ML18079A952

Deficiency Rept:On 780928,during Routine Design Review, Undersizing of Refueling Water Storage Tank Was Discovered. Tank Storage Capacity Has Been Increased.Design Drawings for Tank Piping & Foundation Will Be Modified

ML18079A952
Person / Time
Site:	Salem
Issue date:	10/28/1978
From:	YANKEE ATOMIC ELECTRIC CO.
To:
Shared Package
ML18079A951	List: ML18079A952
References
NUDOCS 7909240038
Download: ML18079A952 (4)

Text

. .

ATTACHMENT Report* oo the Seabrook Statioo ....

• Refueling ~ater Storage Tank Design Deficiency

~*..,

Prepared by: Yan..~ee Atomic Electric CoEpaoy for Public Se:x:vice Cowpany of Ne~ Hanpshire October 28, 1978 790924 O<Y~'b

Introduct ~on I

On Septe!Lber 28, 1978 Mr. Seth Folsom of the Region l office of the ~RC vas inf or-;::ed by Mr. John Haseltine of Yankee Atomic Elec~ric Coc?3oy, Seabrook Station Project Hanager, of a design defic.ieoc.y assod. .ated 'Jith the refueling Yater storage tap~s (?.~ST) at Seabrook Station. - The deficiency

.. as reported under 10CFR50.5Sa(e) and ... as reported Yithin 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of con£ iri::a tion of the deficiency. !he deficiency 'l.iaS reported to Mr. Folsom as an undersizing of the RWST.

The undersizing ..,as discovered in a routine design revie',;o of the Net Positive

_Suction Read (NPSH) for the containment spray and residual heat reooval I

pm:ps. The design r~vie'l.I' verified proper ~iPS'H but it also sho.,.ed that there ~*

did not appear to be sufficient ..,ater inthe RWST to col!lplete the traI!Sfer  ;

of purq> suctions from the tank to the cootainnent suop before it vas e;:;ptiedc

[_

This report ser-ves as a '.1ritten :report under the 30 cay req uire!t-cnts in 10CFR50.55a(e)(3) and is arranged t-0 provide the infor:::.ation requested ~

in the regulation. I Descrintion of Deficiencv I

r The vater stor-age capacity of the RWST serves dual purposes in the I station design. One is to flood the cavity and canal area above the reactor prior to refuelings and the other is to supply bor-ated ~ater to the containment spray and injection puw;:is during a less of coolant accident.

As reported to the IE Inspector on Septe::;ber 28, the deficiency concerns tl~e undersizing of the RWST's useable volume for a loss of coolant accident.

The tank ~as adequately sized for the flooding of the re.actor cavity and c.a.nal.

The tank capacity is dependent: upon a *nu~er of parameters. The major para~eter is the flooding of the containn:ent during a loss of coolant:

accident to a height *.Jhic.h Yill provide sufficient NPSH. for the contaim::ent spray and residual heat removal pumps and proper containment sucp screen submergence. The 375,000 gallon design submitted in the Preliminary Safety Analysis Report (?SJ...R) met thls requirement. Approxiwately 350,000 gallons r

I of the tank ~as assumed to be injected and this volume did alloY sufficient:

flooding of the contairu:;:;e.nt to meet pump NPSH requireoents and screen coverage. The deficiency arose in the ability of the rewaining 25,000 gallons in the tank to meet the other design parameters.

The other design paraoe.ters are instrument error, a vorking allo~ance, transfer alloYance, single failure, and unuseable voluoe. The f ollc*ing is an explanation of vhat these parameters are and ho~ they vere satisfied or not: satisfied in the original design of the R~ST.

1. Instrusent er~or - Increased volume is necessary to account for tbe accuracy of the instru~ntation reading the R~ST level. Initially a nominal 1% full scale instru~~nt accuracy error ~as assur..ed ~hich trar£lates into 3500 gallons. This error h2..S to be taken t~ice - once

. I"*-

.4 to 2s;;_ire that the tank does initially contain 350,0C'O ::;2llo::is anc a sec.ond time to assure that 350,000 gallons has b~cn ~.'.")jE:cte~. r:-:u s' 7 ,000 gallons of capacity ...*as necessary for i.::s ::.ru=e: n::. error.

Recently, ~e *ere infon;ied by ~estinghouse, ~~o is res?onsible [or the RWST level instrumentation, that they prefer a 3~ full scale ins trut:1ent error ...,hich causes an increase in the allo*-*ance for instruoent error to 21,000 gallons.

2. Yorking allovance - Some allo...,ance above the required capacity of 350,000 g2llons is necessary to prevent alar.::s ~1th only minor ~ater losses from the tank. It ...,ould not be practical to maintain a level at the e.y~ct setpoint and thus, a na.:::inal 3,000 gallons over and above the design capacity is desired for this ::iargin.

3. Trar.sfer allo*ance - Additional ~WS! capacity must be provided to accomr:Ddate a reasonable delay ti~e associated vith the trar.sfer of certai.n pump suctions from the RWS! to the contairu:;ent sump. As stated in RESAR-3, "w-hen the prop-er quantity of *.:ater has been trap.sferr-ed--* _,

to the contain=ent, the cootain:::ent suu:rp valves open. This au to~ tically sh i f ts RHR and contain~ent spray pu:::;p suet ions to the

• containment su'Op. Ho"'*eve r, the safety inject ion and charging pu:::ips continue to draw from the RWST.If you assuDed a 10 r:llnute realigo~eot time, the allo*ance would be 24,000 gallons since the flow rate of the pumps remaining on the RWS! is 2,400 gallons per ninute (gpm).

4. Single failure - Because of the autocatic trar£fer of R...~R and contail:lUent pucps to the Su::lp, sone allo~ance oust be included for a single failure. The most limiting single failure for tank capacity r_esults if the contain:;;ient suu;p valve for one of the trains failed .

~o open upon a low level signal from the RWST. Should this ...,.ors t singl~

failure occur, the associated R..~R and containment spray pumps in "that train 'IJill continue to dra~ from the RWS! at a rate of 7800 GPM. If you assume it takes 5 minutes to recognize and correct the condition, then an additional. 39,000 gallons ..... ould have to.be added to the tank to compensate for the single failure.

5. Unuseable volume - Once the i-:::J.ve rt of the tank pump suet ion pipes are reached the pumps lase suction and arry rewaining vater in the tank is unusable. This unusable volu~ ~ust be included in the design calculations and for a 375,000 gallon tank it is approximately 14,000 gallons.

The design parameters for instrm:ient error, transfer allow-ance and single failure have changed since the original sizing of the tank. The following tabl~ is a su!Iill'.ary of required capacities under the old and revise design parameters:

Design Para=eter

• Requi:-ed Old r:',...1S 1 Vo l'Jme

!~ e'wl

1. Injection C3pacity 350,000 gal 350, 000 gal

2. Instruoent error 7,000 gal 21, 000 gal

3. ~orking allo~ance 3,000 gal 3,000 gal

4. Transfer allo*.:ance 24, 000 gal

5. Sing],_~ failure. 39 000 gal I

6. Unusable voluce 14' 000 £al 14.000 ~al Total 374,000 gal 451,000 gal As sho~n by the table, the original 375,000 gallon tank "t.;Ould have met all the old des!gn parameters but is substantially undersized for the n~ or revised para;:;eters.

Analysis of Safetv IllrDlications If the tank size had rec.ained at 375,000 gallons, i t is possible that the puwps t.::king suction from i t "t.;Ould have nm out of 'l.:ater and lost suction. The safety injection and charging pun;ps are multistage pu<=?S and upon less of suc~ion may cease and beco~e i~operable. Thus,* the path from the discharge of the residual heat re:::ioval pumps -::hrot!gh these pumps r::.ay have been lost. Bo~ever, even 'wlith ~erst single :ailure for tank sizing, the containment spray and residual heat rei:;:ioval patbs into the reactor coolant system and containment respectively ~ould have re~ined operable and ~ater could have been supplied to the core and containwent at:;iosphere.

No accident analysis ~as made of a loss of coolant accident under these conditions since the design has been c~anged and these conditions are no lo~ger credible.

Corrective Action The R~ST storage capacity has been increased from 375,000 gallons to 475,000 gallons. This larger size is more than adequate for all the design considerations and leaves some raa.rgin for possible future cha~es.

No vork on the tank foundation, piping or erection has started and

- thus, no field uudification needs to be taken to enlarge the tan..~. P.o~ever, design drawings for the tank, piping and foundations ~ere COQplete and certified for construction and ~11 ba.ve to be codified. The spray cheoical addition tack ~ill also have to* increase in size proportional to the R~ST or the concentration of checicals in it will have to increase to co~per~ate for the RWST e~2a~~~went. A decision on ~hat ~111 be done to the spray c:iei:;iical addition tank ":oJill be r...ade in the near future and docm::en-::ed in the final Safety Analysis Report.

• ----.--- -*--~-----------