L-84-246, Forwards Clarifying Info Re Postulated Reactor Coolant Pump Seized Rotor & Loss of All Nonemergency Ac Events,Per 840213 Request

Forwards Clarifying Info Re Postulated Reactor Coolant Pump Seized Rotor & Loss of All Nonemergency Ac Events,Per 840213 Request

ML17215A557
Person / Time
Site:	Saint Lucie
Issue date:	09/11/1984
From:	Williams J FLORIDA POWER & LIGHT CO.
To:	John Miller Office of Nuclear Reactor Regulation
References
L-84-246, NUDOCS 8409140177
Download: ML17215A557 (12)

Text

l j ~

REGULATORY i FORMATION DISTRIBUTION SYS(EM -(RIDS)

'CCESSION NBR;840914P177 DOC ~ DATE! 84/09/11 NOTARIZED:

NO FACIL'.,50 335 St, Lucie Plantr Unit ii Flor ida Power L Light Co.

AUTH,NAME AUTHOR AFFILIATION

'WILLIAMS< J.N, F 1 or i da-Power 8 Light Co, RECIP ~ NAME>

RECIPIENT AFFILIATION MII LER'iJ ~ R,

. Operating Reactors Branch 3

DOCKET 05000335 05000335

SUBJECT:

For wards, cl ari fy'ing -info re postulated, r eactor coolant pump seized rotor

.L,, loss of all non emergency AC eventspper 840213 request' DISTRIBUTION -CODE; A001D COPIES RECEIyED~LTR: ENCL +

SIZE:

',,4j TITLE:

OR Submittal'.

General Distribution

/

NOTES:

OL i 02'/01/76 RECIPIENT'D CODE/NAME NRR'- ORB3 BC- '1.

INTERNAL: ADM/LFMB NRR/DE/MTEB NRR/DL/ORAB NRR/DS I/RAB RGN2

'COPIES LTTR ENCL 7

7 1.

0 1=

1-1 0

ii 1

1 RECIPIENT ID 'CODE/NAME ELD/HDS2'RR/DL DIR N

METB G FI 04 COPIES LTTR ENCL" 1

0.

1 1

1 1

1 1

EXTERNALj ACRS NRC PDR NT?S 09 02' 6

1 1

1 LPDR NSIC 03 05 TOTAL NUMBER OF 'COPIES'EQUIRED:

L'TTR 26 ENCL 23

f

'I

h g<<<<} <<g

~

1 II 1

<<'h}'t ~ I4 11~

ti i<<h} ii;i<<g h

y,)

1 q <<}r 0i,~,h I }II I I ~

qr<<

> > ~1 r 1

11 r q 2 r-i}'1'<<S)

~ N(<<<<i <<<<I'I

~

II 1

F I

FLORIDAPOWER & LIGHTCOMPANY September 11, 1984 L-84-246 Office of Nuclear Reactor Regulation Attention:

Mr. James R. Miller, Chief Oper ating Reactors Branch g3 Division of Licensing U. S. Nuclear Regulatory Commission Washington, D.C.

20555

Dear Mr. Miller:

Re:

St. Lucie Unit 1 Docket No. 50-335 Reactor Coolant Pump Seized Rotor and Loss of all Non-Emer enc AC Power In response to your 'letter of February 13, 1984, concerning the postulated reactor coolant pump seized rotor and loss of all non-emergency AC events, attached please find the cl ari fying information requested.

Very truly yours, J.

W. Willians, Jr.

Group Vice President Nuclear Energy JWW/PLP/js Attachment cc:

J.

P. O'ei lly, Regi on II Harold F. Reis, Esquire P NS-LI-84-215 pgp ggaCK 05000335 8409i40i77 840911 '

':.;;,; 'PDR ~

j PEOPLE... SERVING PEOPLE h

Page l of 5 ATTACKKNT uestion 1:

Reference 1, guestion 9 (two pages before graph):

Claims 3-second time delay between turbine trip and LOOP is )ustified by C-E analysis.

But this is justified only if St, Lucie output is less than 5% of subregional "grid power per your Ref.

14" p. 13, last two lines.

Is this the case7 Answer:

The above question indicates a misunderstanding that the power "output" pro-vided by St.

Lucie 1 comprises less than 5% of the "grid power."

The document referred to as Reference 14 in the question specifies that the plant power gen-eration is about 8% of subregional grid generation durin minimum load condi-tions.

The 8% generation loss to the grid upon turbine trip was calculated by assuming that St. Lucie 1 was.operating at 100% capacity at the time minimum load con-ditions exist on the grid.

Therefore, the percent generation loss to the grid can be defined as the plant generating capacity (802 l%e) divided by the mini-mum yearly load.

The minimum yearly load was estimated using typical daily, weekly and annual load patterns presented in Reference g1-1.

The minimum yearly load was determined to be approximately 50% of the sumer peak load.

The suer peak load for the Rorida subregion in 1983 was estimated to be 20,046 Ne based on data presented in.Reference gl-2.

These values lead to the estimated 8% subregional grid generation loss.

i.e.

802 hke (Unit Capacity) 50% of 20.046 Ne (estimated 1

ly 1 d

~dd tt If the generation loss to the grid were greater than 8%. 'the time to LOOP would decrease as shown in the attached figure.

However, the loss of St.

Lucie 1 is believed to be conservatively bounded by an 8% generation loss to the grid upon turbine trip.

Even if the loss of St. Lucie 1 constituted a

9% generation loss to the grid, the time del~ to LOOP would still be greater than three seconds.

It should be noted that the aathodology used to generate the attached figure incorporated several other conservatisms:

~

No credit was taken for load shedding programs designed to arrest frequency decline.

(i.e., initial rate of frequency decline is assumed.)

~

No credit was taken for spinning reserve.

~

The Florida model assigned "island" conditions (no support from neighboring grid systems, i.e., lower rotating inertia).

~

Loss of offsite power occurs instantly at a grid frequency of 58.0 Hertz for SL1.

~

All grid systems are standardly operated and specifically designed against unstable frequency oscillations resulting from loss of the largest generating unit on that grid.

Reference gl-1:

C.

R. Nusick, R.

M. Knapp.

"C-E Meets Utility Load Change Capability Needs with System 80 NSSS." TIS-6215, April 1979.

Reference g1-2:

Electric Power Su 1

and Demand 1981-1990, National Electric

+

~ rgi'....)u r 198.

V

gyp gX lO TO~INCH 7 X lO WCHKS 46 l320

~

~I

~ ~

page 3 of 5

Question 2

Ref.

2, Q84 says AFM has

3. minute delay ior ZNTERIN system.

Ref. 3,,Qf5 says AFW has 2 minute delay for SHART system.

Are these the numbers that were used in analyzing system performance?

Answer The system performance following the LOOP and Seized Rotor Events analyzed in Reference 1 assumed a

2 minute delay time for the AFW SMART system.

Question 3

A.

Ref.

3, Q05 says delay time is inconsequential to the results because AFW will be shut off to SG side having a faulted ADV.

B.

Ref.

2, QP4 says the SMART AFM system is assumed (par.

1), and it also says that (par. 3), for both LOOP and. SR, the AFM is shut off to the SG having a

faulted ADV.

C.

Statements AEB appear to conflict with statements in ref.

1 which say that AFM enters the affected SG.

Specifically:

Ref.

1, Att. 1, Table 4, 1056e7 sec for Seized Rotor for AFW entry.

Ref.

1, Att. 2, Table 3, 1186.1 sec for LOOP for AFM entry.

Note:

Both of the cited Tables show the operator closing the affected ADV's block valve at 1800 seconds.

D.

Question:

Was "interim" AFW system used to calculate system performance and SMART AFM system used to calculate dose, or what?

For SR?

For LOOP?

Answer to 3C The reanalysis for the LOOP and Seized Rotor Events presented in Reference 1

assumed, as a single failure, a stuck opened ADV in the, affected steam generator.

The ADV is located on the main steam line, upstream 'of the main steam line isolation valve (MSIV).

Results of the calculatioris presented in Reference 1 indicate that the affected steam generator's pressure never decreases below the MSIS setpoint pressure during the first 30 minutes of the transient.

As

such,

.the affected MSIV remains open 'uring that period.

Since pressure differentials in the steam lines are low following turbine trip, the check valve located by the affected MSIV remains

open, allowing steam leakage from the unaffected steam generator.

Therefore, both steam generators release stem through the stuck opened ADV.

As a

result, both steam generators depressurize virtually at the same

rate, showing no apparent asymmetry.

However, since the affected steam generator discharges more steam through the stuck opened ADV than the unaffected steam generator due to its proximity to that valve, its level reaches the AFW actuation setpoint, initiating the delivery of the AFM following a 2

minute delay time.

h

~.

d

5y 4

Page 4 of 5 Answer to 3)

The analysis performed in Reference 1 used the SMART system to calculate system response and off-site doses for both the Seized Rotor and LOOP events.

The interim system was never used for any part of the analysis.

For the purposes of conservatively bounding the off-site dose calculating a partition factor of 1.0 was assumed.

This factor maximizes off-si-te dose magnitude.

Such a partition factor could only arise if a large differential pressure developed between the two stean generators reaching the AFAS cutoff setpoint and resulting i n subsequent dry-out of the affected side stean generator.

This differential was not predicted to develop during these analyses and hence the sequence of events table shows AFW reaching both sides.

However, by assuming a 1.0 partition factor, we have conservatively bounded the dose calcul ation.

As pointed out in prior correspondence, the upgraded or "STUART" AFW system was anticipated to be installed within less than a year of the Reference 1 submittal.

It was felt to be prudent to'include its effects in the analyses.

Subsequent correspondence has remained consistent with the initial submittal.

It should be realized, however, that final implementation of the system has been delayed.

The currently functional "interim" system, which has received NRC review is also conservatively bounded by this analysis.

Question 4

I Ref.

3, last

pages, repeats Reference list from Ref. 2 and says it is the same as the one missing from Reference 1.

I question, this specifically, referring to Reference 1,

Attachment 1: that document has two citations to a "reference 1" which don't appear to match the "reference 1" included in Ref.

2 and Ref.

3.

The two Citations in Reference 1 ~

attachment, 1

are (1) first page, Discussion, lines 1

and 2,

and (2) second

page, second paragraph, last sentence.

Do these both refer to the same "reference 1" as in Ref.

3V If so, can you provide it?

answer The two citations in Reference 1 refer to the same "Reference l" as in Reference 3.

However, in that reference, the Seized Rotor Event was re-evaluated and the discussion was based on a previous analysis, which is presented as Reference 5 herein.

0

~M I

Page 5 of 5 Question 5

Rtf.

1, Att. 2 Table 3 says AFW begins to enter the affected SG at 1186.1 sec.;

Ref.

2, Qf4 indicates that such AFM flow could occur with the existing "interim" system but not with "SMART" system that ms claimed to be used in the analysis.

Please describe shat is being asseaed in the analysis, and

<that is not being

'assumed about kFW system performance.

Please summarize

results, including corrected time data per Rtf. 3,

QtB, bottom.

And. ii available, provide graphs covering 0 to 20 seconds.

Snsuer As discussed in the answer to question 3C, even when assuming the SMART AFW

system, the AFV begins to enter the affected steam generator at 1186.1 aeconds since the LOOP reanalysis in Reference 1 does not show a calculated differential in the two steam header pressures great enough to reach the AFW shutoff aetpoint.

References:

1-Reanalysis of 8-31-82 for LOOP and/or Seized Rotor: Letter, R.

E.

Uhrig (FFC,L) to D.

E Et.senhut (RRC), "St. Lucte Dntt

)

Docket 50-335 Stretch Power - Additional Anal sis",

L-82-381, August 31, 1982.

2.

LOOP Clarification of 11-21-83:

Letter, J.

W. Williams (FP&L) to J.

R.

Miller (NRC), "St. Lucie Unit 1 Docket No.

50-335 Loss of Offsite Power-Additional Information Re uest,"

L-83-567, November 21, 1983.

3.

Seized Rotor Clarification of 12-22-83:

Letter, J.

W. Williams (FP&L) to D.

G. Eisenhut (NRC), "St. Lucie Unit No.

1 Docket No. 50-335, Seized 'Rotor Event Additional Information Re uest,"

L-83-60, December 22, 1983.

4.

Letter, J.

R. Miller (NRC) to J.

W. Williams (FP&L), "Reanalysis of Reactor Coolant Pump Seized Rotor and Loss of All Non-Emergency AC Power,"

Docket No. 50-335, February 13, 1984.

5.

Letter, A.

S.

Jameson to R.

W.

Winner, "Cycle 4 Design and Safety Report for Stretch Power," F-CE-7126, March 19, 1980.

If t

J