ML20112C921
| ML20112C921 | |
| Person / Time | |
|---|---|
| Site: | Harris  |
| Issue date: | 01/08/1985 |
| From: | Eddleman, Eddleman W EDDLEMAN, W. |
| To: | |
| References | |
| CON-#185-025, CON-#185-25 82-468-01-OL, 82-468-1-OL, OL, NUDOCS 8501110480 | |
| Download: ML20112C921 (4) | |
Text
.. O N _'_ _ _. _ _.
.w.
R ne UNITED STATES OF AMERICA
- C January 8, 1985 (Extension of time NUCLEAR BEGULATOBY COMMISSIOy egr j1 NO:3ppproved Jan 2 and v
Jan 3,19?5 by the Board, Staff and J ME aua.c.:.Annlicants)
BEFORE THE ATOMIC SAFETY AND LICENSWG BOARD'm h
Glenn O. Bright
~'"
Dr. Jares H. Carpenter James L. Kelley, Chairman In the Matter of CAB 0 LINA POWER AND LIGHT CO. et al.
)
(Shearon Harris Nuclear Power Plant, Unit 1)
ASLBP No. 82-1468-01
)
OL Joint Intervenord Pronosed Findings re Joint Contention 7 (Stean Generators Multiple Tube Ruptures)
Wells Eddlenan hereby files on behalf of hinself and the Conservation Council of North Carolina, Kudzu Alliance, and chanel Hill Anti-Nuclear Group Effort (CH_ANGE), the following pronosed findings of fact for the Joint Intervenors on their Contention VII:
1.
Anplicants' clains of innrovenent in tube failure rates are not reliable because the rate of stean generator tube failures appetrs to be stable or increasing.
To illustrate this, consider the US Westinghouse plants (for which it is OK to construct a stat!stic' by the sane nethod Anplicants ' witness Hitchler used for all 'festinghouse nuclear plants in the world, see Tr. h078, Hitchler on cross ).
- First, their failure rate is higher ( 1; ruptures in P.1456 nillion tube years)
O
@g (ruptures, see Hitchler Table 7; U.S. 51 plant tube years, Hitchler wu h8 table 1, attached to his nrefiled testimony folowinD Tr.14012).
~<
J h
Applying Hitchler's 10% conservative discountirg method to the con.o (t %v U.S. W tube years, gives It fa lures in about 2.2 nillion tube years, a failure rate of 1.8 x 10-6 runtures ner tube year.
Pgirability assessment involves judgment (Tr.1 0911-P5 ) and judg-ment is used in these findinEs following:
gT3
3 s.
'm Since the Harris plant is a U.S. Westir.ghouse plant with nearly 1h000 steam tubes (Hitchler orefiled, calculation of failure probabilities at p.10 following Tr. 401P)(h578 tubes, 3 stean generators at Harris),
it will incur about 0.5 to 0.6 million tube years in a 40 year overcting life.
This means an expectation of one tube rupture, over this life.
The 1.8 x 10 tube rupture exnerience in US Westinghouse nlants is higher than Hitchler's overall historical figure of 1.6 x 10-6 (H prefiled at A.10, p.6).
While this difference may not be very statistically significant, it is not reassuring. It is three times his o.6 x 10-6 estinate for the Harris plant. That estimate is nnt credMe.
2.
The U.S. experienced tube rupture failure rate for Westinghouse design nuclear plants was not decreasing over the period covered in witness Hitchler's tables.
Hitchler testified (Tr. h137) that it was OK to comnare pre-October 1979 operating experience (on the sane basis )
with ext >erience after that date.
For exannle, no loose-narts ruptures had occurred by October 1,1979, but two occurred in the next three years (seeTr.h1h37. (l/I3hM)
When Hitchler's Table 1 data are used to get the tube ynans as of 9-30-79 (it ends on 7-31-83), the nrocedure is to remove all plants with connercial operation dates af ter 7-31-83, and for the remaining plants, to subtract 3.833 years (3 years and ten months, the time fron 9-30-79 to 7-31-83) from the years of operation for all other plants.
That gives 9-30-79 operating years for all plants in commercial operation by that date.($1ese data are unadjusted for canacity factor, Tr. h052-53)
The result of this calculat* on
- s about 1 36 million tube years -
for US Westinghouse plants as of 9-30-79.
There had been two ruptures in these n1 ants by that date (steam tube ruptures), a failure rate of about 1.7 x 10-6 (discounting 136 million tube years 105 for conserva-tism consistent with Hitchler's nethodology: 2/1.2 n!111on 7.7 x 10" ).
1 w 1.\\b But in the next 1.09 million tube ywars (P.h56 nillion ninus 1 36 nillion) 3 fron October 1,1979 thru July 31, 1983, there were P failures also, a rate (same methodology) of P.0 ner nill!on tube years.
7-w.
a 3.
There are not many onerating years' experience with the "imorovenents" Hitchler clains for the Harris plant. Tr. h133-36.
This together with the above facts leads to the adoption of a failure rate based on recent failure rates in U.S. W nlants (1.8 to P.0 per million tube years).
Using the Chi-scuared methodology Hitchler recommended (when he finally got it straight -- Tr. 4144, see also 4146, 4127) of 10 degrees of freedon (2 x 4 ruptures in US W ninnts, plus 2, see Tr. k1kk), the 95% confidence linit based on that is 18.3 (from Chi-squared tables)/2 x (2 to 2.2 nillion tube years ) or from h.6 to h.2 failures by rupture per nillion tube years.
These correspond to 2 or more rup'tures for the Harris niant over its onerating life as projected (ho years).
Thus a multiple rupture at Harris is not inpossible; indeed, excluding connon-node failures, it has a probability of 1 in a nillion or nore; and enmmor-node failures cannot be excluded.
4 There is no rigorous analysis of nultiple tube runtures in the Harris FSAR (Tr. h016, see also h022-23 and h15P-53).
Hitchler has never done one and doesn't know the snecifics (Tr. h017, see h016-17).
The Staff did no statistical analysis for Harris, Tr. hP3h-35
- 5. Witness Hitchler has not reviewed Harris OA nrocedures (Tr. h136).
This further undernines the credibility of his "inprovement" estinates since QA failures could lead to tube ruptures, even connon-mode ones.
- 6. Of the four US West!nghouse steam generator tube ruptures, none leaked beyond specification limits before rupturing (Tr. hP17) and only one apparently leaked at all before rupturing and it was "below the (technical snecification) limit un to the time when it increased way over the (same) limit" (Staff witness Marsh, Tr. hP17)'iradeouate proceduvesg(
Thus rupture events strike without warning.
Worse accidents can show 7.
Shearon Harris has a relatively high tech spec leak linit of ga g
500 gallons per stcan generator per day (1500 gallons / day for all 3)
Tr. 4210-11.
This makes leak-before-rupture less likely for Harris.
8.
There can be steam generator tube ruuture events with con-sequences beyond the design basis (Tr. 4218-19,see also h227) These include neltdowns (Hitchler, Tr. h012 ff (prefiled) Ans 21, p.13).
althouc Hitchl meltdo9n risk (er believes this is only a 3% addition)to so runture to. include multiple ruptures -- ibid.
- 9. Most stean generator tubes are plugged due to corrosion (Tr. h0h7-8 and it is nossible for any tube to conrode (Tr. h036). There are other forms of corrosion besides the nain ones (Tr. h037-38).
Nevertheless, actually corroded tubes have not bean bu"9t tested (Ptare, Tr. h183-85).
- hus the burst-test derived probabilities of multi low., The Harris tube plunging criterion remains h0%nle ruuture are too thickness, Tr. h179-60 (SEaff).
degradation of wall 10.
Theactual frecuency of tube ruutubs for U.S. Westinghouse nuclear olants is 1 in h5 years or rather higher (Hitchler nrefiled A.7 n.5; see Findings 1 through 3 above).
If, as concluded above, the frecuency is nore than one event in h0 years, classification of tube runture as a "Condit!on IV" event (Hitchler orefiled, A.6, p.4, ffg. Tr. h012) as is now done, is inproper.
Therefore steam generator tube rupture should be classified as an event exnected more of ten than once in plant life.
Multiple tube ruptures should be rigorously analyzed for the Har"!s plant because the Atomic Energy Act recuires the NEO to out safety first.
It is so ordered.
TP M D Ad L
w UNITED STATES OF AMERICA NUCLIAR REGULATORY C010iISSION In the matter of CAROL!hA POWER k LIGkiT CO. Et al. )
Docket 50-h00 Shearon Harris Nuclear Power Plant, Unit 1*
)
0.L.
CERTIFICATEOF SERV!CE I hereby certify that copies of Wells Fddlenan's Proposed Findi.ngs on C_ontentions hl,116 and 9; and of J.I. Pron. Findings on Joi.nt. 7 EAYE been served this h. day of January 1981,bydepositin the US W il, first-class postaEe prepaid, upon all parties whose nanes are listed below, except those whose nanes are arked with an asterisk, for when service was acconplished by
- Extension of time fron Jan. 4 apnroved by Annlicants, Staff and Board Jan.2 Judges Ja tes Kelley, Glenn Bright and Jar.as Carpenter (1 copy each)
Atonic Safety and Licensing Board US Nuclear 9egulatcry Connission Washin6 ton DC 20555 George F. Trowbrid e (atterney for Applicants)
E Shaw, Pittman, Potts & Trowbridge lluthanne G. Miller 1600 M S t. h*4 ASLB Panel WashinEton, DC 20036 USNRC Washington DC 2C55 5 Office of the Executive Legal Director
[
Spence W. Perry Attn Docke ts 50-400/401 0.L.
pW [ cst.SW das ngton DC 20555 CNM washingt n DC 20740 Dan Read Docketing and Service Section (3x)
CEA!L'E/FLP Attn Docke ts 50-h00/h01 0.L.
Waleigh,ft07 Waveross NC 2760~6 Office of the Secretary USNRC Dr. Linda W. Little Washington DC 20552 Governor's Waste Mgt. Bd.
(E plan only) 513 Albenarle B3dg John Munkle Steve Rochlais 325 N. Salisbury St.
CCNC FEMA-Suite 700 M
137.1 Peachtree St.NE 'aleigh, NC 27611 307 Granville Rd Chapel Hill Nc 2751h Atlanta GA 30309 Bradley W. Jones Robert Gruber USNRC Region 11
'Travi.J Payne Exec. Director 101 Marietta St.
Edelstein & Payne Public Staff Atlanta GA 30303 Blex 12601 Box 991 Raleigh NC 27605 Raleigh NC 27602 Richard Wilson, M.D.
Certified by h
729 Hunter St.
Apex NC 27502