ML20044B433
| ML20044B433 | |
| Person / Time | |
|---|---|
| Issue date: | 01/17/1991 |
| From: | Chandler L NRC OFFICE OF THE GENERAL COUNSEL (OGC) |
| To: | Gray J NRC COMMISSION (OCM) |
| Shared Package | |
| ML20042D089 | List:
|
| References | |
| FRN-57FR14514, REF-GTECI-B-56, REF-GTECI-EL, RULE-PR-50, TASK-B-56, TASK-OR AE06-1-061, AE6-1-61, NUDOCS 9101290303 | |
| Download: ML20044B433 (7) | |
Text
.. - -. -. -
Afo& -l DC'T~
u UNIT ED sT ATES t
l S
NUCLEAR REGULATORY COMMISSION 4f WASHINGTON, D. C. 20555
'%,,, g January 17, 1991 9
'I MEMORANDUM T0:
Joseph R. Gray, Legal Assistant C.
Office of Commissioner. Curtiss c
e" FROM:
Lawrence J. Chandler, Assistant General Counsel for Hearings and Enforcement Office of the General Counsel Stuart A. Treby, Assistant General Counsel for Rulemakug and Fuel Cycle Office of the General Counsel
SUBJECT:
REQUEST FOR INFORMATION ON STATION BLACK 0UT AND DIESEL GENERATOR RELIABILITY During a briefing of Commissioner Curtiss on December 18,1990, regarding Generic Issue (GI) B-56, Diesel Reliability, and 10 CFR 50.63, " Station Blackout", the Commissioner requested written background information on this issue. The Staff responded to this request in a question / answer formatted document directed to Dave Trimble, the Commissioner's technical assistant (attached to a January 4, 1
1991, Memorandum from Ashok Thadani, Director, Division of Systems Technology, NRR to James Blaha, Assistant for Operations, Office of the ED0) 0GC has prepared a separate response, which follows. Please contact us if you have any questions.
Q.l.
What does the Station Blackout (SB0) Rule require with respect to:
(1) meeting a specified diesel generator reliability value, (ii) monitoring diesel generator reliability, (iii) utilizing specific
" trigger values" as a basis for instituting corrective action, and (iv) reporting failures to meet either trigger levels or reliability goals to the NRC?
A.I.
Section 50.63(a) requires that nuclear power plants be able to withstand a station blackout for a "specified duration."
l This "specified duration" must be based upon four factors, one of which is "the reliability of the onsite emergency ac power sources." See 10 CFR 50.63(a)(ii).
Section 50.63(c) requires licensees to submit to the NRC. "a proposed station blackout duration to be used in determining compliance with paragraph (a) of this section, including a justification for the selection based upon the four ' factors identified in paragraph (a) of this section." However, Sect. ion 50.63 does NOT identify a specific reliability value (e.a., 95%
gWY j
-~
g, (6hh4803 '
j
-g y
k.
i 2
reliability) which must be achieved.
Thus, licensees are free to select the reliability value that they use in determining the "specified duration" of SB0 for each of their plants.2 Section 50.63 also does NOT contain any requirements for monitoring i
emergency ac source reliability, the use of " trigger values" as a basis' for implementing activities intended to assure compliance with the licensee-specified reliability value, or reporting any non-compliance with licensee-specified reliability values to the NRC.
i Q.2.
Can the NRC require licensees to comply with the provisions of Regulatory Guide (RG) 1.155 with respect to specified diesel generator reliability levels, use of " trigger values,"
and f
monitoring and reporting?
j A.2.
There are three potential ways in which the NRC could require licensees to comply with RG 1.155.
Each of these presents some potential problems.
First, under Section 50.63(c)(3), the NRC is to notify licensees regarding the adequacy of their station blackout duration, and the equipment modifications and procedures deemed necessary by the licensees to comply with Section 50.63(a). If the NRC has yet to make such notifications (OGC under' tands that such s
notifications have not been issued), the NRC could withhold positive notification until licensees have committed to implement RG 1.155.
One problem with this approach is that failure to obtain positive NRC notification is not a prerequisite to continued operation; licensees have no incentive per se to obtain a positive NRC notification.
Another problem is that licensees could challenge NRC's failure to issue such notifications on the basis that Section 50.63 does not require the actions specified in RG 1.155 and therefore that the NRC has no valid basis for withholding notification. Moreover, a backfit claim under 10 CFR 50.109 could be raised.
Second, the NRC could issue a rule (which could be in form.of a modification to Section 50.63) which requires licensees to comply with RG 1.155.
A backfit analysis would have to be performed for such a rule, since the SB0 Rule did not require implementation of RG 1.155.
There would also be some delay in order to comply with the notice and comment provisions of the Administration Procedure Act.
Nonetheless, this seems to be the best way to achieve i
compliance with RG 1.155.
2 It should be noted that the SB0 Rule does not itself require the i
licensee-selected reliability value to remain fixed.
The licensee may change this value without notice to the NRC.
Of course, if the reliability value is changed, the licensee must ensure that the "specified duration" of the SB0 is not altered.
i
\\
3 The final alternative is to issue an order to each licensee (or possibly issue a " generic order" to licensees) pursuant to 10 CFR 2.204, requiring implementation of RG 1.155. As with the case of a rule, a backfit analysis would be required before the orders could 1
be issued. Moreover, each order could be challenged and a hearing requested by each licensee.
Q.3.
Are there other Commission requirements that would require licensees to monitor and report to the NRC failure to meet reliability or trigger values?
l L
l A.3.
In OGC's view, the followin'g Commission requirements could arguably impose reporting obligations on licensees with respect to emergency ac power source reliability values: 10 CFR 50.71(e) together with 10 CFR 50.59; the parallel reporting requirements in 10 CFR 50.72 and 50.73; and 10 CFR 50.9.
l 10 CFR 50.71(e) requires periodic updating of a. nuclear power plant's FSAR "to assure that the information included in the FSAR is the latest material developed." If the analyses required by the l
580 Rule and the licensee-selected reliability values must be incorporated into the FSAR by 50.71(e), they may not be changed except through 50.59 or by license amendment.
This would impose a " notification" requirement if the licensee found that it had to request a license amendment.
However, it is not clear whether i
Section 50.71(e) requires incorporation of the SB0 Rule analyses into the FSAR. Under Section 50.71(c) the FSAR update must include:
all the changes necessary to reflect information and analyses submitted to the Commission by the licensee or prepared by the licensee pursuant to Commission requirement...The updated FSAR shall be revised to include the effects of: all changes made in the facility or procedures as described in the FSAR; all safety evaluations performed by the licensee either in support of requested licensee amendments or in support of conclusions that changes did not involve an unreviewed safety question; and all analyses of new safety issues performed by or on behalf of the licensee at Commission request.
10 CFR 50.71(e) (emphasis added).
It is unclear whether 50.71(c) only requires that those aspects of the FSAR which are affected by new analyses (e.a., SSC and procedure changes which were implemented because of the results of new analyses) must be updated, or whether all new analyses required by the NRC must be incorporated into the FSAR.
The statement of considerations for the proposed and final version of Section 50.71 are ambiguous with respect to this matter.
See 41 Fed. Rea. 49123 (November 8,1976)(proposed rule), 45 Fed.
Rea. 30614 (May 9,1980)(final rule).
If the NRC were to now adopt the position that Section 50.71(e) requires the FSAR to be amended to include the analyses underlying the licensee's SB0 duration, licensees may well raise a backfit claim. The backfit claim would
4 likely be premised on the argument that the Staff has not previously indicated that the updated FSAR should include the SB0 analyses.
In addition, the NRC Staff's understanding of the requirements of 50.71(e) as applied to other analyses, e.o., evacuation estimates in the context of emergency preparedness, may not be consistent with the position that SB0 analyses must be included in the updated FSAR.
Assuming that 50.71(e) could be interpreted to require incorporation of the SB0 analyses into the FSAR and thereby imposing a
" notification" requirement, licensees would be required to comply with their selected reliability value and could not deviate from it or change the value unless they either made a 50.59 determination, or were granted a license amendment by the.NRC.
Incorporation of the SB0 analyses into the FSAR would not, nowever, be a basis for implying a monitoring requirement. Conceptually,:a licensee could be required to report important new information (i.e., a notification requirement), but have no obligation to positively seek out such information (i.e., a monitoring requirement).
The parallel provisions of 10 CFR 50.72(b)(ii)(A) through -(C) and 10 CFR 50.73(a)(2)(1)(A) through (C) may also afford the basis for a reporting requirement.
Under these sections, licensees are required to report to the NRC:
Any event or condition during operation that results in the condition of the nuclear power plant, including its principal safety barriers, being seriously degraded; or results in the nuclear power plant being:
(A)
In an unanalyzed condition that significantly compromises plant safety; (B)
In a condition that is outside the design basis of the plant; or (C)
In a condition not covered by the plant's operating and emergency procedures.
One could argue that failure to meet the licensee-selected reliability value is an event or condition that " seriously degrades" safety. In 0GC's view, this argument would be difficult to sustain in light of the promulgation of the SB0 Rule as a "stfety enhancement."
A case could be made that the failure to meet a licensee-selected reliability value places the. plant in an "unanalyzed condition that significantly compromises plant safety" which is reportable under subparagraph (A).
The term, "significantly compromises" arguably connotes a somewhat reduced safety importance as compared to the term, " seriously degrades." In addition, if the SB0 Rule was issued as a " safety enhancement" which involves a " substantial increase" in the overall protection of the public health and safety under 10 CfR 50.109(a)(3), then there should be no problem showing that
J.
0 5
failure to comply with the SB0 is a situation which "significantly compromises plant safety."
llowever, a problem in - relying upon subparagraph (A) is whether failure to meet the selected reliability value is a condition which is "unanalyzed."
The Statement of Consideration for the change to Section 50.73 points-out that
" licensees may use engineering judgment and experience to determine whether or not an unanalyzed condition existed. It is_ not intended that this paragraph apply to minor variations in individual parameters, or to problems concernino sinale pieces of eouioment (emphasis added)." 48 f_q6 Ben. 33850 (July 26,1983).
It is difficult to determine whether the failure to meet a licensee-selected reliability value which alters the "specified duration" of the plant's SB0 and renders invalid the plant's coping analysis, constitutes a " condition which is outside the design basis of the plant" which must be reported under subparagraph (B). The definition of " design basis" in subparagraph (8) is not clear. The statement of considerations for the changes to Sections 50.72 and 50.73 do not explain the meaning of design basis in this context.
Set 48 Fed.
Rea. 39039 (August 29,1983) (Section 50.72), 48 fad. Rea. 33850
(' July 26,1983) (Section 50.73). If " condition...outside the design basis" is interpreted as noncompliance with a NRC-required analysis of a " design basis event" as that term is defined in 10 CFR 50.49, z
the Environmental Qualification Rule, then licensees would - be required to report to the NRC their noncompliance with the licensee-selected reliability value.
By contrast, if " condition...outside the design basis" means non-compliance with a NRC-required analysis of a " design basis accident *,
then failure to meet the licensee-specified reliability value may
.not be reportable.
In the definition of " Alternate ac source" in 10 CFR 50.2, SB0 is described as a "non-design basis accident:"
" Alternate ac source" means an alternating current (ac) power source that... meets the following requirements:...(4) Has sufficient capability and reliability for operation cf all systems for coping with station blackout and for the time required to bring and maintain the plant in safe shutdown (non-design basis accident).
Another argument against reliance on subparagraph (B) is that a plant's design basis may be limited to those analyses that are necessary to demonstrate reasonable assurance (or no undue risk) to the public, and do not include analyses demonstrating compliance 10 CFR 50.49 refers to SSCs "important to safety" as being those required 2
)
to " remain functional during desian basis events (emphasis added)...."
Design basis events are in turn identified as " conditions of normal operation, including i
anticipated operational occurrences, design basis accidents, external events, and natural phenomena for which the plant must be designed to ensure [certain specified] function ~s...." See 10 CFR 50.49(b)(1).
l 6
with requirements deemed to be safety enhancements.
Pursuant to l
50.34(a)(3)(ii), the FSAR is required to set forth the " design bases and the relation of the design bases to the principal engineering criteria."
Under Section 50.34(a)(3)(1), the principal design criteria-are the General Design Criteria (GDCs) in 10 CFR Part 50, Appendix A.
Appendix A defines the GDCs as "the necessary design, fabrication, construction, testing and perfomance requirements for structures. systems and comDonents imDortant to safety; that is, structures, systems and components that provide _ reasonable assurance that the facility can be operated without undue risk to the health 1
and safety of the oublic." The association between design bases and reasonable assurance is further reinforced by a November 20, 1981 Memorandum from Harold R. Denton, Director, NRR to all NRR personnel which sets forth definitions for commonly-used safety classification terms which equates SSCs important to safety as those necessary to i
provide reasonable assurance. Since reasonable assurance of no undue risk is equivalent to adequate protection, then _it would be difficult to treat the SB0 Rule, which was characterized as a " safety enhancement," as part of the design bases which is. necessary to assure reasonable assurance.
Failure of a plant's diesel generator to meet its assumed reliability rating, thereby causing the plant's "specified duration" to change from that previously determined, may be reportable under subparagraph (C) as a S" condition not covered by the plam's operating - and emergency procedures." However, some plants' operating procedures may well cover the circumstance of loss of emergency ac power, which.
is equivalent to a reduction of the reliability of the diesel generator below its specified reliability value.
Some Staff consideration of the technical implications of a change in the specified duration and the impact upon plant procedures is necessary to fully explore this alternative.
Turning to 10 CFR 50.9, it is unclear whether that section imposes an independent obligation on licensees to monitor failures to meet either their selected reliability values or any set of "tr.igger l evel s. "
As discussed above, Section 50.63(c) explicitly requires only a one-time submission of information to the NRC.
Section 50.9(a) does not transform that one-time obligation into a continuing i
obligation to keep the NRC apprised of any changes in that information.
By its terms, Section 50.9 states that "information provided to the Commission...or information required by statute or by the Commission's regulations, orders or license conditions to be maintained" must be complete and accurate in all material respects.
Although 50.9(a) was not intended to independently transform one-time information reporting requirements into continuing requirements to maintain the accuracy of information and report changes to the NRC, the Statement of Considerations accompanying Section 50.9 suggests that there is an obligation to correct information that, while accurate when submitted, later turns out to be erroneous because of newly discovered information or advances in technology.
liowever, even if the reliability of the onsite emergency power
= - - -
^
)
7 sources should change, if this change does not affect the specified duration, the change in reliability arguably may not be material.
Secticn 50.9(b) may possibly be relied upon as a basis for requiring licensees who identify failures to comply with their selected reliability value to report such failures to the fiRC.
- liowever, Section 50.9(b) states that the licensees' reporting obligation extends only to "information identified by the... licensee as having for the regulated activity a sionificant implication for public health and safety or common defense and security (emphasis added)."
Licensees may well argue that failure to comply with selected reliability values is not reportable since it is not "significant" from a health and safety standpoint inasmuch as the SB0 Rule was adopted as a " safety enhancement."
The reasonableness of the licensee's belief in this regard does not appear to be relevant, as the Statement of Considerations for the. final rule, 52 Fed. Reg. 49362 (December 31,1987) indicates that the standard of materiality is "one of a licensee's own recognition of information with a significant health or safety or common defense or security implications."
Q.4.
Can an " interpretive rule" under Section 4 of the Administrative Procedure Act (APA), 5 U.S.C. 553(b)(A) be issued which would require licensees to: (1) meet either 14RC-specified or licensee-selected diesel generator reliability goals, (ii) monitor diesel generator reliability, (iii) utilize reliability " trigger values" as a basis for implementing corrective action, and (iv) report failures to meet either trigger levels or reliability goals to 'the tiRC?
A.4.
Yes, an interpretive r6e could be issued. However, if such r rule is to be binding and have the force of law, then it must bc usued under the notice and comment provisions of the APA.
See Union of Concerned Scientists v. tiRC, 711 F.2d 370 (D.C. Cir.1983), c.f.
Limerick Ecoloav Action. Inc. v. IIRC, 869 F.2d 719 (0.C. Cir.1989).
YA d
Lawrence 4. Chandler Assistant General Counsel for Hearings & Enforcement Office of the General Counsel
/
huar;tA.Tr/$/.Mb
- w5' eby Assistant General Cou sel for Rulemaking & Fuel Cycle l
Office of the General Counsel l
cc:
l M. Karman l
S. Crockett l
K. Cyr
.