ML20039D678
| ML20039D678 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 12/17/1981 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20039D677 | List: |
| References | |
| TASK-2.E.4.2, TASK-TM NUDOCS 8201050436 | |
| Download: ML20039D678 (7) | |
Text
<1
.j-
- pa trog'o UNITED STATES 83 T q[gg NUCLEAR REGULATORY COMMISSION ENCLOSURE.2 i
1; f a C WASHINGTON, D. C. 20555
^
l o,[':
j-
,r gvf SAFETY EVALUATION bV THE OFFICE OF NUCLEAR REACTOR REGULATION SUPPORTING CONTAINMENT VENTING AND PURGING DURING NORMAL OPERATION f
_0 F_
BROWNS FERRY NUCLEAR PLANT, UNITS NOS.1, 2 AND 3 TENNESSEE VALLEY AUTHORITY DOCKET NOS. 50-259, 50-260 AND 50-296 1.0 Introduction A number of events have occurred over the past several years which directly relate to the practice of containment purging and venting during normal plant operation. These events have raised concerns relative to potential failures affecting the purge penetrations which-could lead to degradation.in containment integrity. and, for PWRs, a degradation in ECCS performance. By letter. dated November 29, 1978, the Commission (NRC) requested the Tennessee' Valley Authority (TVA/ licensee) to respond to certain generic concerns about containment purging or venting during normal plant operation. The generic concerns were twofold:
(1) Events had occurred where licensees overrode'or' bypassed. the safety actuation isolation signals to the containment isolation valves.
These override features were found to be desioned such that a legitimate automatic closure would not have occurred if needed.
These events were determined to be abnormal occurrences and were so characterized 'in our report to Congress in January 1979.
L.
(2) Recent licensing reviews have required tests or analyses to show 1
that containment purge or vent valves would shut without degrading containment integrity during the dynamic loads of a design basis loss of coolant-accident (DBA-LOCA).
The NRC position of the November 1978 letter requested licensees to cease purging (or venting) of containment or limit purging (or venting) to an absolute minimum. Licensees who elected to purge (or vent) the J-containment were requested to demonstrate that the containment purge (or vent) system design met the criteria outlined in the NRC Standard i-Review Plan (SRP) 6.2.4, Revision 1, and the associated Branch Technical
- t Position (BTP) CSB 6-4, Revision'l.
'i 4
~
KOohr 1
m
. As a consequence of the. incident at THI-2, implementation of a number of new requirements has been recommended for operating reactors. These new requirements are described in NUREG-0660, "NRC Action Plan Developed as a Result of the TMI-2 Accident," May 1980, and NUREG-0737, " Clarification of Ti1I. Action Plan Requirements," November 1980. The NRC staff has also requested licensees to submit information sufficient to permit an independ -
ent evaluation of their response to these new requirements.
One of the TMI-2 Action Plan requirements directly related to containment venting and purging is action Item II.E.4.2 on containment isolation dependability. Task II.E.4.2 sets forth 7 positions on requirements to insure containment isolation. All licensees, including TVA, were required to be in conformance with positions 1 through 4 by January 1,1980 as part of the Catogory "A" Lessons Learned requirements. Positions 5 through 7 are being evaluated 'as part of the generic issue on containment vent and purging.
2.0 Discussion Purging and inerting at Browns Ferry Nuclear Plant (Browns Fer"y) are performed through isolation valves of various sizes, the largest of which are 10,18, and 20-inch butterfly-type valves. The exhaust lines from the drywell and torus are connected to a common exhaust header which leads to the Reactor Zone Exhaust System and th; Standby Gas Treatment (SBGT) Systen.
Two 18-inch butterfly isolation valves configured in series (one of each is bypassed by a 2-inch line containing 3n isolation valve) are provided to isolate the exhaust line from the drywell. Two 18-inch butterfly isolation valves configured in series (each is bypassed by a 2-inch line containing an isolation valve) are provided to isolate the exhaust line from the torus. The purge supply lines to the drywell and torus are connected to a common supply header. One 20-inch butterfly isolation valve is provided for the conmon header. One 18-inch butterfly isolation valve (which is bypassed by a 2-inch line having an isolation valve) is provided to isolate the supply line to tne drywell. One 20-inch butterfly isolation valve is provided to isolate the nitrogen supply to the supply header.
The licensee responded to the NRC position letter of November 1978, by their letters of December 21,1978, March 1,1979 and June 12,1979 in which they reported that they had been conducting a program of limited and intermittent purge and inerting and that their present Technicil Specifications for each unit allow use of primary containment purge systems 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> before unit shutdown and during the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following unit startup. The licensee indicated that, on the average, about five entries are made into the containment each year for maintenance and that purging and de-inerting operations are necessary for personnel safety.
The total combined purge time for all three units for 1977 was 634 hours0.00734 days <br />0.176 hours <br />0.00105 weeks <br />2.41237e-4 months <br />,
- j and for 1978 was 665 hour0.0077 days <br />0.185 hours <br />0.0011 weeks <br />2.530325e-4 months <br />. The licensee indicated that the expected purge operation time during a typical cycle for one unit is 296 hours0.00343 days <br />0.0822 hours <br />4.89418e-4 weeks <br />1.12628e-4 months <br /> per year.
We have reviewed the licensee justifications for the estimated annual usage of the purge system and find them to be acceptable.
i
, _r
. The licensee has conducted an analysis'to determine the amount of air.
and steam released in the event of a LOCA during purging operations.
The results were presented in TVA's letter to us of March 17, 1980.
The analysis performed was based on the design basis accident that produces the highest drywell and torus pressure. The two valve closure times, 5 and 15 seconds, bound the calculations for a spectrum of break sizes.
As stated in their letter of June 2,1981, TVA has or will (by fall 1982) reduce the valve closure time for the large purge valves from approximately 15 seconds to less tnan 2.5 seconds, which will reduce the offsite and onsite doses well below those analyzed and determined to be within NRC limits.
For larger breaks, contaimnent isolation will be initiated shortly into the incident.
For smaller breaks, containment pressurization rates are slow and isolation logic is tripped promptly by high radiation monitors.
The amount of air released for a large break LOCA is 955 lbs. and the amount of steam released is 2,903 lbs. In the event of a small break LOCA the amount of air released is 3,572 lbs. and the amount of steam released is 8,425 lbs. We have reviewed the licensee's assumptions and find them to be acceptable.
The licensee has provided the results of an analysis to determine the capability of provisions made to protect structures and safety-related
~ equipment located beyond the outboard purge isolation valves against a loss of function resulting from exposure to the environment created by the escaping air _and steam, during an accident.
In their letter of-March 17,1980, the licensee indicated that their initial analysis showed that the blowout panels on the refueling floor would open as a result of the pressure transient. The licensee indicated that the Standby Gas Treatment System would not be able to maintain sufficient negative pressure in the reactor building as a result.of the blowout panels relieving. As a result, the licensee stated in their letter dated March 17, 1980 that purge operations will be limited at hot conditions "until this situation is resolved."
TVA's letter of June 2,1981 presented the results of their reanalysis of containment purge valve operability under DBA LOCA forces. As noted above, TVA is modifying the 18" and 20" butterfly isolation valves to reduce the valve closure time to less than 2.5 seconds. (The analysis actually indicates that the disc will be fully closed under maximum torques in about 1.9 seconds.) The reduced valve closure time not only results in lower radiological doses but also protects against pressurization of appurtenant duct work and secondary containment. The reanalysis shows that, when modified, the containment vent and purge system at Browns Ferry will meet Branch Technical Position CSB 6-4, Item 5.b.
and is acceptable.
The licensee indicated (by letter dated March 17, 1980) that the maximum closure time for a purge valve at Browns Ferry was 15 seconds and committed to modify the valves to achieve a closure time of 5 seconds in
- l order to be in compliance with Item B.l.f of BTP CSB 6-4, Revision 1.
As noted above, the closure time on the large valves is being reduced to less than 2.5 seconds.
k e
l y
- -The licensee provided a sketch of a design for a debris screen to be added where each line penetrates the drywell or torus. The licensee states that the debris screens are designed to withstand the maximum LOCA pressure while completely clogged with debris without failure.
^
The design of the screen incorporates a 2 mesh,10 gauge stainless steel wire screen welded across a maximum opening of 6 x 6-inches created by a grid of A-36 (1-1/2" x 1/4") steel bar. The licensee has indicated that the debris screens are consistent with Seismic Category I requirements and that the installation of the screens will not adversely affect the seismic quali'fication of the piping.
We conclude that for Browns Ferry Units 1, 2 and 3, the only open issue with respect to containment vent and purging is: to incorporate in the Technical Specifications a surveillance requirement.to periodically test that the closure times for the larger butterfly valves, including instrument delays, is less than the 2.5 seconds used in TVA's evaluation.
In addition, as a result of numerous reports on the unsatisfactory performance of resilient seats in butterfly-type isolation valves due to seal deterioration, periodic leakage integrity tests of the 10,18,
and 20-inch butterfly isolation valves in tha purge system are necessary.
The purpose of the leakage integrity tests of the isolation valves in the containment purge lines is to identify excessive degradation of the resilient seats for these valves. Therefore, they need not be conducted with the precision required for the Type C isolation valve tests in 10 CFR Part 50, Appendix J.
These tests would be performed in addition to the quantitative type C tests required by Appendix J, and would not relieve the licensee of the responsibility to conform to the requirements of Appendix J.
We will pursue with the licensee the results of operating experience on seal deterf ation in the larger butterfly-type isolation valves at Browns Ferry and the possible need for additional surveillance requirements and/or leakage integrity tests on these valves.
Subject to successful' implementation of the above recommended actions, we find the purge / vent system design and operating practices for Browns Ferry to be acceptable. The environmental qualification of the valves and control systems is being evaluated as part of our review of licensee responses to IE Bulletin 79-018.
As discussed in the Introduction above, TMI-2 Action Plan Item II.E.4.2 is being evaluatsd as part of the generic issue on containment vent and purging. We have comoleted our review of the licensee's submittals on position 5 of II.E.4.2.
Position 5 requires that the containment pressure setpoint that initiates containment isolation for non-essential system containment vessel penetrations be at, or reduced -to, "...the minimum compatible with normal i
operating conditions." Response evaluation was based upon the values provided for the following parameters:
I
~
e
. (1) The maximum observed or expected containment pressure during normal operation.
(2) The loop error and observed drift in the pressure sensing instrumentation providing the isolation signal (see note).
(3) The containment isolation pressure setpoint.
NOTE: The clarification document (NUREG-0737) provided only the expected margin for instrument error and did not specify
~
acceptable values for instrument drift or atmospheric changes contributing to the total sensing loop error.
Additional staff guidance established a limit of 3.0 psi for an isolation setpoint margin over the normal containment pressure to account for total loop error.
In addition, for subatmospheric contaimnents, a 3.0 psi setpoint margin over atmospheric pressure is also considered acceptable.
In consideration of these values, the isolation pressure setpoint is to be as low as practical without increasing the probability of inadvertent activation of the isolation signal.
The letter of December 23, 1980, submitted by TVA, provided sufficient information to conclude that the containment isolation pressure setpoint for Browns Ferry Nuclear Plant, Units 1, 2 and 3, meets the NUREG-0660/0737 requirements, is within the additional limiting guidelines provided by the NRC staff and is acceptable.
As noted previously, we had concluded as part of our " Category A" Lessons Learned evaluations that TVA was in conformance with positions 1, 2, 3 and 4 of Task Item II.E.4.2.
This Safety Evaluation concludes that TVA is also in:conformance with position 5 of this task item.
Our review of the plant design with respect to positions 6 and 7 is continuing. The results of our evaluation will be presented in a future Safety Evaluation on this subject.
l i
T
?
I ENCLOSURE 3 4
SAFETY EVALUATION REPORT BROWNS FERRY NUCLEAR STATION, UNITS 1, 2 AND 3 OVERRIDE OF CONTAINMENT pVRGE ISOLATION AND 4
OTHER ENGINEERED SAFETY FEATURES ACTUATION SIGNALS
~
Introduction Instances have been reported at nuclear power plants where the intended automatic closure of containment purge / ventilation valves during a postulated accident would not have occurred because the safety actuation signals were inadvertently overridden and/or blocked, due to design deficiencies. These instances were determined to constitute an Abnornal Occurrence (!78-5). As a followup action, NRR issued a generic letter requesting each licensee to take certain actions.
Evaluat_ ion 1
The enclosed report (EGG-ll83-4163) was prepared for us by EG&G, San Ramon, as part of our technical assistance contractor program. The report provides their technical evaluation of the design compliance with NRC-provided criteria.
The report determines that, upon completion of agreed-to plant modifi-cations to preclude re-opening of certain valves upon reset, the design of the Primary Containment Isolation System (PCIS) satisfies the NRC-criteria for this review.
The report also identifies that the override features associated with i
post-accident venting of the containment to the Standby Gas Treatment System does not meet the criterion regarding the blocking of more than one type of safety actuation signal. This design feature for containment atmosphere control is another engineered safety feature that can function only after the reactor mode switch is taken out of the Run mode. This bypass is acceptable.
l The contractor's report recommends that plant modifications be undertaken l-to provide systems-level manual actuation features for the PCIS. The Browns Ferry plants were designed and licensed to IEEE Standard 279 of 1968. This standard required manual actuation features but did not require those features to be at the system-level. While system-level manual actuation features (as required by the later edition of IEEE 279) may be desirable, the design satisfies the standards in effect at the time of licensing and thus is acceptable.
- I The contractor's report also concludes that other ESF systems satisfy the criteria for this review.
t
. Conclusions Based upon our review of the contractor's technical evaluation, we conclude that completion of modifications by the licensee has brought the design of the PCIS into conformance to the criteria for this review. The design of other ESF systems with manual override capability conform to our criteria, e
e
, t.
ll