ML20040B881
| ML20040B881 | |
| Person / Time | |
|---|---|
| Site: | North Anna  |
| Issue date: | 01/22/1982 |
| From: | Campe K Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20040B860 | List: |
| References | |
| NUDOCS 8201260508 | |
| Download: ML20040B881 (11) | |
Text
r UNITED STATES OF AMERICA j
NUCLEAR REGULATORY COMMISSI0fl BEFORE THE ATOMIC SAFETY AND LICENSING APPEAL BOARD In the Matter of
)
)
VIRGiflIA ELECTRIC AND POWER
)
Docket Nos. 50-338 COMPANY (VEPCO)
)
50-339
)
(florth Anna Power Station, Units 1 )
and 2)
)
NRC STAFF TESTIMONY OF KAZIMIERAS M. CAMPE ON THE EFFECT OF RECENT WESTINGH0l'SE TURBINE CRACKING EXPERIENCE ON ESTIMATED HISTORICAL TURBINE FAILURE RATE Q.
Please state your name and position with the NRC.
A.
My name is Kazimieras M. Campe.
I am employed at the U.S. Nuclear Regulatory Commission as a Senior Site Analyst in the Siting Analysis Branch.
Q.
Have you prepared a statement of your educational and professional qualifications?
A.
Yes.
It is attached to this testimony.
O.
What is the purpose of your testimony?
A.
The purpose of my testimony is to respond to the Appeal Board's Memorandun and Order of October 23, 1981 with respect to the relation-ship between the resolution of the disk cracking problem and the estimated turbine failure probabilities presented in our previous!y filed testimony on florth Anna Units 1 and 2 turbine missile risk.
8201260508 820122 PDR ADOCK 05000338 T
pon
- 0.
In the Staff's April 27, 1979 estimates of the turbine missile risks for North Anna Units 1 and 2, how was the potential for turbine failure factored into the analysis?
A.
The Staff evaluation of the risks associated with potential turbine missiles for North Anna Units 1 and 2 was based, in part, on the estimated frequency of the expected failure rate of turbine disks by brittle fracture. A failure rate of 6 X 10-5 per turbine year was used in the Staff's analyses.
Q.
What is this failure rate based on?
A.
This rate is derived from considering the historically observed number of turbine failures (with missile ejection) and the total number of turbine years of operation for both nuclear and conven-tional steam units.
It was not derived from an application of identifiable failure mechanisms such as metallurgical or mechanical disk properties or crack formation and growth due to environmental conditions involving temperature, stress, and corrosion.
It should be noted, however, that the above environmental failure causes were considered by the Staff when describing the conservatisms associated with the use of the historical turbine failure rate.
Q.
How does the recent Westinghouse turbine disk cracking experience affect the turbine failure rate used by the Staff?
A.
The new turbine information stemming from inspection of Westinghouse turbine disks and discovery of stress corrosion cracking indicates that there is a potential for brittle fracture failure of the disks
during operation. A logical conclusion that can be drawn is that continued operation of a disk with cracks would lead to disk failure within a time period characterized by the intial crack size and the j
crack growth rate associated with the disk material and operational environment.
Consequently, if all Westinghouse units suspected of a
cracked disks were permitted to operate indefinitely without any corrective measures, failures would presumably occur, and this would affect the current estimate of the average turbine failure rate by brittle fracture.
However, this is not the case, as evidenced by the Staff activities currently being pursued with respect to all operating reactors, including North Anna Units 1 and 2.
The i
inservice inspection, the crack growth rate estimation, and the corrective measures being taken with respect to defective disks, give assurance that the conservatism associated with the use of the previously estimated turbine failure rate is not invalidated.
Q.
What is the basis for your position that the previously estimated
]
turbine failure rate is still appropriate for use in the probablistic risk assessment with respect to turbine missiles?
}
A.
The following considerations support this view:M a.
Current inservice inspection intervals are based on crack growth rates which incorporate the recent stress E A more detailed description of the disk cracking experience and the measures being taken to the likelihood of disk failure is presented in "NRC Staff Testimony of Warren S. Hazelton and Clifford D.
Sellers Regarding Turbine Disc Cracking" (Testimony of Hazelton and Sellers).
J
. corrosion cracking experience. These intervals are much shorter than those that were recommended at the time when the Staff did the turbine missile risk analysis for florth Anna Units 1 and 2.
b.
The ability to detect key way cracks in the field, as recently developed by Westinghouse, permits a more thorough inservice inspection than previously
- possible, c.
Corrective measures taken by the utilities with respect to defective disks returns the rotor integrity to an acceptable level until the subsequent inservice inspec-tion.
In view of the above, the Staff believes that the continued use of the historical turbine failure rate in nuclear plant risk assessments, including North Anna Units 1 and 2, is justified.
Q.
In "t1RC Staff Testimony Regarding Turbine Missiles" (dated April 27, 1979) a factor of improvement in the probability of disk failure and missile ejection, P, at florth Anna Units 1 and 2 was presented.El i
1 SI All further page references are to this April 27, 1979 testimony.
5-How does the Westinghouse turbine disk cracking experience affect the factors of improvement?
A.
The P failure probability is made up of two distinct modes (pp. 4-5),
y the design overspeed failure and the destructive overspeed failure.
The recent Westinghouse cracking experience does not have an effect on the destructive overspeed failure probability since this is a failure mechansim that is not material dependent. Rather, it occurs because of a component or system failure in the overspeed protection system.
Therefore, the factors of improvement in the value of the destructive overspeed probability presented in our earlier testimony (pp.19-20) are still valid.
However, the factors of improvement in reference to disk integrity (which affects the probability of design overspeed failure) are subject to reconsideration with respect to the recent Westinghouse operating experience.
Specifically the factor of improvement in the desion overspeed probability was perceived in two parts:
(1) improvement due to an 1
increase in material fracture toughness (pp. 28-33) and (2) improvement I
due to preservice inspection (pp. 34-35).
In the discussion of the factor of improvement due to materials toughness, it was assumed that crack growth was due to stress corrosion assisted fatigue. On this basis the factor was determined by taking the ratio of the density function of cracks that could grow to critical size in two materials with differing materials toughness (p. 31).
The i
initial crack size was calculated from stress corrosion assisted fatigue loading over the expected life of the turbine.
This sane calculation of the initial flaw size necessary for a failure to occur may be made l
O I !
for any assumed crack growth mechanism and for any period of time.
Or, stated in another way, the factor of improvement that was calculated in the design overspeed failure probability does not rely on the method of crack grcwth.
Rather, over a given period (load cycles or time), it depends upon the disk material fracture toughness level.
However, it appears that the crack growth mechanism that has actually been found in Westinghouse turbine disks is primarily stress j
corrosion cracking. Hence, the factor of inprovement attributed to increased fracture toughness is no longer a principal source of l
consideration with respect to the estimated turbine failure proba-bility used by the Staff.
The improvenent due to preservice inspection, however, is still valid, since that is not affected by stress corro-sinn considerations.
In summary, the improvement factors described in our previous
{
testimony are affected (in part) by the Westinghouse disk cracking experience. Specifically, the improvement derived from high fracture toughness is offset by the potential for stress corrosion cracking, i
1 Q.
If the improvenent due to higher fracture toughness is offset by the j
susceptibility to stress corrosion cracking, then aren't the pre-viously described improvement factors somewhat reduced, 30 that there is less conservatism with respect to tne historfral failure proba-bility used by the Staff?
l A.
No. Once stress corrosion cracking was identified as a generic problem, Westinghouse undertook an inspection effort in order to screen for cracks which may lead to disk failure.
Part of this effort involved a
a
1 -
the development of an ultrasonic testing (UT) technique which permits l
a turbine disk to be inspected for flaws without turbine rotor dis-assembly.
The technique is capable of detecting flaws much smaller l
than critical size with good reliability, thus allowing the affected material to be repaired, replaced, or returned to service with a known flaw for a given period of time, as determined from a failure evaluation.
In addition to the UT program, Westinghouse has developed a set of criteria with respect to the inspection and certification of low pressure nuclear turbine disks.
The Staff has reviewed and evaluated the Westinghouse criteria for crack growth rate, critical crack size determination, and the fraction of the critical crack size that is allowed for continued operation. We have found the Westinghouse criteria to be acceptable and believe that adherence to them will result in a low likelihood for disk failure by brittle fracture.
In view of the above, it has become more appropriate to relate improvement factors to the preservice and inservice inspection schedules, instead of the increased material fracture toughness.
Q.
What degree of improvement do you expect from the implementation of the inspection program referred to above?
A.
Although numerical estimates of the improvement factors have not been made, it is the Staff's judgment that they are sufficiently large such that the disk failure probability is no higher than that used by the l
)
t
, Staff in the Safety Evaluation Report, as well as in our previously filed testimony.
Q.
What is the basis for this viewpoint?
A.
The UT inspection, in con. function with crack growth modeling, is a reliable method of detecting disk flaws and preventing them from reaching critical size.
This represents a vest improvement over the conditions that were in effect prior to the seventies, when in-the-field rotor inspection, especially near the hub, was not possible. Presently, the turbine material degradation, if any, will be monitored and corrected throughout the proposed forty year life of the plant (see Testirony of Hazelton and Sellers).
In our previous testimony, we assumed that if a crack was present in the turbine disk, it will never be detected and repaired. Hence it is reasoreble to conclude that the turbine inspection program leads to a substantial improvement factor with respect to the failure probability used in our turbine missile risk assessment.
O.
With respect to another aspect of the turbine missile risks, Westing-house has indicated that the results of recent scale model tests with turbine disk ruptures lead to some revisions of the estimated disk fragment exit energies for their turbines. How does this affect your previous testimony?
A.
The impact of this on the Staff's evaluation of turbine missile risks is believed to be negligible for the following reasons:
-9 Westinghouse scale model tests were conducted to deter-a.
mine the effects of non-symnetric impacts of disk fragments on turbine blade stator rings. This effect was anticipated by the Staff during its evaluation of North I
Anna Units 1 and 2, and prior to the Westinghouse tests.
The Staff missile energy estimates were based on the assumption of non-symmetric impacts.
Staff evaluatior of North Anna Units 1 and 2 was based on b.
neglecting the presence of plant structures and barriers.
Hence if there were some changes in the estimated missile energies, this would not affect significantly the risk evaluation.
In view of the above, we do not believe that the revised missile energies will change our estimates of the turbine missile risks for North Anna Units 1 and 2.
In summary, how does the Westinghouse turbine rotor cracking experi-O.
ence relate to the Staff's evaluation of the turbine missile risks for North Anna Units 1 and 2?
The Staff believes that our previous estimates, based in part on the A.
historically detemined turbine failure probability (P ), are still 1
The Westinghouse inspection program is acceptable and provides valid.
for a lower failure probability than what was used in the Staff's analyses (see Testimony of Hazelton and Sellers).
s KAZIMIERAS M. cat:PE PROFESSIONAL QUALIFICATIONS SITING ANALYSIS BRANCH OIVISION OF ENGINEERING I am a member of the Siting Analysis Branch of the Office of Nuclear Reactor Regulation of the United States Nuclear Regulatory Commission.
My duties include the identification and evaluation of hazards to the safe operation of nuclear power plants due to natural and man-made events external to the plant. This involves the review and independent assessment of nearby industrial, transportation, and military facilities with respect to potential hazards such as explosions, fires, missile impacts, ground subsidence, and toxic gases.
Prior to being assigned to the Siting Analysis Branch, I was employed within the Accident Evaluation Branch (formerly known as the Accident Analysis Branch) where I initially began my review efforts in the area of offsite hazards.
While at the Accident Evaluation Branch, I was a lead technical reviewer on turbine missiles, tornado missiles, and control room habitability systems.
I have prepared most of the technical input for the Regulatory Guide 1.115 and Standard Review Plan 3.5.1.3 on turbine missiles.
I have also prepared revised Standard Review Plan (SRP) 2.2.1-2.2.2, " Identification on Potential Hazards in Site Vicinity", SRP 2.2.3, " Evaluation of Potential Accidents,"
SRP 3.5.1.5, " Site Proximity Missiles (Except Aircraft)", and SRP 3.5.1.6,
" Aircraft Hazards."
o In 1974 I co-authored a paper with K. C. Murphy of the Accident Analysis Branch on an overview of methods for evaluating control room habitability In 1975 I co-authored a paper with Dr. J. Read of the Accident systems.
Analysis Branch on the subject of high trajectory turbine missile strike probabilities.
I graduated from the University of Connecticut where I received B.S. and M.S.
Between 1960 degrees in Mechanical Engineering 1958 and 1960, respectively.
and 1962 I completed some advanced mathematics courses at the Rensselaer Polytechnical Institute branch in East Hartford, Connecticut.
During this period I was employed by Pratt and Whitney at the CANEL Analytical Physics Group as an analytical engineer.
From 1962 to 1966 I attended Purdue From 1966 to University, where I received a Ph.D. in Nuclear Engineering.
1972 I was employed by Hittman Associates, Inc. where I worked in the Radioisotope Department. During this period my responsibilities included radiation shielding analyses, radioisotopic generator design, and computer code development for reactor core physics calculations.
I have been employed by the Nuclear Regulatory Commission since 1972.
--