ML20028A677
| ML20028A677 | |
| Person / Time | |
|---|---|
| Site: | Rancho Seco |
| Issue date: | 11/23/1982 |
| From: | Sheomaker C, Shoemaker C NRC ATOMIC SAFETY & LICENSING APPEAL PANEL (ASLAP) |
| To: | NRC OFFICE OF THE EXECUTIVE LEGAL DIRECTOR (OELD), SACRAMENTO MUNICIPAL UTILITY DISTRICT |
| References | |
| ALAB-703, ISSUANCES-SP, NUDOCS 8211240219 | |
| Download: ML20028A677 (16) | |
Text
..
./
00f.KETED LRNRC UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION
'82 Nm/ 23 P3:00 ATOMIC SAFETY AND LICENSING APPEAL BOARD
??' E CF ECETAW i U L SERV! E Administrative Judges:
i w.c4 Alan S.
Rosenthal, Chairman stavEuNOV231982 Dr. John H. Buck Christine N.
Kohl
)
In the Matter of
)
)
SACRAMENTO MUNICIPAL UTILITY DISTRICT)
Docket No. 50-312 SP
)
(Rancho Seco Nuclear Generating
)
Station)
)
)
MEMORANDUM AND ORDER November 23, 1982 (ALAB-703)
This special proceeding was instituted to consider whether certain actions ordered by the Commission in the wake of the March 1979 accident at Three Mile Island were necessary and sufficient to assure that the Rancho Seco facility could safely respond to feedwater transients.
The Licensing Board essentially answered that question in the affirmative (see LBP-81-12, 13 NRC 557 (1981)), and no appeals from its initial decision were taken.
Following our sua sponte review of the Licensing Board's initial decision, we issued a memorandum and order in which we withheld our final conclusions about the case, pending receipt of further information in several identified 8211240219 821123 PDR ADOCK 05000312 G
. -, =. - -
l 1
-l l
2 l
,m areas.
See ALAB-655, 14 NRC 799 (1981).
1/
We had hoped to receive this information and complete our review quickly.
Several factors, however, contributed to delay.
Principal among these was the shutdown of the Rancho Seco facility in April 1982 during which cracking in a high pressure injection (HPI) nozzle assembly was discovered.
Because one of the matters on which we sought more information concerned the number of thermal cycles that the HPI nozzles can 1/
Specifically, we requested the following information:
1.
Status reports from [ licensee] SMUD and the staff on the six recommendations in BAW-1564 to enhance AFW [ auxiliary feedwater) safety and reliability; 2.
Status reports from SMUD and the staff on SMUD's commitments to improve AFW reliability, as described in CEC Exhibit 21 (Enclosure 2);
3.
Status reports from SMUD and the staff on the installation of the safety-grade anticipatory reactor trip; 4.
Status reports from the staff and SMUD on the need for the additional analyses identified in the Staff Evaluation at 19, 23 [(see 14 NRC at 809)];
5.
Staff comments on the March 25, 1981, letter from B&W to SMUD concerning " Reactor Coolant Pump Suction Small Break LOCA";
6.
SMUD and staff schedules for HPI [high pressure injection] analyses; and 7.
Staff clarification of its position on the need vel non for extended pressurizer level indication.
14 NRC at 817.
3 withstand (id. at 810-11, 817), the discovery of the cracked nozzle and related problems resulted in our posing still further questions to both licensee and the NRC staff.
See Memorandum and Order of April 15, 1982 (unpublished).
Responses to those questions were submitted over the next five months.
In the meantime, the Rancho Seco facility returned to operation in mid-August, following various repairs to the HPI nozzles and other hardware.
2/
We have by now received the information originally solicited in ALAB-655 for all but one area.
Under item 2, the staff committed to review licensee's revised reliability analysis of proposed modifications to the auxiliary feed-water (AFW) system and to apprise us of its evaluation.
See Affidavit of Ernest D.
Sylvester (December 4, 1981) at 3, 4.
That review is apparently still under way, but we expect to receive the staff's evaluation soon. -3/
Of course, until
--7/
The staff monitored these repairs and authorized the return to operation after concluding in its safety evaluation that licensee's corrective actions were i
j acceptable.
We, as well, were satisfied with the nozzle repairs as described to us during an August 13, 1982, conference call with licensee and the staff.
See note 6, infra.
3/
Installation of the AFW system modifications that are
~-
the subject of the staff's review is not scheduled until mid-1984.
Letter from staff counsel to Appeal Board (October 15, 1982).
l l
4 we receive all the material solicited and are satisfied that the gaps we initially perceived in the record have been filled, we cannot reach any final judgment as to the overall adequacy of the Licensing Board's initial decision.
We are able at this juncture, however, to summarize our conclusions about the items not relating directly to the improvements proposed for the AFW system.
1.
In items 1, 3, and 4, we requested status reports from licensee and the staff on various recommendations or ccmmitments to pursue further action.
14 NRC at 805-06, 808-09, 817.
The concern underlying these requests for information was that matters that had assumed enough safety significance during the hearing to provoke licensee's commitment to further consideration might later be overlooked, inadvertently or otherwise.
The status reports received from licensee and the staff, however, reflect that adequate attention has been devoted to each item identified in our requests.
Further, licensee and the staff are in general agreement as to which additional actions may still be warranted and which are not.
No further comment or involvement on our part with respect to items 1, 3, and 4 appears to be necessary.
2.
In ALAB-655, we discussed a letter from Babcock and Wilcox (B&W) to licensee concerning the fact that the loss-of-coolant accident (LOCA) analyses relied on in this proceeding did not consider a pump suction line break where
5 AFW flow is delayed.
Id. at 809-10.
1/
Instead, the analyses considered a pump discharge line break and demonstrated that operator actions to start either the AFW flow or the HPI flow within 20 minutes will result in acceptable conditions.
The B&W letter noted that it had not been shown whether this 20-minute delay in AFW actuation was acceptable to accommodate the greater rate of coolant loss associated with a pump suction line break.
In view of this apparent deficiency in the LOCA analyses, we solicited comments on whether further analyses were necessary before the safety of the Rancho Seco facility can be reasonably assured.
The staff stated in response that it did not regard the failure of the LOCA analyses to include the " pump suction break / delayed AFW" scenario as significant to the continued safe operation of the plant.
It concluded that
" demonstration of the 20 minutes for operator action is not an absolute requirement since the subject scenarios are outside the design basis for Rancho Seco."
Affidavit of Walton L.
Jensen, Jr., on Item No. 5 (November 24, 1981) at 3.
Because we found the staff's comments not fully responsive and we had become aware of another affidavit on 41 The March 25, 1981, letter was not part of the record
~~
below but was provided to us by licensee's counsel following issuance of the initial decision.
6 the same matter recently provided to the Licensing Board in the TMI-1 Restart proceeding, we sought clarification of the staff's position.
Order of January 28, 1982 (unpublished).
The staff reaffirmed its view that additional analyses are unnecessary.
It based its conclusion on four factors: (a) regardless of the break location, the vessel water level would initially drop to the same approximate elevation; (b) the added loss of primary coolant inventory from a pump suction pipe break would be from water in the cold leg; (c) in the absence of emergency feedwater, the operator has 20 minutes to initiate IIPI, regardless of the location of the break in the cold Icg piping; and (d) emergency procedures require the operator to initiate HPI immediately, regardless of break location, if a loss of all feedwater has occurred.
Affidavit of Walton L. Jensen, Jr. (February 5, 1982), at 2.
Although we still find the staff's responsc somewhat unclear, we agree that analysis of a pump suction line brev is not necessary.
Admittedly there has been no demonstra-tion that, in fact, the 20-minute period for initiation of AFW flow found acceptable for a pump discharge line break is also acceptable for a suction line break.
The time factor, however, assumes less importance in view of post-TMI emergency procedures that now direct the operator to activate HPI immediately.
Thus, even if further analysis were to show that substantially less time is available to restore feedwater flow following a pump suction line break,
7 immediate actuation of HPI will assure acceptable conditions.
3.
A matter that warranted our attention in ALAB-655 was the effect of thermal stress on high pressure injection nozzles and the number of HPI initiation cycles permitted for each nozzle at Rancho Seco.
The record and decision below showed that the number of design basis cycles (40) might soon be reached.
While we concluded that the Licensing Board's characterization of the design basis limit as "' overly conservative'" might well be justified, we sought supplementation of the record on this point.
Specifically, we asked the staff and licensee to provide
" analyses of (1) the maximum allowable number of thermal cycles on the HPI nozzles; (2) methods of detecting thermal cycle effects on the nozzles; (3) possible means of prolonging the useful life of the nozzlec; and (4) technical specifications or operating procedures that might reduce the use of the HPI without endangering the core."
14 NRC at 810-11.
In responding to our request, licensee stated that it had reevaluated and increased the design basis limit for HPI cycles and, consequently, that limit is not being approached more quickly than anticipated. -5/
It also identified a
--S/
At the outset, we solicited only licensee's and the staff's schedules for performing additional HPI analyses.
14 NRC at 817.
Licensee, however, responded with information addressed directly to the four areas of our concern.
s 8
change in operating procedure designed to reduce the thermal stress on the nozzles.
Affidavit of Robert A. Dieterich (December 11, 1981) at 6 6-4.
After receiving licensee's submittal, the staff replied that it would need more time to review it, but indicated that it considered plant operation safe over the near term.
Affidavit of Mark L.
Padovan (January 5, 1982) at 2.
The staff's review led to its own requests for further information from licensee concerning several aspects of the latter's HPI analyses.
The staff and licensee thus traded information over the next few months.
In the meantime, cracking in the HPI makeup nozzles was discovered at the Crystal River and Oconee facilities --
like Rancho Seco, both B&W plants.
Thermal cycling was considered as the possible cause of the cracking.
Our concern about the implications of these events for Rancho Seco precipitated our request of the staff for still more information related to thermal stress on HPI nozzles.
Memorandum and Order of March 8, 1982 (unpublished).
After meeting with the staff, licensee agreed to shut down Rancho Seco for inspection of the nozzles.
Ten days after the shutdown, the staff notified us that cracking was found in the makeup nozzle and that its thermal sleeve was missing.
The staff speculated that the sleeve could have traveled through the reactor coolant system to the bottom of the
9 reactor vessel, where it might remain trapped.
No cracking or missing sleeves were discovered with respect to the three other nozzles, although the sleeve in nozzle B had moved about one inch upstream, apparently as a result of missing or loose weld buttons.
Board Notification BN-82-37 (April 13, 1982).
Again, we sought more information from the staff and licensee -- this time on the consequences of the movement of the thermal sleeve through the reactor system and the nature of the repairs.
Memorandum and Order of April 15, 1982 (unpublished).
And again, the staff determined that it would need more information from licensee before it could respond to our questions or conclude that the plant could be safely restarted.
Affidavit of John F.
Stolz (April 21, 1982) at 2.
Licensee complied with the staff's request for further analyses, completed its repairs, and returned the facility to operation in mid-August after receiving staff authorization.
5I The staff later provided us with its
_6/
The plant had remained shut down for approximately four months, during which time the thermal sleeves for the makeup nozzle and nozzle B were replaced and unrelated problems concerning excessive hydrogen in the primary coolant system and deformation of the steam generator internal AFW header were resolved.
See Board Notification BN-82-41 (April 30, 1982).
Our April 15 order had directed licensee to notify us at least three business days before the scheduled return of Rancho (FOOTNCTE CONTINUED ON NEXT PAGE) l l
w 10 final analysis of the maximum allowable number of thermal cycles on the HPI nozzles, concurring with licensee's results.
Affidavit of Shou-Nien Hou (September 2, 1982).
Now that we have the results of the staff's and licensee's analyses and the nozzle cracking problem appears to be remedied, we can summarize our findings based on this supplementation of the record.
As a result of a more analytical evaluation than appeared in the record below (see, e.a., Tr. 2014-15), licensee and the staff agree that the design basis for Rancho Seco includes 70 allowable cycles on each nozzle due to manual HPI initiation; 40 rapid depressurization cycles (automatic HPI); 40 test cycles; and 240 heatup and cooldown cycles.
Affidavit of Robert A.
Dieterich (December 11, 1981) at 6-4; Affidavit of Shou-Nien Hou (September 2, 1982) at 2; letter from licensee's counsel
, 6_/
(FOOTNOTE CONTINUED FROM PREVIOUS PAGE)
Seco to operation.
Because we had not yet received the staff's comments on the adequacy of the nozzle repairs at the time licenses notified us of Rancho Seco's imminent restart, we held a conference call with licensee and the staff on August 13, 1982.
During the call we discussed, among other things, the thermal sleeve and AFW header repairs and the status of the staff's review of licensee's HPI analysis.
Pursuant to our request, both licensee and the staff followed up with letters incorporating the salient points of the conference call.
The staff also submitted a safety evaluation report in which it concluded that licensee's corrective actions for the thermal sleeve problem were acceptable.
O 11 to Appeal Board (July 8, 1982), Enclosure (" Calculation Data / Transmittal Sheet").
According to the staff, "[a]
simplified ratio method was utilized to extrapolate stresses calculated for the rapid depressurization transients as shown in the original stress reports, which was based on the nuclear power piping [ASME] Code B31.7, 1968 draft."
Affidavit of Shou-Nien Hou (September 2, 1982) at 2.
Despite some earlier misgivings about licensee's calculations (see Affidavit of John F.
Stolz (February 25, 1982) at 2), the staff now finds this to be a valid method of estimating stress and has determined that the load combinations used are acceptable.
Affidavit of Shou-Nien Hou (September 2, 1982) at 2-3.
The actual number of thermal cycles experienced by each nozzle as of April 1982 is: nozzle A (makeup), 19; nozzle B, 33; nozzle C, 30; and nozzle D, 30.
Affidavit of Mark L.
Padovan (April 16, 1982) at 2-3.
1I Since the hearing, licensee has changed its cperating procedures in order to limit thermal stress.
For manual post-trip coolant system volume control, licensee now requires operators to use only the HPI nozzle that is used for system makeup.
- Thus, because the flow is continuous, the nozzles are not subject
_1/
If licensee's repairs (see note 6, supra) included replacement of nozzles A and B themselves (rather than just the thermal sleeves), the number of thermal cycles experienced by those two nozzles presumably can now be considered zero.
I l
O 12 to thermal stress upon manual HPI initiation after a reactor trip.
Affidavit of Robert A. Dieterich (December 11, 1981) at 6-3.
In view of this change in operating procedure and the number of cycles already experienced relative to the reevaluated design basis limits, it no longer appears that "there is a substantial chance that the permitted lifetime number of HPI cycles for each nozzle will soon be reached."
14 NRC at 810.
We are therefore satisfied that even if there is some increase in HPI actuations due to the modifications originally ordered by the Commission in this proceeding, it is unlikely to result in diminished effectiveness of the HPI nozzles. -
Although our original concern has thus been assuaged, the discovery of cracking in the makeup nozzle and its apparent relationship to a thermal sleeve missing from the same nozzle provide a new object for our attention.
The staff han concluded that the design modifications and replacement of the thermal sleeves in two of the four nozzles are acceptable corrective actions, and we have no cause to doubt that assessment.
See Affidavit of Mark L.
Padovan (August 17, 1982) at 2 and Enclosure (Safety
_8/
Despite the agreement of the staff and licensee on the reevaluated design basis number of allowable HPI cycles, it is not clear whether the additional 70 cycles for manual HPI initiation have been incorporated in documents pertinent to the Rancho Seco operating license (e. g., the Final Safety Analysis Report).
If not, licensee should take steps to modify these documents to reflect the new evaluation.
C' i
13 Evaluation Report).
S!
Our concern, however, lies with the adequacy of licensee's inspection program vis-a-vis the remaining two nozzles (C and D) for which no design changes were made.
According to the staff, ultrasonic and liquid penetrant inspections of HPI nozzles to safe-end welds are required only once every ten years.
Affidavit of Mark L.
IEI Padovan (March 31, 1982) at 3.
More frequent inspec-tions of the remaining two original nozzles (already in service six years) might detect missing weld buttons and a loosened sleeve before it has the opportunity to travel through the system -- as the sleeve from the makeup nozzle has done already.
Licensee has agreed to perform an additional radiographic examination of these two nozzles at the next refueling outage.
Letter from licensee's counsel to Appeal Board (August 16, 1982) at 2.
We have tentatively concluded, however, that a radiographic inspection of these nozzles should be performed at each refueling cutage in the future, until they have been replaced or modified in the 9/
The thermal sleeve from the makeup nozzle, thought to
~~
be lying at the bottom of the reactor vessel, is to be removed at the next refueling outage (January 1983).
Affidavit of Mark L.
Padovan (August 17, 1982) at 2.
10/
Until the discovery of cracked nozzles at Oconee and Crystal River focused attention on the matter, no such inspection had been performed yet at Rancho Seco.
Affidavit of Mark L.
Padovan (March 31, 1982) at 5.
I
I s,
14 same manner as nozzles A and B.
We recognize that nozzles C and D are not used for continuous system makeup and that they showed no signs of degradation when inspected during the plant's most recent shutdown.
Affidavit of Robert A.
Dieterich (April 21, 1982) at 4.
But we do not regard radiographic examinations at each refueling outage as a significant burden on licensee, given the problems associated with the two other nozzles of the same original design. 11/
We will give licensee the opportunity, how-ever, to explain why a license condition incorporating these additional examinations in its inservice inspection program is not warranted.
Licensee's comments should be filed with us by December 14, 1982.
The staff may reply by December 29, 1982.
4.
In its initial decision, the Licensing Board agreed with what it perceived as a staff recommendation for steps to prevent the loss of pressurizer level indication.
13 NRC at 584-85.
It thus " direct [ed] the licensee and 1
Staff to proceed directly with plans for" such instrumen-
~~11/
Moreover, licensee's calculations reevaluating the design basis number of allowable thermal cycles for each nozzle appear to assume the presence of a thermal sleeve in each nozzle.
See letter from licensee's counsel to Appeal Board (July 8, 1982), Enclosure
(" Calculation Data / Transmittal Sheet").
If the presence of an intact sleeve cannot be assured, the reliability of the design basis as reevaluated might be seriously undermined.
f 15 tation.
Id. at 586.
Because we were uncertain as to the nature of both the staff's position and the Board's
" direction" to the parties, we asked the staff to clarify whether it believes extended pressurizer level indication is needed at Rancho Seco.
14 NRC at 814-15.
The staff has replied with an affidavit that is still somewhat confusing.
It nonetheless unequivocally concludes that extension of the existing pressurizer level indication range is not necessary.
Affidavit of Walton L. Jensen, Jr.,
on Item No. 7 (November 24, 1981) at 3.
The staff apparently believes extended pressurizer level indication is not necessary because post-TMI guidelines instruct the operator to rely on a subcooling meter in the control room i
to monitor primary system inventory.
Further, the long-term modifications of the main and auxiliary feedwater system proposed by licensee are designed to keep the pressurizer level on scale after a reactor trip, thus obviating extended level indication.
Id. at 2-3.
In view of the staff's clarification of its position and licensee's proposed feedwater modifications, we see no need to formalize the Licensing Board's instruction "to proceed directly with plans for extended pressurizer level indication."
13 NRC at 586.
w 16 Licensee's comments on the need for additional radiographic inspections of the two unmodified HPI nozzles are due December 14, 1982.
The staff's reply is due December 29, 1982.
Our final ruling in this proceeding is deferred pending receipt of the comments noted above and the staff's evaluation of licensee's revised reliability analysis of proposed modifications to the AFW system.
It is so ORDERED.
FOR THE APPEAL BOARD G..G u W Secre$n Shoemaker C.
J tary to the Appeal Board