Resolution of Generic Safety Issues: Issue 30: Potential Generator Missiles - Generator Rotor Retaining Rings (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )
The April 13, 1979 incident at Sweden's Barseback-1 nuclear plant involving the failure of a generator rotor retaining ring was identified by AEOD in 1982 as a potential generic safety issue.490 As a result of the AEOD concern,799 NRR agreed to prioritize the issue.
There have been 32 recorded events of failure of generator rotor retaining rings. The retaining rings restrain the radial forces generated by rotor coil ends, insulation, and packing blocks. For an 1100 MWe plant, the generator rotors may be up to 6 feet in diameter and 7 feet long and may weigh approximately 4,000 pounds. In 6 recorded cases, the failure of these retaining rings resulted in major damage to the generator and, in 2 cases, to the plant structure. The pieces of the broken retaining rings are expelled axially as compared to the turbine missiles (see Item A-37) which are expelled tangentially. The failures have been principally experienced on the exciter end of the generator, the end away from the turbine. Of concern is the potential of these large missiles to do damage to the plant. Further details of retaining ring failure can be found in EPRI EL-3209.819 All plants are affected by this issue.
Since current requirements for turbine and generator orientation are to preclude damage from turbine missiles, the positioning may not be optimum to protect against missiles resulting from failed rotating rings. In such cases, the only solution may be the erection of shields to restrain these missiles. In addition, most of the failures have resulted from stress corrosion cracking induced principally by water. A higher frequency of inspections for crack presence plus added precautions to assure a dry environment would reduce the probability of crack initiation and growth leading to catastrophic failure.
The same assumptions used in the turbine missile evaluations in Item A-37 will be used for this issue. These are: (1) the probability that a missile penetrates a barrier is 1, and (2) the probability of radioactive release given damage to safety-related equipment is 1.
An estimate of retaining ring failure likelihood is 10-3 per retaining ring.819 For the 32 recorded events of retaining ring failure, 6 led to extensive damage for a failure rate of 0.2 extensive damage occurrences per retaining ring. Six extensive damage events per ring failure are greater than the number of events resulting in missile ejection, since extensive damage events included events in which only the generator received extensive damage. However, due to the uncertainty involved in these infrequent events, 6 events are used in these calculations. There are 2 retaining rings per plant (generator). Assuming an average plant life of 40 years, the frequency of expected extensive damage events is given by:
The probability of a safety-related target being struck by a missile is estimated to be 0.01. This value is based upon previous staff efforts related to turbine missiles where it was found that the prediction for low trajectory missiles for unfavorably-oriented generators generally fall in the range of 10-2 to 10-3.820,821 Thus, the frequency of an extensive event occurring and damaging some safety-related equipment is 10-7/RY.
As in the case of turbine missiles (Item A-37), a realistic estimate of a radioactive release to the environment is a gap release and not a core-melt. The consequences of the most likely release would be 75,000 man-rem per occurrence for the 90 PWRs and 40 man-rem for the 44 BWRs, based on the radioactive release categories described in WASH-1400.16 The computer program CRAC2 applied to a typical Midwest site meteorology (Braidwood) was used for the dose calculation.64 An average population density of 340 persons per square mile was used over an area which extended from an exclusion zone of one-half mile about the reactor out to a 50-mile radius about the reactor. This results in a public risk exposure of 19.1 man-rem for the PWRs and 5 x 10-5 man-rem for the BWRs, a total of less than 20 man-rem for all reactors. The change in transient-induced accident frequency created by generator ring missiles was calculated. However, the low initiating event frequency of 10-5/RY (as compared to transient initiators which are greater than 1/RY) does not result in a significant change in risk.
Industry Cost: Assuming that shields would be required on 10 plants at $1M/plant results in total utility costs of $10M.
NRC Cost: It is estimated that NRC costs are $100,000 to evaluate this issue better.
Thus, the total cost associated with the solution to this issue is $10.1M.
Based on a public risk reduction of approximately 19 man-rem, the value/impact score is given by:
This issue should be DROPPED from further consideration.