laboratory analysis of the particulates in the FOSTs determined them to be normal bottom sediment. The analysis concluded that the sediment consisted of dissolved metal material from normal degradation of the tank lining, dead biological matter from the annual use of a blocide and normal fuel by products from the effects of temperature changes on the fuel. The amount, type and location of the sediment indicates that it built up gradually over a long period of time.

Fuel in the FOSTs is recirculated monthly. The fuel in the tank had been recirculated and new fuel had been added to the tank within two days prior to its being sampled. This caused the bottom sediment, which had not previously been detected at the level at which the samples were taken, to rise above the level of the outlet line (EIIS DC-0UT) from the tank. The sediment settled back below the outlet within 3 days. During this time, any fuel transferred from either of the FOSTs would have had a high sediment level.

Po,M 3esa ,

g". vi =vea m u v aron ,cou. .o=

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION

<*aovtoovewo n . m LKPint$ S'3000 faceLety hanet og DOChlt NUM8in tal gga ggugga g paog m vtan 68 0V8 h'44 ** v 8'DN Calvert Cliffs, Unit 1 "*"" "*"*

~ ~

rnn w==.e m e = = 1 , : _ me n ,muw nn f '

Monthly recirculation of the FOSTs has been performed since September 9, 1985.

Since that time, there have been approximately three days per month when contaminated fuel could have been passed to the day tanks (EIIS DC TK) . High '

particulate levels were not found in the day tanks when they were cleaned and examined in early December, 1989.

Fuel is added to the tanks intermittently. This also disturbs bottom sediment.

It is probable that in this case, the combination of recirculating the tanks and adding fuel disturbed the bottom particulates more than usual.

One reason this condition was not detected earlier is that, prior to July, 1989, FOST fuel oil samples were taken from the tank drain line (EIIS DC DRN) . The prucedure governing these actions did not require a sufficient amount of oil to .

l be drawn off to provide a sample of oil in the tank rather than oil in the drain ,

line. Suspended solids in the tank which did not enter the drain line would not be detected. The sampling procedure was revised in July, 1989 to require I samples to be taken from within the tank itself. The sample tested in August, l 1989 was within ASTM limits. A sample taken on November 28, 1989, after the l particulates had settled back to the bottom of the tank was also within ASTM j limits.

CAUSE OF CONDITION The particulates in the tanks were normal bottom sediment which had built up l over time. They had not been discovered previously due to inadequate sampling technique.

l ANALYSIS 1

! The POSTS supply fuel to all three Emergency Diesel Generators (EDGs) (EIIS EK DG). The suction pipe, located approximately nine inches above the bottom of the POST, leads through a redundant header to one of the three motor driven fuel transfer pumps (EIIS DC P), each of which is associated with an EDG fuel oil day tank. Each day tank holds enough fuel (485 gallons) to operate its associated EDG for 154 minutes at 3000 kilowatts. Fuel demand at this load is

. 3.15 gallons per minute. From the day tank, fuel oil runs through a 5 micron duel strainer basket filter (EIIS DC FLT) on its way to the EDG. The basket assembly is of the duplex type which permits flow through one filter at a time.

In the event ora filter becomes fouled, the operation may be switched to the other filter without interrupting EDG operation. Had dirty fuel passed via the day tank to any EDG, the filter would have removed all of the sediment.

It is possible for particulates to build up on the filter to the point that they restrict fuel flow. The fuel line is instrumented (EIIS DC PA) to provide g*ow ms .c .

tpAC Seem agha

" u 8 IsuCatG ht";ut? tony consaslepto,e

• LICENSEE EVENT MEPORT (LEM) TEXT CNTINUA110N **aovio ous no sino-oits I APihtS Sial mi f ACl4th miAast 49 ppcggg gg,,,,, ,

Calvert Cliffs, Unit 1 "M@t'  ?#.0 recin . u:: m e, ,,w.,nn l l 0l0 nl4 W nl5 indication of low fuel oil pressure in the event that a fouled filter is cutting off the flow of fuel to the EDG. This signal feeds into the remote EDG combined trouble alarm provided in the control room. The local fuel flow alarm, based on the pressure change across the filter, is set at 10 psi. This value is half the manufacture's recommended setpoint of 20 psi.

Should the EDG combined trouble alarm sound, an operator would be dispatched to the EDG room, where the low fuel pressure alarm would be sounding, indicating a dirty filter. The operator would switch flow to the other filter basket and initiate actions to have the dirty filter replaced.

There may be a chance that the filter could back up quickly enough to prevent an operator from switching flow to the other filter basket before flow is cut off to the EDG, However, without sufficient data on how fast the filter would have clogged, it is not possible to determine whether or not an operator would have arrived in time to change the filter before fuel flow was constricted to the point that the EDG tripped. Excessive particulates were not found in the day tanks and hence , did not reach the filters. A review of the operating history of a recently replaced falter found that impurities had built up gradually over time and that there had been no sudden increase in the amount of sediment collecting on the filter.

After careful consideration of the filter assembly design, the low alarm setpoint, the operating history of the filter, and the other controls described above, we determined this item not to be reportable.

CORRECTIVE ACTIONS Both tanks were emptied, cleaned and filled with cican fuel oil.

Future sampling of the POSTS will include taking an extra sample from the bottom of the tanks in addit'on to the three presently taken. The bottom sample will serve as a trend indicator.

All sample analysis results, including trends, will be provided to the- system engineer.

Biocide is presently added to the FOSTs annually to prevent growth of biological matter. A review will be conducted to determine the optimal amount and frequency of biocide addition to prevent biological growth in the tanks.

grow ma .,

i itellt #wo seta W s NuctiA2 78;uL.*.tomy consemessose )

• UCENSEE EVENT REPORT (LER) TEXT CONTINUATION ***xovieove=a seo-eio. j axmts: s sus 54C644TY Isaagg gg, N E1N M th m tik NuhsSIR 16) P&OE (3s k'

,gan stoutwting moon Calvert Cliffs, Unit 1 "'**'* """

0 l6 l0 l0 l0 l 3[1 l 7 8l 9 0 l 2l 6 -

0l0 ol5 0F 0 l5 rixin . % wwws mmavem 1

l \

l ADDITIONAL INFORMATION There have been no similar events at Calvert Cliffs which have been described in LERs in which fuel oil quality has threatened operability of the EDGs.  ;

i e

g ,. o.. ... ..

. . . - . - - - - -