ML20087N392

Forwards Responses to Draft SER Open Items 64 & 67 & FSAR Questions 430.97,430.100 & 480.26.Responses Will Be Incorporated Into Upcoming Rev 6 to FSAR

ML20087N392
Person / Time
Site:	Beaver Valley
Issue date:	03/28/1984
From:	Woolever E DUQUESNE LIGHT CO.
To:	Knighton G Office of Nuclear Reactor Regulation
References
2NRC-4-032, 2NRC-4-32, NUDOCS 8404030407
Download: ML20087N392 (10)

Text

{{#Wiki_filter:. o 2NRC-4-032 (412)787 - 5141 Telecopy 8 Nuclear Construction Division March 28, 1984 Robinson Plaza, Butiding 2, Suite 210 Pittsburgh, PA 15205 United States Nuclear Regulatory Commission Washington, DC 20555 ATTENTION: Mr. George W. Knighton, Chief Licensing Branch 3 Office of Nuclear Reactor Regulation

SUBJECT:

Beaver Valley Power Station - Unit 2 Docket No. 50-412 Open Item / Question Response Gentlemen: This letter fo rward s res pons es to the issues li s ted below. Duquesne Light Company plans to incorpo rat e the res ponses to the FSAR que s tions into FSAR Amendment 6. The following items are attached: At tachment 1: Res pons e to Open Item 64 of the Beaver Valley Power Station Unit No. 2 Draf t Safety Evaluation Report. At tachment 2: Response to Open Item 67 of the Beaver Valley Power Station Unit No. 2 Draf t Safety Evaluation Report. At t achment 3: Response to FSAR Questions 430.97 awl 430.100 (Power Systens Branch) forwarded by your letter dated September 19, 1983. : Res pons e to FSAR Ques tio n 480.26 (Cont aiment Sys tens Branch) contained in your letter dated January 9,1984. DUQUESNE LIGHT COMPANY By E. ld. Woolever Vice President KAT/wjs cc: Mr. H. R. Denton, Director NRR Mr. D. Eisenhut, Director Division of Licensing Mr. G. Walton, NRC Resident Ins pector Mr. M. Lacitra, Project Manager SUBSCRIBED AND SWORN TO KFORE ME THIS g?fM DAY OF 77//4cd , 1984. Ab 1 Not ary Public ANITA ELAINE REITER, NOTARY PUBLIC ROBINSON TOV!NSHIP, ALLEGHENY COUNTY ) f.tY COMMISSION EXPlRES OCTOBER 20,1986 8404030407 840328 PDR ADOCK 05000412 E PDR

t Unit ed Stctco Nuclear Rrgulttory Commission Mr. G2crgs W. Knighton, Chief Page 2 COMMONWEALTH OF PENNSYLVANIA ) ) SS: COUNTY OF ALLEGHENY' ) On this 4fd day of Neff /(([ , before me, a Notary Public in and for said Commonwealth and County, pe rsonally appeared E. J. Woolever, who being duly sworn, deposed and said that (1) he is Vice President of Duquesne Light, (2) he is duly authorized to execute .and file - the foregoing Submitt al on behalf of said Company, and (3) the statements set ' forth in the Submittal are true and correct to the be s t of _. his knowledge. Y c' Notary Public ANITA ELAINE REITER, NOTARY PUBLIC ROBINSON TOWNSHIP, ALLEGHENY COUNTY MY COMMISSION EXPIRES OCTOBER 20,1986

1 p. ATTACHMENT 1 Response to Open Item 64 of Beaver Valley Power Station Unit'No. 2 Draf t Safety Evaluation Report Drafc SER Section 7.3.3.12 Steam Generator Level Control-and Protection Three steam generator level channels are used in a two-out-of-three logic for isolation of feedwater on high stema generator level. One

of the three level channels is used for contr ol.

This design for actuation of feedwater isolation does not meet the. requirements of Paragraph ' 4.7 of IEEE '279, " Control and Prot ect ion Systen Int e r-act ion," -in that the failure of the level channel used for control

could_ require prot ect ive - action and the remainder of the pr ot ect ion system channels would not satisfy the single-failure criterion.

The applicant has not responded to this concern. This is an open iten. E

Response

Y' - Each steam generator (S/G) has three protection grade level transmit-lters (LT), each of which fe ed s a Hi-hi bistable (B/S) and a Low-low ' bistable. One of the' LT's, 2FWS*LT476 (on S/G A, for example), also provide's a signal ' to the feedwater control system. Section 4.7.3 of _IEEE Std. 279-1971 states in part : O "Where' a single random ' failure can cause a control sys tem I. ' act ion that results in a generating ~ tation condition requir-- s ing. protective action 'and can also prevent proper action of a prot ect ion system channel designed to prot ect agains t the i ~ condit ion, the remaining redundant' protection channels shall be capable.of providing the prot ect ive ~ act ion even when degraded by a second random. failure." In pr evious. FSAR review meetings, the ICSB reviewer has expr es sed concern that a failure of LT 476 'would cause a transient for which protection - would not be prov ided. This~ is not a concern if the LT fails'high because the ' core is prot ect ed agains t low level ' by the diverse - trip! on high steam / feed irate conincident with low S/G 1evel from ' the other LT's. The. concern arises in the case that the trans-mitter fails ' low and causes the S/G to overfill. A safety analysis of '?Feedwater System Malfunction Causing an' Increas e in Feedwater Flow" is presented in the Beaver Valley Unit 2 Final Safety Analysis Report. ~ IEEE Std. 279-1971 is not - applicable - to this portion of the sys ten design 'because this condition does not require - protective act ion as the core remains above the DNB limits (see FSAR 15.1.2.2). Addition-ally, ' the plant has.been analyzed for' steam line break accidents (FSAR 15.1.5) and 'the DNB ' design -limit is met in this case' also. Therefore, the concern resulting ' from the overfill of the S/G is one of protec-tion 'of the. turbine, a commercial concern. As IEEE 279 is applicable to ". ' the ' safety-related ' functional performance . o f pr ot ec-t ion sys tems '...," i t is not applicable to this function. Neve rt he-

less, the.. standard Wes tinghouse - design -for ' this system uses three h igh-quality, safety grade level transmitters to provide indication f

cnd S/G cutomrtic initiation funct ions. In the m:ny reactor-yeara' experience of operating-this de s ign, no known eve nt of S/G overfill has occurred. Still, in order for a steam generator overfill event to occur, several very specific failures must occur in a very short period of time. The initiating failure must be that of the controlling ch annel (LT 476) failing low - such that the S/G 1evel begins to rapidly rise. Should the other two channels be operating prope rly, the hi-h i S/G level turbine trip and feed water pump trip would function to tenninate the t rans ient. Should one of the ot he r ch annel s fail high, the same action would occur when the operating channel reached the hi-h i setpoint. If one of the other channels failed low, the reactor would be tripped by the low-low S/G 1evel trip and feedwater would isolate 14 seconds later on low T avg. coincident with P-4. One case can be postulated where automatic sys tem prot ect ion is not provided. In orde r fo r this to hap pe n, two very specific failur es would need to occur within a sho rt time pe riod. The controlling channel (LT 476) would have to fail low causing the fe ed regula t ing valve to open and S/G to fill, and the high level bistable on one of the other level ch annels (LT 474 or 475) would have to fail to turn of f upon receiving the appropriate analog signal. The failure of the high level bistable in the "on" position would be extremely unusual. If these two specific failures were to occur, then it is conceivable that without operator action, a steam generator might overfill. However, this rare set of failures is protected against by the follow-ing process. Beaver Valley Unit 2 Technical Specifications require a channel check every 12 hours which would detect any channel fa ilure. Addit ionally, the technical specifications require a channel func-t ional tes t every 31 days to de t ect any inoperable bistable.

Thus, the above described scenario of very specific compo nent failures and failure modes must occur in a short time frame res ult ing in an extremely low probability event.

Even considering this case, this event is very easily di'hgn os ed. Tab le 1 li s ts ala rms which would actuate as a result of this transient and Wes tinghouse analysis shows that the operator has gr eater than ten minutes to isolate fe edwater before steam generator overfill could occur. Since this event is a standard malfunction used in reactor operator training courses, it is apparent that this ten minute time span is su f ficient for ope rato r act ion and is cons is tent with othe r as sumptions made in safety analyses in the Beaver Valley Unit 2 FS AR. In conclusion, IEEE Std. 279-1971 is not ap plicab le to the S/G high level event. The analysis presented in the Beaver Valley Unit 2 FSAR demonstrated that there is adequate core prot ect ion agains t DNB fo r excessive fe edwate r. flow trans ient s. In ad di t ion, recent pla nt-specific analyses conducted by Westinghouse on double failures in the Steam Generator Level Control and Protection Systems for Beaver Valley Unit 2 have shown that, when one cons ide rs the trans ient res po ns e, including actuation of other protective fun ct ions, the protection and control systems design of Beaver Valley Unit 2 prov ides adequa te protection against exces sive feedwater flow transients resulting in a steam generator overfill event.

TABLE 1 t TIME SEQUENCE OF ALARMS AND ANNUNCIATORS FOR A FEEDWATER CONTROL MALFUNCTION WITHOUT REACTOR TRIP Core Condition Event -Time (sec)_ 1. Beginning of Life Bistable 476 A C Channel 476, lo-lo SG 1evel 0 Low level deviation alarm 0 Feedwater Control Valve fully open, loop 1 9 Channel 475, hi-hi SG 1evel 14 3 Bistable 475C 14 3 2. End of Life Bistable 476 A O Channel 476, 10-10 SG 1evel 0 Low level deviation alann 0 Feedwater Control Valve fully open, loop 1 9 Channel 475, hi-hi SG 1evel 14 6 Bistable 475C 14 6 r y sr -e, e,,-, e .e,- --.v = r -.-,4 e-,,--

T. ~ ATTACHMENT 2 Response to Open Item 67 of Beaver Valley Power Station Unit No. 2 Draft Safety Evaluation Report - Draf t SER Section 7.3.3.15 Power Lockout for Motor-Operated Valves Cert ain motor-operated valves, such as' those for cold-leg accumulator isolation, require power lockout (removal) to meet the single-failure criterion. The power lockout scheme used by the applicant uses ' an additional, manually controlled (via - removable banana plugs) contac-tor. The staf f has concluded that a short or relay failure in this circuitry could cons titute a nondetectable failure and thu s violate the single-failure criterion. The staf f has expressed this concern to the applicant and cons iders this item open subject to its review of the applicant's pending response.

Response

The circuitry for motor control on these valves will be modified as shown on the attached figure 12241-E-6HN to provide a white indicating light powe red from the secondary side of the slave contactor control t rans fo rmer. 'Under normal plant. operating conditions, this white light will be extinguished. When power is applied to the slave - cont actor, the light will -be energized indicating that powe r is available.

I l l } -l 6, l $$ runw4(c.iiAss e) .to.'mc. w,,wt ~ ' ' 43d e t im e r ownca war w.i c avuop=o y 3' A, s s A 3 C Je, _ VALVE POS*TiOp4 ,'~~'~~~~ "~~' - 0 c \\,, %' A 5 N " ' * = a * *'5 n hs m, _ ,/,t. gt ~ e sm.

i. - - - - -.,

.). ). ), .r - or

44.

3 estEN LT r **.s owet. EO e s 4 cPE N t.imiT I g O Rt. _I D _M,, O L] M L8.,.I. N g g Q M ,swas '~ 4 6 e -.is-in. s-in. .. iA - is-4 2 3

• we-s A9-

.W-.g68 i onus u:^ ::[" CLD3G f n t1C T -~ '~~ or n* y y - - e, os vor og ,-on g t.o.. z ,.ts, _4._ c s-a s'

• MM v=4 v

.e 0 5 es ao e a. ,,,,,S..,,,, t,1 tre ue 4~ILat5 L.*dos- _2. . _it a inf er"-

• CT e

ig -{~ - _6C ~ ~6 ~E==ia Tr a ni,cAWaF1 ffir'7MA A) 4 5 g C3- 7 I - sr. WeF. IIE[f} 8' h ei .q' o+ 42 4een cry AT AR.A6sE ass.Q s a( u,, an,.y,) 14LJL.O . 4 '0E tt j JS unum SPARE. ~~ 5' 4 '*4 Imam sru e. k 37,M

• a MU TwJ M 42.

42 42 h b S-*h 3 -o - - o - -o C -C-C c M ], W5 36 s eA e c 1, NTI) A r '~ '~ + dgIh6%- 3.j.c.e g Cl 'V ik e 4 IN,ST 62 rR2 ggggg; i3 $20-

8. ALL N*inaeff LDCATED AT RESPECTn'E C

,, g*,,c 42C-NGC E*LEFT A*, NOTED b 42 __.-_,,,4 SJ Lan is: LaLS.

2. APftsc AaLE L(niC C LttL-26-S 1 C6 m m s mecc'lataRAAA pe p6 /&

a m peo. o C 4 Tostour, stoiTcetES TGS-O ( Tas-c STop (MF E M Dtt DQ ' ucTCsp. sp HECH AN<Al. 04RLDAD Occtc9t5 i,' " 8 sitea escunf1ED DusMNf3 OPE *enNCi OR CLD",iA8Cb CYCLA. L50s ATnON VALVE 2 Lienst 'asi*c 4

5. fD8t LDCAT CN LEFEstENCE SEE 4.-2%

E7c@C'IA42 TIC 8tl 08.51S A M(.W O6bA(AO) @ VALW322 S(vd 6224b2006.4ds-cos-322 E L 6 REFER TD il2 44 2cos.408 -oca - 620 , 54f'ETY Ib43CCTe0N ACCOAq Ot7rLET uch geoeA ci(r NCLSISAs4 HCCat-(Ob ,d. heoosuon m ano i - -..n-~~ n.~.< g4ae 1, (Tsus our.) F coo:ht/utvg c,wrm ee. et r, c. er a4.. sax = ev strm. "'s erv

• 11 2] h.,Caasa #Emmessiv8.

cwe "l*/.*' ,,,g ro r rutLY k n o.a. n, # --n-- .p m 42 <m => g% ,,, rs....., g:[p.ng,--T.c sioN mA= s m I m "*'*1 se4ee a.,,T, 3) f'"[f$0 4 de i o$4 E,*' e, NUCLEAR SAFETY RELATED Nr*6$ M - h " ""' 44 A r(ut .s-'*m f agpu2 22T eg" -T .. sis.a ' n if o.M ms PEEUMINARY 15 v., u. , h-W2 dp t" y i G '*:7 v.T T.. iA. sisA>4 o w A EM cim.mw wnaui.4eov uca . stSAH *W p , gg L py g g._g y g my 44,suacm,, ao ~~ u m.. 18 F ? gr-P BEM9. anr1 lowr.R.T.Imc 3 c$.g ,4o nu w vctos /, i f tv.SL:Ne uom ccmyn n-mi.- ,, g g) %g* lL%; @ ]J" Uf li . * '" C slaws cc.4ucion vv 2s15ewwe65A(40) at *aDANT $81 I ** / 4 3 a e (gT. no.ir3AY HCCE2 E05 entnt ATiNCa LEim ctT N0 5LI A E d 7 8 12241-E-6HN g g e 8

.r. - _ i ATTACHMENT 3 Respone to FSAR Questions 430.97 and 430.100

Question 430.97 (Section 9.5.6)

Provide a discussion - of the measures that have been taken in the der ~~ of the standby diesel generator. air starting system to preclude the.toulingl of the air start valve or filter with moisture and contan-inants such as oil carryover and rust (SRP 9.5.6, Part III).

Response

Refer to the response to Question 430.100. Question 430.100 (Section 9.5.6) A study by the University of Dayton has shown that accumulation of water in the start ing air - system has been one of the mos t frequent . causes of diesel engine failure to s tart on demand. Condensation of entrained moisture in compres sed air lines leading to control and starting air valves, air start mot' ors, and condensation of moisture on the ; working surf aces of these components has caused rust, sc ale, and water itself to build up and. score and jam the internal working part s of. these vital compo nent s thereby preventing starting of the diesel generators. In the event 'o f loss of offsite powe r, the diesel gene rato rs must function since they are vital to ' the safe shutdown of the reactor (s). Failure of the diesel engines to start from the ef fects of moisture condens ation in air s tart ing systems and from other cau ses have lowered _ their operational reliability to substantially less than the desired reliability of 0.99 as specified in Branch Technical Position ICSB (PSB) 2, " Diesel Generator Reliability Testing," and Regulatory Guide 1.108, " Periodic; Testing of Diesel' Generator Units used as Onsite Electrical Power Systems at Nuclear Power Plants." In an af fort toward. improving diesel. engine starting reliability, we require that. compressed air ' starting system designs include air dryers - for the removal of entrained moisture. The two air dryers most com-monly used are the des sicant. and ' refrigerant types. Of these two type s, the refrigerant type is the one most suited for this applica-tion and, therefore, is preferred. - Starting air should be dried to a ~ dew point 'of - not more - than 50*F shen ins talled in a nonnally con- _ trolled 70*F environment, otherwise, the starting air dew point should be - controlled to at le as t 10*F less than the lowest expected ambient - t empe rature. Revise ; your design of the diesel engine air s tart ing sys ten ac cord-ingly', describe this feature of your design. Also expand your FSAR to ' discuss the procedures that will be followed to ensure the dryers are working properly 'and the freque ncy of checking /tes ting (S RP 9.5.6, Parts II and III).

Raponw The University of Dayton study which concluded that water accumulation was a leading cause of diesel starting failures was based upon analy-sis of Licensee Event Report s (LER) issued from 19 69 through 1977. Improvements in industry practices since that time, and the success of such pract ices must al so be cons ide red. During the firs t ye ar of operation (1976-1977) at BVPS-1, two failures of the diesel generators to start becau se of moisture were reported via LER. To prevent fur-ther occurrences, a system of periodic blowdown as well as blowdowns prior to testing was implemented. No moisture related start failures have occurred which result ed in LE R' s in the seven ye ars since (approximately 1/6 of the expected plant li fe ). In addition, a more recent ' study by the Institute for Nuclear Power Operations used 450 diesel related LE R's issued since January 1980 to analyze failure

data, this study shows that only 5% of the more recent failures were due to moisture. Improved operation and maintenance pract ices applied by the industry when the moisture problem first became apparent on an indus try-wide basis in the late 1970's is undoubtedly part of the reason for this improvement.

Such practices at BVPS-1 have proven to be highly successful. As described in FSAR.9.5.6, the method of starting the diesels used at -BVPS-2 involves admission of air directly into the engine's cylinders. This eliminates the need for air start stors used on some other designs which are, as s tated in NUREG/CR 0660, more suscept ib le to moisture related damage. Another impo rt ant de s ign fe ature diich allows these diesels to be more tolerant of any starting systen fail-ure is the redundancy of the start sys tems. Each diesel is provided with redundant start ing system component s to enh ance its start ing reliability as stated in FSAR 9.5.6. In accordance - with GDC 17, suf ficient independence and redundancy is provided to as sure performance of satety funct ions assuming a single failure. Loss of a single diesel generato r is cons ide red in the safety analyses. As demons trated at BVPS-1, loss of a diesel due to moisture is a rarity when appropriate operating and maintenance prac-t ices are ap plied. Coincident loss of both diesels due to misture would be even less likely and is prevented in part by the independence of the air starting systems. The use of operating pr ocedures similar to those in use at BVPS-1 snould provide diesel reliability similar to that of BVPS-1. Addi-l t ionally, periodic tes ting will confirm the availability of the di es els. Thus, it cannot be concluded that installation of the air drye rs, as requested by the NRC, would provide substantial additional protection which is required for the public health and safety. There-fore, thie proposed backfit is not justified under 10CFR50.109.

r ATTACHMENT 4 Response to FSAR Question 480.26 Question 480.26 (Section 6.2.2) Provide a table listing quantities and locations of the various types of insulation employed inside the cont airunent. Specify this for each cont airunent subcompartment, for both components and piping systems 8-iaches in diameter and larger. -Response: The information requested requires significant extra ef fort to produce and would not be justified because the guidance of Reg. Guide 1.82 was used as the basis of sump design. The Reg. Guide indicates that an assumed 50% blockage is cons ervat ive and, as with other Reg. Guides, it indicates that it provides methods acceptable to the staf f. DIE is aware of unissued draft ve rs ions of the SRP 6.2.2 as well as NUREG 0897 and Reg. Guide 1.82 which appear to be the source of the' ques-t ion. If so, it appe ars that this issue could be int erpreted as an unauthorized backfit since these new criteria which are still unissued came about well after is suance of the cons truct ion' pe rmit and even af ter docketing of the operating license ap plication. In addi t ion, assuming that the staf f is justified in using new unissued material to base the question on, the draf t NUREG 0897 and several other NUREG/ C R' s conclude that the only plants for which the 50% blockage is not . conservative are those having large quantities of fiberous insulation. As point ed out in previous res ponses. to ques t ions 480.2 and 730.1, BVPS-2 has a minimal amount of such insulation. .}}