ML20141M950
| ML20141M950 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 02/26/1986 |
| From: | Standerfer F GENERAL PUBLIC UTILITIES CORP. |
| To: | Travers W Office of Nuclear Reactor Regulation |
| References | |
| 0398A, 398A, 4410-86-L-0033, 4410-86-L-33, NUDOCS 8603030112 | |
| Download: ML20141M950 (5) | |
Text
{{#Wiki_filter:L i GPU Nuclear Corporation g g{ Post Office Box 480 Route 441 South Middletown, Pennsylvania 17057-0191 717 944 7621 TELEX 84-2386 Writer's Direct Dial Nurnber: (717) 948-8461 4410-86-L-0033 Document ID 0398A February 26, 1986 TMI-2 Cleanup Project Directorate Attn: Dr. W. D. Travers Director US Nuclear Regulatory Commission c/o Three Mile Island Nuclear Station Middletown, PA 17057
Dear Dr. Travers:
Three Mile Island Nuclear Station, Unit 2 (TMI-2) Operating License No. DPR-73 Docket No. 50-320 Clarification of Basis for Technical Specification Change Requests 49 and 51 With Respect to Control Room Emergency Air Cleanup System INTRODUCTION Technical Specification Change Request (TSCR) No. 51 (Reference 1), among others things, requested deletion of all requirements for availability and operation of the class IE Diesel Generators. One of the loads supported by the emergency diesel generators is the Control Room emergency air cleanup system. A request was made in TSCR No. 49 (Reference 2) to delete certain functions of this system which relied on the diesel generators in the event of loss of off-site power. Based on" discussions with the NRC staff, GPU Nuclear has re-evaluated its request to delete this system and now proposes to retain in the Technical Specifications the operability requirements for this system and requests only that requirements for back-up on-site AC power supply requirements be deleted. This letter presents the basis for supporting this system with off-site power without the requirement for emergency diesel generator support. O 8603030112 860226 i Ipgj ADOCK 050 g O y PDR P I GPU Nuclear Corporation is a subsidiary of the General Public Utilities Corporation
-O Dr. Travers February 26, 1986 4410-86-L-0033 BACKGROUND An evaluation (Reference 7) prepared in support of TSCR No. 49 discussed the consequences of various accidents on the habitability of the TMI-2 Control Room. This evaluation concluded that there is no credible accident at TMI-2 that can challenge the habitability of the Unit 2 Control Room. The most limiting radiological event was a TMI-l LOCA with loss of off-site power. The evaluation demonstrated that by taking credit for a protection factor of 50 for full-face respirators with iodine canisters that Unit 2 Control Room occupancy could be maintained under this combined event without relying on the Control Room emergency air cleanup system. In fact, the skin dose was reduced by isolating the Control Room ventilation system. The NRC staff expressed concern with this approach and GPU Nuclear responded with the basis for the assumed protection factor in GPU Nucleat letter 4410-85-L-0231 (Reference 3). Further discussion with the NRC staff resulted in the understanding that although the protection factor of 50 may be acceptable under specific actual emergency conditions, it was not appropriate to delete engineered protection features using this assumption. It was also agreed that the combined event of a Unit 1 LOCA and a complete loss of off-site power may be so unlikely that an emergency on-site diesel generator power supply need not be required to assure Unit 2 Control Room habitability. The following sections of this letter demonstrate that the combined Unit 1 LOCA and loss of off-site power is an occurrence of such low probability as not to require Unit 2 emergency on-site AC power. Further, if this extremely low probability event does occur, some time is available to return off-site power before Control Room dose levels exceed current limits if alternate protective devices, such as respirators, are used. PROBABILITY OF COINCIDENT UNIT 1 LOCA AND LOOP, RECOVERY LESS THAN 1 HOUR The probability of a coincident and independent LOOP event at Unit 2 with a Unit 1 LOCA is negligible. This is based upon the site specific LOOP frequencies from Reference 6 and LOCA probabilities typically predicted for similar plant designs (e.g., Reference 4). Therefore, the issue becomes the likelihood of a Unit 1 LOCA occurring which is initiated by a LOOP; the LOOP must also have a duration that will result in overexposure of Unit 2 personnel. From Reference 4, the most rapidly progressing core melt sequence results in a source term release in one (1) hour. Thus, if the LOOP was the initiating event for such a sequence, (and assuming that the LOOP fault also affected Unit 2), off-site power must be recovered within one (1) hour to allow the Unit 2 ventilation system to protect Control Room personnel against any appreciable dose accumulation. The probability of a LOOP induced LOCA which results in a core melt was estimated from References 4 and 5. This probability was calculated to be in the range of 5x10-6/yr to lx10-5/yr. The probability of not recovering 2x10-{/ demand.off-s te power within one (1) hour was estimated from Reference 6 as Thus, the probability of a Unit 1 LOCA and a LOOP resulting
Dr. Travers February 26, 1986 4410-86-L-0033 in unavailability of the Unit 2 ventilation system was determined to be less than 1x10-5/yr. This probability is sufficiently low as to justi'y that the Unit 2 ventilation system need not be placed on the emergency di Jels. CONSEQUENCES AND PROBABILITY OF NOT RECOVERING OFFSITE POWER PRIOR TO 1 HOUR The combined occurrence of a Unit 1 LOCA and Unit 2 LOOP lasting one (1) hour is sufficiently unlikely that further analysis is not necessary. However, GPU Nuclear extended the analysis to consider the consequences and probability of not returning off-site power for longer periods. The study of Unit 2 Control Room personnel exposure during the Unit 1 LOCA referenced in TSCR No. 49 (Reference 7) demonstrates that during the first two hours after Unit 1 core damage (or three (3) hours after initiation of the event) the average thyroid dose will be 135 R/hr with 1350 cfm inleakage and no Control Room air filtration or supplemental protection. This level assumes Regulatory Guide 1.4 specified source terms, i.e., 25% of iodine core inventory is released. For a realistic consequence analysis, the results of NUREG-0772 can be used to evaluate actual Control Room doses, with credit taken for a respirator iodine canister protection factor. NUREG-0772 (Reference 8) states that "under most LWR accident conditions CsI (or HI) would be the expected chemical form of iodine in the vapor phase in the primary system, although the formation of some elemental iodine cannot be precluded for certain accident conditions." A protection factor of 50 has been applied to 90% of the source term iodine, as a particulate, and 10% of the iodine has been assumed to remain as elemental iodine and therefore unfiltered. These results are therefore still conservative because other test results referred to in Reference 2 demonstrate the effectiveness of charcoal in filtering elemental iodine. Taking credit for a protection factor of 50 for the particulate iodine fraction for the full-face respirators with iodine canisters, which Control Room personnel would use as back-up, the two hour thyroid dose is on the order of 30 Rem, consistent with General Design Criteria 19 and the Standard Review Plan. At this point, i.e., three (3) hours after the initiation of the combined event the probability of not recovering off-site power is estimated to be 2x10-$ emand resulting in the probability for the combined event of less /d than lx10-6/ year, which can be considered incredible. Additionally, the necessary Control Room staff can be exchanged at this point, i.e., three (3) hours after the event initiation and because of declining dose levels the new staff can man the Control Room for another three (3) hours without exceeding 30 Rem thyroid. The probability of not recovering off-site power at the end ofthisperiod,nowsix(6]hoursafterinitiationofthecombinedeventis reduced to less than 5x10 . The resulting probability of a Unit 1 LOCA and a LOOP exceeding six (6) hours is so low that such an occurrence can also be considered incredible. Finally, it should be noted that TMI-2 is unique in that no actions are required to be taken from the Unit 2 Control Room to maintain Unit 2 in a safe shutdown condition. During a Unit 1 LOCA, ongoing recovery activities at TMI-2 would be terminated and systems placed in a safe configuration. Thus,
I Dr. Travers February 26, 1986 4410-86-L-0033 the TMI-2 operators' sole responsibility would be monitoring of ambient Unit 2 conditions. The uniqueness of TMI-2 is also reflected in the limited consequences of potential Unit 2 accidents, as presented in Reference 9. That reference, which was reviewed by the NRC, analyzed a broad spectrum of j possible accidents at TMI-2 and demonstrated that for any of these accidents, the resulting off-site doses were well within the guidelines of 10 CFR Part
- 100, i
SUMMARY
This letter is to reconcile concerns expressed by the NRC staff with respect 2 i to TSCR No. 49 and No. 51 that deal with the Control Room emergency air i cleanup system and deletion of the energency diesel generators. GPU Nuclear agrees to retain the Control Room emergency air cleanup system. This letter demonstrates that the probability of the combined Unit 1 LOCA and loss of off-site power is sufficiently low as to not require energency diesel generator power to the Control Room air cleanup system. Further, the availability of respirators with iodine protection canisters, assuming a PF of 50 as described above, adds assurance that Control Room manning can continue until the Control Room Ventilation System can be re-established. Sincerely, I . R. Standerfer i l Vice President /D rector, TMI-2 FRS/EDF/eml Attachment 1 i l r i I a i i
ATTACHMENT 1 (4410-86-L-0033) REFERENCES 1. GPU Nuclear letter 4410-85-L-0135, Technical Specifications Change Request No. 51 and Recovery Operations Plan Change Request No. 33, dated July 31, 1985. 2. GPU Nuclear letter 4410-85-L-0110, Technical Specifications Change Request No. 49 and Recovery Operations Plan Change Request No. 31, dated June 18, 1985. 4 3. GPU Nuclear letter 4410-85-L-0231, Technical Specifications Change Request No. 49 and Recovery Operations Plan Change Request No. 31 - Response to NRC Comments, dated November 20, 1985. 4. US NRC NUREG/CR-1659, Reactor Safety Study Methodology Applications Program: Oconee No. 3 PWR Power Plant, Revised May 1981. 5. US NRC NUREG-1032 DR, Evaluation of Station Blackout Accidents at Nuclear Power Plants, May 1985. 6. GPU Nuclear Report, Probability of Loss of Off-site Power at TMI-2, Risk Assessment Section, Licensing and Nuclear Safety Department, Revision 1, 4 4 January 1986. l 7. Burns and Roe Report, Study for Evaluation of the Control Room ~ Ventilation and Cleanup System for the Technical Specifications Reduction Program (Task BI-04), Revision 2, December 1984. i 8. US NRC NUREG-0772, Technical Basis for Estimating Fission Product Behavior During LWR Accidents, June 1981. t 9. GPU Nuclear letter 4410-85-L-0077, GPU Nuclear Corporate Seismic Design. Criteria, dated April 16, 1985. 3 i } .}}