ML20031H536
| ML20031H536 | |
| Person / Time | |
|---|---|
| Site: | North Anna  |
| Issue date: | 10/21/1981 |
| From: | Christman J HUNTON & WILLIAMS, VIRGINIA POWER (VIRGINIA ELECTRIC & POWER CO.) |
| To: | |
| References | |
| ISSUANCES-OL, NUDOCS 8110280083 | |
| Download: ML20031H536 (31) | |
Text
f-t I ot c) b s+-
6 I
7 <
(.
i
,9 October 21, lgWLMETED
.9l QG
- qq u:>NRC
[Q o.s.qKgG M og\\r -
- ' h, N
11 OCT 23 P4:29
(
0;y
$//
NUCLEAR REGULATORb{ Q5ilI'CA UNITED STATES 70 0FFICE OF SECRETARY 00CKETg ERVICE OMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING APPEAL BOARD
^
In the Matter of
)
)
VIRGINIA ELECTRIC AND POWER
)
Docket Nos. 50-338 OL
- COMPANY,
)
and 50-339 OL
)
North Anna Nuclear Power Station,
)
Units 1 and 2)
)
MEMORANDUM OF VEPCO'S COUNSEL ON NORTH ANNA 1 AND 2 TURBINE MISSILE ANALYSIS Purpose of this Memorandum The purpose of this memorandum is to summarize and draw together into a conceptual framework the various documentsM that make up Vepco's case on the turbine missile risk at North Anna 1 and 2 -- to provide a sort of " road map" of the evidence.
This memorandum is not, of caerse, intended itself as evidence, though it may be taken to represent Vepco's position on matters of law.
In effect, it will serve as before-the-fact proposed findings of' fact and conclusions c:f law, or as a trial brief.
3 g50 S
1/
Most of the documents on which Vepco relies contain information proprietary to the Westinghouse Electric 1
Corporation.
811028 COB 3 38 POR ADOCK 0 00 pg O
f s
4 Summary of Vepco's Case Vepco submits that, with respect to the risk of turbine missiles at North Anna 1 and 2, the NRC regulations for the issuance of an operating license are met by Vepco's commitment to ultrasonically inspect the North Anna turbine discs-periodically, on a schedule determined by Westinghouse's
" Criteria for Low Pressure Nuclear Turbine Disc Inspection,"
MSTG-1-P (Applicant's Exhibit V-1).
s Using the methods described in Exhibit V-1, Westinghouse calculates the recommended amount of operating time between inspections (the inspection interval) for each separate terbine disc.
Westinghouse also has calculated and reported, in Applicant's Exhibits V-12 and V-13, the expected energies of turbine missiles produced by the rupture of each separate disc.
For some discs the energy is found to be insufficient to escape the turbine casing; such discs are said to be " contained" in the event they rupture.
Vepco submits that it is appropriate, strictly from a health and safety standpoint, to base the inspection interval on only the discs that are not contained, because only missiles from those discs can damage the plant's safety systems and i
l present a risk to the public.
For the present North Anna 1 and l
2 discs the inspection intervals selected on this basis are 43.3 operating months for North Anna 1 (baced on disc 2) and i
39.9 operating months for North Anna 2 (again based on disc 2).
P
._.~,
_. ~
T t
Because of the need to protect the turbine and to schedule turbine inspections during refueling outages so as to minimize plant downtime, however, Vepco may follow an inspection schedule shorter than stated above.
Vepco likely will inspect every 36 calendar months (about 33 operating months) at alternate refueling outages.
This length of time happens to fall within the inspection interval based on the most critical disc regardless of whether it is contained or not (34.9 operating months for North Anna 1 and 33.2 operating months for North Anna 2).
Although Vepco basea its case on the analytical methods in Exhibit V-1, which are deterministic rather than e
probabilistic, it can be shown that an inspection interval of 33 to 36 operating months is e; proximately in line with the NRC
~N Staffts probability guideline of 10 per year.
Dependtug on the assumptions used, particularly the assumptions about
" scabbing," the calculated probability P of unacceptable 4
turbine raissile damage over three operating years (36 operating
-4
-6 months) can range from 1.53 x 10 to 2.11 x 10 with the l
l correct number likely to be closer to the latter than the former.
(Neither number takes credit for a number of j
I conservative assumptions.)
By comparison, a plant that meets
-6 the 10 per year NRC Staff guideline has a P over three years 4
-6
-of about 3 x 10 l
r V
r e
y
-g
'awi-e 3w e
y y
wr y
9
e t
1 Vepco proposes to rely on the Westinghouse deterministic inspection criteria to deal with stress corrosion cracking (Applicant's Exhibit V-1) rather than the updated probabilistic analysis P (Applicant's Exhibit V-4) for the 4
following reasons:
1.
The deterministic method has been reviewed and approved by the NRC Staff; the probabilistic approach is still under review.
2.
The probabilistic approach has not been used heretofore to set inspection intervals.
It is, in Vepco's view, only a rough screening device to determine whether further analysis is required.
3.
To use the probabilistic approach to set inspection intervals requires that there be a probabilistic safety goal.
But no such goal has been established.
4.
The probabilistic approach relies on many assumptions, some of which are extremely difficult to quantify.
Accordingly, it can be used to calculate more or less extreme upper bounds to the true probability, but calculating a "best estimate" of the true probability is i
difficolt and subject to great l
uncertainty.
i
~
Background of this Case l
This case begar,, for all practical purposes, in the spring of 1979, when the Appeal Board, as a result of its sua spogie review of the North Anna 1 and 2 operating license decision of the ASLB, decided that further evidence on the risk of turbine missiles was needed.
i 2 The risk of turbine missiles at North Anna is evaluated against General Design Criterion 4 in 10 C.F.R. Part 50, Appendix A (1981), which provides that structures, systems, and components important to safety "shall be appropriately protected against dynamic effects,' including the efffects of missiles.
(NRC Staff Memorandum of Proposed Pindings Regarding Service Water Pumphouse Settlement and Turbine Missile Risk, Aug. 6, 1979, at 43.)
Vepco and the NRC Staff presented evidence on the turbine missile issue at a hearing on June 20, 1979.
Vepco attempted to carry its burden of proof on this issue by a probabilistic analysis ^ suggested by Regulatory Guide 1.115 (Attachment 3 to this memorandum) and the NRC Staff's Standard Review Plan 2.2.3 (Actachment 4).
That is, Vepco calculated the yearly probability (called Pg) that a turbine missile might be generated and do unacceptable damage to plant safety nystems.
The i1RC Staff's guideline was that the risk was
~
acceptable if P was 10 per year or less, if realistically 4
-6 calculated, or 10 if conservatively calculated.
Vepco's analysis, starting with Westinghouse's calculation of the i
probability of missile generation Py, showed that P4 was about
-7 4.22 x 10 per year, conservatively calculated.
Since this
-6 was below the Staff guideline of 10 no additional protective measures were required, in Vepco's view.
1 Vepco's 1979 calculation of P was based on turbine 4
disc rupture at design overspeed (about 120 per cent of rated speed) and destructive overspeed.
But the calculation did not take into account the probability of missile generation at
~
operating speed, which was thought to be negligible.
Since the hearing, however, evidence of cracking in the keyways and bores of specific discs on specific turbines, attributed by Westinghouse to stress corrosion, began to arise.
Because of this development, and at the NRC Staff's urging, a task force consisting of Westinghouse and its customers, called the Disc Integrity Task Force, was formed to study the disc cracking phenomenon.- During the past two years Westinghouse and the NRC Staff have worked to come to grips with the turbine disc cracking phenomenon and to develop means for analyzing the safety implications.
The method developed to deal with the cracking problem, at least until design changes
~
or other measures can be instituted, is to ultrasonically inspect the discs perii lically on a schedule determined by Westinghouse report MSTG-1-P, " Criteria for Low Pressure Noclear Turbine Disc Inspection" (Applicant's Exhibit V-1).
The NRC Staff have approved this approach (see Attachment 2),
and Vepco has committed to use it (Applicant's Exhibit V-5).
It is this method that Vepco asks the Appeal Board to approve as an acceptable way of complying with General Design
i t
1 Criterion 4.
Part I of this memorandum summarizes the method.
Part II presents an update of the probabilistic approach used in 1979 and shows that it is not inconsistent with a decision to rely on the Westinghouse analysis described in Applicant's Exhibit V-1.
Part III addresses, very briefly, the cause of the disc cracking and the issue of nonsafety grade equipment to mitigate transients.
Attached to this memorandum are Attachment l',
a list of Vepco's Attachments and Exhibits (the Exhibits are numbered V-1 through V-22); Attachment 2, the letter from the NRC Staff asking Vepco to commit to the Westinghouse turbine disc inspection criteria; Attachment 3, Regulatory Guide 1.115; and Attachment 4, Standard Review Plan 2.2.3.
I.
THE WESTINGHOUSE INSPECTION CRITERIA Because of the recent cracking experience, Westinghouse has advised Vepco to institute a program of periodic inspection of the turbine discs by ultrasonic means.
'lepco has accepted this recommendation.
A.
Summary of the Criteria The question now is how often the inspections must be made in order to give reasonable assurance that a crack will not grow to critical size between inspections.
The method for calculating these intervals that Westinghouse has developed is
- contained in its Memorandum Final Report entitled " Criteria For Low Pressure Nuclear Turbine Disc Inspection," MSTG-1-P (June 1981) (Applicant's Exhibit V-1).
This method is deterministic (see Applicant's Exhibit V-16 at 5), as contrasted with the probabilistic approach Vepco used in 1979.
The method of calculating the time between inspections is quite simple in theory.
The expected rate of crack growth da/dt for each disc is calculated; the depth of the crack that could cause the disc to rupture (the " critical crack size" acr) is determined; and the latter is divided by the former to give the amount of time required for a crack to_ grow to the point where it could reach critical crack size and rupture.
The time between inspections in months of turbine-operating time (called the inspection interval) is based on the Westinghouse criteria document (Applicant's Exhibit V-1) and the NRC Staff review of it (Attachment 2).
How a and da/dt are determined is explained in ct Applicant's Exhibit V-1.
The crack size a is a function of ct the flaw shape. parameter Q, the fracture toughness K f the IC metal in the disc, and the bore stress sigma (Ezhibit V-1 at 6).
The calculation of the crack growth rate da/dt is discussed in Applicant's Exhibit V-1 at pages 10-19.
The materials properties, stresses, and operating temperatures for the North Anna units that are the input for these calculations L
< are shown on Applicant's Exhibits V-10 and V-ll.
(Exhibits V-10 and V-11 are submitted only to show the basic parameters of materials properties, temperatures, and stresses; the da/dt and a numbers shown on the print-outs are now outdated.)
ct B.
Concervatisms The significant factors that make the Westinghouse inspection criteria conservative are discussed on pages 20-22 of Applicant's Exhibit V-l.
(The NRC Staff prefer Westinghouse's " Alternate Inspection Criterion" on pages 22-23, and that is what has been used to calculate the North Anna inspection intervals.)
For example, Westinghouse recommends an inspection not after the calculations show that a crack would have reached critical size, but at only 1/2 of that time, to provide considerable margin.
The Westinghouse Report notes on page 22 that a safety factor of 4.6 to 6.9 exists for uncracked discs and 9.2 to 13.8 for cracked discs.
Vepco believes that regular inspections based on the Westinghouse analysis will provide reasonable assurance"that turbine missiles will not be generated at all.
Moreover, there is an additional " conservatism," if that is the correct word for it, inherent in the fact that even if a missile is generated, it may not incapacitate any safety systems; it may miss safety-related equipment altogether or be stopped, i
deflected, or slowed by barriers, or it may destroy a
.-r
.. -. +,. - - - -,,,
6 t safety-related system for which a redundant backup system exists.
These sr.fety factors find their expression in the P2 and P f the probabilistic analysis.
3 C.
Containment of the No. 1 Disc Westinghouse also has developed a method for calculating the energy of a turbine disc fragment once released from the spinning turbine.
This method is detailed in the Westinghouse topical report WSTG-2-P, " Missile Energy Analysis Methods for Nuclear Steam Turbines" (May 1981) (Applicant's Exhibit V-?).
Part A of the report describes the calculation of the internal energy of the disc fragment, and Pact B discusses the ability of the turbine casing to absorb the fragment's energy.
The results of applying Applicant's Exhibit V-2 to North Anna 1 and 2 are found in the Westinghouse " Turbine Missile Reports" (Applicant's Exhibits V-12 and V-13).
As j
Table 9 on page 19 of each of these reports shows, fragments of the number 1 disc are generally contained by the turbine casing.
(The only exception to this is the 90% disc burst from destructive overspeed; destructive overspeed is not affected by the stress corrosion cracking with which the inspection criteria are concerned.)
4 m
s r--
g.
7 4
.--g--
y,
a
_11-D.
Results of the North Anna Analysis Using the Westinghouse inspection criteria produces an inspection interval for North 7.nna 1 of 34.9 operating months (using disc no. 1, which is contained by the turbine casing in the event of rupture) and 43.3 operating months (using disc no.
2, which is not contained):
i e
I f
. North Anna 1 LP#1 Time in Operating Disc
- End, Months 1
GVN 35.9 1
GNN 42.9 2
GVN 43.5 2
GNN 43.3 LP#2 Time in Operating Disc End Months 1
GVN 47.1 1
GNN 34.9 2
GVN 53.3 2
GNf1 46.2 The corresponding intervals for North Anna 2 are 33.2 months (using disc no. 1, which is contained) inJ 39.9 months (using disc no.
2, the most critical disc that is not contained) (see
~
Applicant's Exhibit V-6).
Thus, Vepco believes that, when applied to the present discs on the North Anna turbines, the analytical method in Applicant's Exhibit V-1 justifies an inspection interval (that is, an allowable amount of operating time-between inspections) of 43.3 months of operating time for North Anna 1 and 39.9 months for North Anna 2, based on the most critical discs "that are not contained by the turbine casing.
b
. Vepco may want to inspect more often than tiis, of t
i course, because the number 1 disc, if it ruptured, would damage the turbine, and Vepco has an interest in preserving-its turbine as well as in protecting the public.
But t'.t i s i s essentially an economic decision best left to the pirnt operator.
Vepco also may inspect more frequently than 43.3 and 39.9 months because of its refueling schedule.
Because refueling outages are scheduled for every 18 calendar months, and because it is desirable to schedule turbine inspectione during refueling outages so as to minimize plant downtime, inspection intervals of 43.3 and 39.9 operating months would in practice mean that inspections would have to be held during alternate refueling outages, that is,'every 36 calendar months.
Accordingly, Vepco will probably follow the
[
Westinghouse recommendations for the most critical disc, l
regardless of whether the disc would be contained in the event of its rupture.
With the present discs, this amounts to an l
inspe,: tion interval of 34.9 operating months for North Anna 1 and 33.2 operating months for North Anna Unit 2.
Since of the 18 calendar months in a refueling cycle about six weeks (a
~
month and a half) is devoted to the refueling outage, during which the turbine is not running, two refuelina cycles or 36 calendar months amount to about 33 turbine operating months.
l
, Therefore, the Westinghouse criteria for the most critical crack, regardless of whether it is contained or not, would allow Vepco to inspect the turbines at alternating refueling outages.
The NRC Staff's Safety Evaluation Report recognizes that in individual cases th'e inspection intervals may be adjusted slightly to coincide with scheduled refueling outages, if the turbine vendor's recommendations are followed.
In light of the sizable conservatisms in the Westingheuse inspection criteria, Vepco believes that this allowance for flexibility is appropriate.
E.
Ultrasonic Inspectiop Method The ultrasonic inspection technique developed by Westinghouse for turbine disc cracking is described in the three inspection reporte for the North Anna rotors, Applicant's Exhibits V-7, V-8, and V-9.
The evidence of the reliability of this technique is given in Applicant's Exhibit V-16 at page 6.
It has been shown that the actual crack depths can be l
l greater than the ultrasonically predicted depth (Applicatt's Exhibit V-1 at 13).
But Westinghouse's inspection criteria incorporate a factor to account for this uncertainty for discs l
l found to have crack indications (id.).
l l
l l
m.
- The discs on the turbines at both North Anna 1 and North Anna 2 have been inspected once, North Anna 2 before going into service.
The inspection reports are Applicant's Exhibits V-7, V-8, and V-9.
The inspection of North Anna 1 revealed crack indications in the LP2 rotor, and so it was replaced by a rotor from another plant, the discs of which had been inspected and found to have no indications.
E.
No need for a Technical Specification Requirement Vepco's position is that its commitment to inspect the-turbine at the intervals stated above (43.3 and 39.9 operating months) provides reasonable assurance that there will be no turbine missiles generated at North Anna during its lifetime, and that no technical specification requirement is necessary, because technical specifications "are to be rescrved for those matters as to which the imposition of rigid conditions or limitations upon eactor operation is deemed necessary to obviate the possibility of an abnormal situation or event giving rise to an immediate threat to the public health and l
safety."
Viroinia Electric and Power Co. (North Anna, Nuclear l
l Power Station, Units 1 and 2), ALAB-578, 11 NRC 189, 217 (Feb.
j 11, 1980), quoting Portland General Cloctric Co. (Trojan l
l Nuclear Plant), ALAB-531, 9 NRC 263, 273 (Mar. 21, 1979).
No l
such situation occurs here.
I l
l l
- F.
Possible Design Changes As Dr. Seth's paper (Applicant's Exhibit V-16) points out, Westinghouse is considering design improvements to provide greater: margin for: operation in~ a corrosive environment.
-(The.
Board will recall that potential design improvements were discussed at the North-Anna site visit on 'ebruary 19, 1981.)
Vepco may be able to take advantage of such improvements in the future.
If so, the need to follow the inspection criteria in V-1 may be eliminated.
Whatever order the Appeal Board enters at the conclusion of this proceeding should be flexible enough to allow for such design improvements.
Also, the Board should be' aware that new information_on the phenomenon of turbine disc cracking will undoubtedly be developed in the future.
Southwest Research Institute, for example,,is examining cracked discs from North Anna.
The
" Interim Data-Summary" from that ef fort is included as Applicant's Exhibit V-22.
i i
t i
II.
THE PROBABILISTIC ANALYSIS UPDATED A.-
Westinghouse's P Calculations y
i, Besides developing-the deterministic approach of Applicant's Exhibit V-1, Westinghouse updated its portion of 4
the.probabilistic. approach by recalculating.tr.e probability of
- turbine missile generation P taking into account the y,
mechanism of stress corrosion cracking.
As noted above, Vepco's original evidence in this case aimed to show, beginning with~a Westin.ghouse fault-tree analysis, that the probability or missile generation from design and destructive overspeed was so low that no additional prctective measures needed to be taken over the life of'the plant.
In-light of the recent cracking experience attributed to-stress corrosion, the probability analysis has changed.
The new method of calculating P is detailed in Westinghouse y
Topical Report WSTG-1-P, " Procedures for Estimating the i
Probability of Steam Turbine Disc Rupture from Stress Corrosion i
Cracking" (May 1981) (Applicant's Exhibit V-3).
i 4
Using'the procedures in Exhibit V-3, it can be seen
' that the calculated probability of missile generation increases t
with. time.
The P results for North Anna, as a function of y
operating time, are given in the Westinghouse Turbine Missile 4
Reports, Applicant's Exhibits V-14 and V-15 and, for
. _..-. _. _... _.. _. _. _., _ _,.. _. _. _ ~. -.... _. _.. - _ _ _ -, _.. _ _ _ _,. ~.,
- destructive overspeed (the analysis of which has not changed since the 1979 hearing), from Table 10.2-1 of the FSAR, entered in the tr script of the June 20, 1979, hearing following Tr.
580-Operating shee(d) oher(design oher(destructive rated P
P P
Unit Time (yrs.)
speed) speed)
~0
-8
-6 North Anna 1 1
1.29 x 10 1.54 x 10 1.7 x 10
-5
-7 2
7.23 x 10 6.42 x 10
~4
-6 3
5.14 x 10 3.90 x 10
-6
-8
~0 North Anna 2 1
1.45 x 10 1.75 x 10 1.7 x 10
-5
-7 2
8.02 x 10 7,77 x 1g
~4
-6 3
5.62 x 10 4.28 x 10 B.
Stone & Webster's P Calculations Before Westinghouse's deterministic approach in Applicant's Exhibit V-1 was fully developed, Vepco asked Stone
& Webster to use the revised Westinghouse E calculations to y
recalculate P The results of that analysis are found in 4
Applicant's Exhibit V-4.
There are at least seven conservative assumptions incorporated into the Stone & Webster analysis, listed on page 12 of the report.
One feature of the Stone & Webster analysis deserves special mention.
Stone & Webster has found that the potential target of most concern is the Main Steam Valve House (MSVH),
. because of possible damage from fragments of the number 2 disc.
No penetration of the MSVH wall is expected, but scabbing is a possibility.
But as noted in the Stone & Webster report, the piping in the MSVH is typically large-bore, heavy-wall carbon steel, and the valves are designed to fail in a safe position if loss of electrical power or air occurs.
Also, the MSVH contains major quantities of large-member structural stcel that supports pipe break restraints.
Accordingly, serious damage is unlikely to occur from scabbing (see Applicant's Exhibit V-4, Appendix D, and Applicant's Exhibits V-17 through V'21).
The probabilities P for North Anna 1 and 2 from 4
Exhibit V-4, Table C-19, are set out below.
They vary depending on what assumptions are made, particularly assumptions about "scaobing."
Three different assumptions are used:
l 1 i Criterion A -
Present conservative NRC approach which equates the initiation of scabbing within a safety regulated cubicle with a damage pr bability of 1.0 Criterion B -
Slightly unconservative approach which neglects scabbina damage if misaile perforation is prevented.
5% Scabbing -
Assumes a realistic probability of 5%
for ensuing safety-related damage if a missile strike results in scabbing without perforation With those assumptions, the calculated prcbabilities are as follows:
TOTAL DAMAGE PROBABILITY (P4)
UNIT 1 P /0 NIT / OPERATING PERIOD Operating Time (yrs.)
Criterion A Criterion B 5% Scabbing
-6
~7
~
1 1.02 x 10 4.81 x 10 5.08 x 10
-5
-6
-6 2
2.27 x 10 1.05 x 10 2.13 x 10 3
1.53 x 10'4 2.11 x 10 9.63 x 10
-6
-6 UNIT 2 P / UNIT / OPERATING PERIOD Operating Time (yrs.)
Criterion A Criterion B 5% Scabbing
-7
~7 1
8.68 x 10 4.79 x 10 4.99 x 10~
-5
-6
-6 2
1.70 x 10 1.38 x '10 2.16 x 10
~4
-6
-6 3
1.14 x 10 3.45 x 10 8.96 x 10 i
l l
l i
1
. Thus the probability of unacceptable damage P over the first g
three operating years for North Anna 1, calculated using a
-6 number of conservative assumptions, ranges from 2.11 x 10 to
~4
-6 1.53 x 10
- for North Anna 2 it ranges from 3.45 x 10 and
~4 1.14 x 10 By comparison, the corresponding P over three 4
-6 years' time for a plant whose annual P is a constant 10 er 4
annum is 1 - (1 6)3, or about 3 x 10 -6 Thus the P 's for North Anna 1 and 2 after 4
approximately three y5ars of operating time are not
~0 inconsiscent with the Staff's guideline of "approximately" 10 per year, in the words of Standard Review Plan 2.2.3, particularly given the seven conservative assumptions used by Stone & Webster.
C.
Comparison of the Deterministic and Probabilistic Methods It would be possible to set an inspection interval using, not the Westinghouse deterministic inspection criteria in Applicant's Exhibit V-1, but Stone & Webster's probabilistic P
analysis in Applicant's Exhibit V-4.
If there were a 4
probabilistic safety goal, one could simply calculate tne length of time necessary for P to rise to the level of the 4
safety goal.
Despite Vepco'.s reliance ca probabilities in 1979, Vepco now prefers to rely on the deterministic approach
- embodied in Applicant's Exhibit V-1.
Vepco believes it is appropriate to do so, for the following reasons.
1 1.
It is permiscible to use the deterministic method The probuoilistic approach derives from Regulatory Guide 1.115 and the Standard Review Plan, SRP 2.2.3.
The Regulatory Guide states explicitly that other methods are permissible:
This guide reflects current NRC staff practice.
Therefore, except in those cases in which the applicant proposes an acceptable alternative method for complying with specified portions of the Commission's regulations, the method described herein is being and will continue to be used in the evaluation of submittale,.
Regulatory Guide 1.115 (Attachment 3) at 1.115-4 (emphasis added).
It appears that the NRC Staff have concluded (see ) that the Westinghouse disc inspection criteria are one such " acceptable alternative method."
2.
It is preferable to use the e
deterministic method What is more, Vepco submits that the deterministic approaca is preferable to the probabilistic for calculating inspection intervals, at least at the present time, for a number of reasons.
. First, the deterministic method has received far more attention than the probabilistic method.
The criteria in Applicant's Exhibit V-1 have been developed after many months of discussions between Westinghouse and the NRC Staff.
They are being relied on by other licensees.
They have been reviewed and approved by the NRC Staff 'see Attachment 2).
By contrast, Westinghouse's probability calculations are still under review by the Staff (Attachment 2, Letter from Robert A.
Clark, at 2).
And as far as Vepco knows, the probabilistic method has never been used to set inspection intervals.
Second, to use the probabilistic method to set an inspection interval requires that'there be a probabilistic safety goal, and there is none.
The Commission's " Plan for Developing a Safety Goal," 45 Fed. Reg. 71023 (Oct. 27, 1980),
has not come to a conclusion.2/
2/
Public meetings to elicit public comment on the safety goal project have been postponed.
46 Fed. Reg. 2672.1, col. 2 (May 14, 1981).
The Plan mentions a set of interim criteria suggested by
~
the NRC Probabilistic Analysis Staff.
These criteria provide that "no action" need be taken if the estimaged probability of severecoredamageperyear_jsless_ghan10 If the estimated probability is 10 to 10 the licensee is supposed to " consider" correcting the probleg.
A prgblem causing a probability of core damage from 10 to 10 is supposed to be corrected in a matter of " years."
Even using the extremely conservative scabbing " Criterion A,"
the cumulative probability after three year's time will only just have passed the level at which, if it were a one-year probability, the problem would need to be corrected in " years."
By that time the licensee, 3ing the 43.3-month inspection interval advocated above, would be preparing to inspect the turbine witnin the next eight months.
- -6 The 10 per year probability guideline is not a safety goal.
Tha Commission, in Florida Power & Light Co. (St. Lucie Nuclear Power Plant, Unit No. 2), CLI-81-12, 13 NRC (June
-6 15, 1981), made clear that the 10 probability figure is only
-6 a rough guideline.
In Vepco's view the 10 guideline should only be used as a sort of rough screening device to eliminate from consideration safety issues that can easily be shown to be
-6 unworthy of further' consideration.
Failure to meet the 10 per year guideline does not mean that a plant is not sufficier' ly safe; it may merely mean that additional analysis is needed.
In the case of turbine missiles, that additional analysis is embodied in Exhibit V-1.
Third and finally, the probabilistic analysie inccrporates a great many assumptions that are extremely difficult to quantify.
For example, there are seven assumptions in Stone & Webster's analysis that cannot be quantified without considerable time and effort.
Likewise, the ri;k of " scabbing" is difficult to quantify.
Depending on the scabbtng assumptions one uses, the probabilistic analysis can justify operation for one, two, three, or more years.
Thus, about all one can do with the probabilistic analysis, without an enormous amount of analytical work, is use it to calculate a sort of rough upper bound on the probabilities by taking care to incorporate conservative
_.~,_ __
o assumptions.
It is useful, once again, as a sort of rough screening device.
1 D.
Conclusions.
In short, Vepco urges the Appeal Board to find the following:
1.
That General Design Criteria 4 is met, and that reasonable assurance that the public health and safety will be protected is provided, by 'lepco's commitment to use inspection intervals calculated in accordance with Applicant's Exhibit V-1, as approved by the NRC Staff in Attachment 2; 2.
That it is appropriate, for health and safety purposes, to use the inspection interval determined by the most critical disc (that is, the one with the smallest inspection int.erval) that is not contained by the turbine casing, as shown by Westinghouse's calculations in accordance with Applicant's sxhibit V-2; and i
i l
.v.,
.y.
,,ry..
,,.,y.,_
y.,,y.,
,.-_.-.~.m,
, 3.
That no technical specification requiring a particular inspection interval is necessary.
The Board's order, or 7epco's written commitment-(Applicant's Exhibit V-6), along with incorporation of the i
inspection schedule in the Station procedures, will be sufficient assurance that the inspections will be carried out as planned.
III.
OTHER MISCELLANEOUS MATTERS A.
Cause of the Recent Turbine Disc Cracking The Board has at various times expressed an interest in the mechanism of the recent disc cracking.
As noted above, Westinghouse attributen it to stress corrosion.
Some information on stress corrosion cracking can be found in Applica".t's Er.hibit V-16 (Dr. Seth's paper) and V-1 (the Westinghouse inspection c:iteria report) at page 11.
1
,-,_.s
---,y-,4 7--
- B.
Non-Safety Grade Equipment Issue The Board said, in ALAB-551, that it would carry forward the question of whether the generic issue of nonsafety-grade equipment used to mitigate transients has a reasonable nexus to the turbine missile issue.
Vepco's counsel.
is aware of no new information that would prove he'.pful on this issue.
Respectfully submitted, 4
/s/ James N.
Christman James N.
Christman, Counsel for Virginia Electric and Power Company 1
Michael W. Maupin, Esq.
James N.
Christman, Esq.
James M.
Rinaca, Esq.
Hunton & Williams 707 East Main Street P.O.
Box 1535 Richmond, VA 23212 I
e h
LIST C# APPLICANT'S EXHIBITS AND ATTACHMENTS Letter of AuTust 31, 1981, from Robert A.
Clark, NRC, to R.
H. Leasburg, Vepco, asking Vepco to commit to use the Westinghouse inspection criteria, and attar.hed Safety Evaluation Report Regulatory Guide 1.115 (Rev.
1, July 1977)
Standard Review Plan 2.2.3 Applicant's Title or Exhibit No.
Description V-1
- Westinghouse Steam Turbine Ganerator Division, MSTG-i-P, Criteria for Low Pressure Nuclear Turbine Disc Inspection ; June 1981) t V-2
- Westinghouse Steam Turbine Generator Division, WSTG-2-P, Missile Energy Analysis Methods for Nuclear Steam Turbines (May 1981)
V-3
- Westinghouse Steam Turbine Generator Division, WSTG-1-P, Procedures for Estimating the Probability of Steam Turbine Disc Rupture from Stress Corrosion Cracking (May 1981)
V-4
- Stone & Webster Engineering Corp.,
Summary Report, Turbine Missile Damage Probability Analysis for North Anna Units 1 and 2 (Rev.
4, Oct. 15, 1981)
V-5 Letter of September 22, 1981, from R.
H.
Leasburg, Vep;o, to Harold R.
Denton, NRC, committing to use the Westinghouse inspectioa criteria
,,-.m
,.e.-
,~,
-~=r,-
+-m---,-~-------.-~+--wc
,e--w--w-*---+=---v*--v-
,ww<---.-v.-
--e-
.4-**nw--
4-V-6 Letter of October 7, 1981, from S. J. Piermont, Westinghouse, to S. Eccles, Vepco, stating the recommended inspection intervals for North Anna 1 and 2, based on the Westinghouse inspection criteria V-7 North Anna 1 LPl Low Pressure Disc Ultrasonic Inspection report, 1/16/81 V-8 Three Mile Island "ni't #2 Low Pressure Disc Ultrasonic Inspection Report, 2/2/81 V-9 North Anna 2 Low Pressure Disc Ultrasonic Inspection Report, 4/7/80 V-10
- Computer print-outs with materials properties and operating temperatures and stresses for North Anna 1, 2/4/81 V-ll
- Computer print-outs with materials properties and operating temperatures and stresses for North Anna 2, 9/22/80 V-12 Turbine Missile. Report (HP296-LP281-LP281), CT-24821, Rev. O, August 1980 (North Anna 1)
V-13 Turbine Missile Report (HP296-LP281-LP281), CT-24859, Rev. O, October 1980 (North Anna 2)
V-14
- Turbine Missile Report, Results of Probability Analysis of Disc Rupture and Missile Generation, CT-24822, Rev. 1, March 1981 (North Anna 1)
V-15
- Turbine Missile Report, Results of Probability Analysis of Disc Rupture and Missile Generation, CT-24860, Rev. 1, March 1981 (North Anna 2)
V-16
- B.
B.
Seth, Notes on Low Pressure Disc Update (paper presented at the Steam Turbine-Generator Technology Symposium, Charlotte, North Carolina, October 21-23, 1980, and Phoenix,
---y-
,y.-
,-y ee--g ieg,-
.,,-*------r----e-w-,
e
-~w++--,,--+-,m-=,y
,e--,,we.
,,ne.
-r--
- -me-w*e---
b
l i Arizona, Govember 5-7, 1980)
V-17 through
-21 Photographs of inside of Main Steam Valve House V-22 Southwest Research Institute, Investigation of Keyway Cracking in LP Turbine Discs, SWRI Project 02-6734-001, EPR.I RP 1398-9, Interim Data Summary (Oct. 16, 1981) l
- Contain information proprietary to the Westinghouse Electric Corporation.
.d
- s PROOF OF SERVICE
'I hereby certify that.I-have this day served Virginia Electric and Power Company's Memorandum of Counsel upon each of the persons named below by placing a copy in the mail, first-class postage prepaid:
Secretary U.S.
Nuclear Regulatory Commission Washington, D.C.
20555 Attention:
Chief, Docketing and Service Section t
Alan S.
Rosenthal Administrative Law Judge Atomic Safety and Licensing Appeal Board U.S. Nuclear Regulatory Commission Washington, D.C.
20555 Dr. John H.
Buck Atomic Safety and Licensing Appeal Board U.S. Nuclear Regulatory Commission Washington,.D.C.
20555 Dr. Lawrence R. Quarles Atomic Safety and Licensing Appeal Board U.S.
Nuclear Regulatory Commission Washington, D.C.
20555 Daniel T. Swanson, Esq.
U.S. Nuclear Regulatory Commission Washington, D,C.
20555 By /s/ James N.
Christman James N.
Christman, Counsel for Virginia Electric and Power Company DATED:
October 21, 1981 4
,,-