ML20199J317
| ML20199J317 | |
| Person / Time | |
|---|---|
| Issue date: | 12/17/1998 |
| From: | Burgess S, Mary Drouin, Guttmann J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III), NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | |
| Shared Package | |
| ML20199J249 | List: |
| References | |
| NUDOCS 9901260122 | |
| Download: ML20199J317 (85) | |
Text
-
Options for Incorporating Risk Insights into 10 CFR 50.59 Process December 17,1998 By Mary Drouin (RES)
Jack Guttmann(RES)
Sonia D. Burgess (Region III)
Michael E. Parker (Region III)
Daniel T. Moy (Region 1)
Contractor Support:
Allen Camp Donnie Whitehead (Sandia National Laboratories)
- n agg22 7g 228 PDR a
TABLE OF CONTENTS
. SECTION PAGE 1.
INTRODUCTION
.I 1
2.
BACKGROUND.................
2 2.1 Current {50.59 Regulation and Process..
2 j
2.2 Limitations of 10 CFR 50.59 Regulation and Process..
3 2.2.1 Summary of Limitations......
.3 2.2.2 Scope Lim itations....................................
4 2.2.3
" Parameter" Limitations.....................
4 2.3 Examples of Facility Changes Using the 10 CFR 50.59 Process 5
l 2.3.1 Facility Changes Needing NRC Approval........
6 2.3.1.1 Error in Dose Calculations for the Process Gas System 6
2.3.1.2 Error in Air Volume of Secondary Containment..
6 2.3.1.3 Calculational Error in Head Loss of Emergency Core Cooling Suction Strainers. '........
7 2.3.1.4 Modification of Main Feedwater Valves.
7 2.3.1.5 Discrepancy Between As-Built System and UFSAR Description.
8 2.3.1.6 Examples of Facility Changes That Are Not Pursued as License j
Amendment Because of the Expense, Uncertainties, or Timeliness of Amendment Approval 8
2.3.2 Facility Changes Not Needing NRC Approval.,
8 2.3.2.1 Residual Heat Removal Service Water Sump Pump Discharge Check Valves Replacement.
8 2.3.2.2 Manual Isolation of the Control Rod Drive Pumps during a Station Blackout or Loss Of Offsite Power Procedure Change.
9 2.3.2.3 RIIR Low Pressure Core Injection Inboard Injection Valve Design Change..
9 3.
APPROACli FOR IDENTIFYING OPTIONS..........
I1 4.
EVALUATION FACTORS 13 4.1 PRA Policy Statement Implications.
14 4.1.1 Enhanced Safety Decisions.......
14 4.1.2 More Efficient Use of Agency Resources.....
14 4.1.3 Reduction in Unnecessary Burden on Licensees.
15 4.2 Facility Coverage implications....
15 4.3 Risk Implications..........
15 4.3.1 Confidence Required From Risk Results...
15 4.3.2 PRA Completeness.
15 4.3.3 Deterministic vs Risk-informed..
16 4.4 Regulatory implications..
16 4.4.1 Regulatory Guidance Needed....
16 4.4.2 Impact on Other Regulations 16 4.4.3 New Rule Making........
17 4.4.4 Impact on Inspection / Training 17 l
1
-i-I l
i o
l l
l l
TABLE OF CONTENTS SECTION PAGE 4.5 Licensee implications..............
17 4.5.1 Internal Procedures Needed 17 4.5.2 Impact on PRA Activities...
17 4.5.3 Impact on Training..................
... 17 4.5.4 Impact on Inspection.................
18 4.5.5 Impact on Facility Changes..........
... 18 4.6 Resource implications..........................
18 4.7 Time Implications.........................
. 19 5.
Test Case Description 20 5.1 Description of Proposed Change...........
20 5.2 Evaluation Again:.t Q50.59 Parameters.........
22 6.
Options
. 25 6.1 Scope and Parameter Changes...............
... 25 6.2 Evaluation Process.........................
. 26 6.3 Scope Change Options...................
..... 28
6.3.1 Option
Limit the Scope to Chapter 15 Events.
28
6.3.2 Option
Limit the Scope to SAR Risk-Significant Changes
. 31
6.3.3 Option
Expand the Scope to Risk-Significant Changes 35 6.4 Parameter Change Options..................
.. 38
6.4.1 Option
Minimal increase in Probability and Consequences to Current l50.59 Param eters...........................
38
6.4.2 Option
Reduction in Margin with Control inputs 42
6.4.3 Option
Delete " Margin of Safety" as a Criterion...
42
6.4.4 Option
Define Margins with Safety and Regulatory Limits 42
6.4.5 Option
Define Margins with Fission Product Barriers - Definition 44
6.4.6 Option
Define Margins with Specified Parameters.
.. 44
6.4.7 Option
Define Margins with Mitigation Capability..
45
6.4.8 Option
Define Margins with No Reduction 45
6.4.9 Option
Define Margins with Minimal increase.
. 46 6.4.10 Option: Define Margins with % Reduction Between Calculated and Acceptance Crit eria........................
46 6.4.11 Option: NEl 96-07 Report.............
47 6.4.12 Option: Replace Parameters with Regulatory Guide 1.174 Principles..
. 47 6.4.13 Option: Frequency-Consequence Curves..
51 6.4.14 Option: Replace Parameters with Modified Regulatory Guide 1.174 Principles 55 6.4.15 Option: Risk Increase Interval (RAW)....
59 l
6.4.16 Option: Maintenance Rule Metrics (RRW/ RAW) 64 6.4.17 Option: Core Damage Frequency........
69 6.4.18 Option: Large Early Release Frequency..
74 t
I l
l
- ii -
1
-. ~... - - - -._..~.
-.--~. -. -... l l.
l i
t i
LIST OF FIGURES FIGURE
\\
PAGE i
I 2-1 Cu rrent 5 0.5 9 Process......................................................... 2 i
j i
5-1
- Aerial View of South Texas Nuclear Power Station............................... 21 i
l 52 Cross-Sectional View of Reservoir and EC Pond................................. 21 i
i LIST OF TABLES TABLE PAGE 6.1-1 Scope and Parameter options............................................... 25 t
6.2-1 Evaluation Factors and Possible Ratings...................................... 27 6.3.11 Evaluation Results for : Limit the Scope to Chapter 15 Events...................... 29 6.3.2-1 Evaluation Factors for Limiting the Scope to Risk Significant Changes to Information Described in the S AR................................................... 3 2 6.3.3-1 Evaluation Factors for Expanded Scope to Risk-Significant Changes Throughout the Facility 35 6.4.1-1 Evaluation Results for Minimal Increase in Probability and Consequences Option
................................................................ 39 l
6.4.12-1 Evaluation Results for RG 1.174 Option.
................................... 49 6.4.13-1 Evaluation Results for Frequency-Consequence Curves Option
..... 52 6.4.14-1 Evaluation Results for Modified RG 1.174 Option............................... 57 6.4.15-1 Evaluation Results for Risk Increase Interval Option.............................. 61 l
6.4.16-1 Evaluation Results for Maintenance Rule Metrics (RRW/ RAW) Option.............. 66 6.4.17-1 Evaluation Results for Core Damage Frequency Option.......................... 71 6.4.18-1 Evaluation Results for Large Early Release Frequency Option.................... 76 s
l I
i l
i t
i i
l i
l l
i l
1
- I"H' --
O
{ 50.59 Risk-Informed Option, intioduction 1.
LNTRODUCTION in the August 1995 NRC Policy Statement the Commission stated its" intention to encourage the use ofPRA and expand the scope of PRA applications in all nuclear regulatory matters to the extent supported by the state-of-the-art in terms ofmethods anddata. " This would Iead to " safety decision making enhanced by the use ofPRA insights,. more efficient use of agency resources; and... a reduction in unnecessary burdens on licensees. " In a letter from Funches to the Chairman,10 CFR Part 50, Section 59 (Q50.59) is identified as a likely candidate to benefit from the application ofrisk insights. In that regard, Mr. Funches stated that "RES will.. assess various options and develop recommendations on this topic, considering approaches such as: develop guidance that licensees could use on a voluntary basis that allows the use ofPRA results; analyzing in advance Ihe risk significance ofsmallchanges toIheplant response to accident conditions which would allowplants to make changesprovidedtheplant response stayed within certain bounds (this approach couldthen be codifiedin a revised 50 59); developing guidelines derivedfromprevious work done (e.g., BNL risk study on class 3-8 accident - NUREG/CR-0603) on the risk basis accidents; and the ACRS recommendations contained in their July 16,1998 letter. This activity would be conductedconsistent with Commission direction on SECY-98-171." The EDO's response to the Chairman's tasking memo dated identifies the activities to be carried out by the staffin the process of making 50.59 more risk informed.
The objective of this report is to provide options for integrating risk insights into the j50.59 process. In order to achieve that objective, the following high-level approach was taken:
1.
Ext.mine the existing s50.59 requirements to identify its limitations and problems, 2.
Develop a range of options with varying levels of PRA uses, 3.
Develop evaluation criteria that are consistent with Commission policy and directives, 4.
Evaluate the options in accordance with the developed criteria in developing the different options, the background (history) of the existing 50.59 process was reviewed to identify its limitations. This examination included reviewing actual examples of where NRC approval was and was not needed and the bases for those decisions. For each example, the examination also assessed the risk significance associated with the changes, in evaluating the options, criteria were developed that focused on the risk, regulatory and licensee implications and their associated resource and time impacts. The previous examples along with a new test case were evaluated against each option to provide further insights.
The options are based on technical, and not policy, evaluations. Ilowever, as stated above, the objective is to provide options for a risk-informed G50.59 process. Therefore, options that do not change the existing process or result in making the existing process more deterministic are not included. Similarly, the Commission has indicated a desire for the regulatory process to become " risk-informed" as opposed to " risk-based." Therefore, all options retain at least some deterministic aspects, for example, dealing with safety margins, defense in depth, and meeting existing regulations, such as Parts 20 and 100, and Appendix A to 10 CFR 50. Any recommendations will be provided as part of the overall effon on risk-informing Part 50.
The remainder of the repon is laid out as follows: Chapter 2 describes the current Q50.59 process, its limitations and sites examples of licensee G50.59 evaluations. Chapter 3 describes the approach for identifying options. Chapter 4 presents the evaluation factors, consistent with Commission policy. Chapter 5 presents a test case provided by one linnsee. Chapter 6 presents the options selected for further evaluation along with the results of those evaluations.
1_
o 4
s50.59 Risk-informed Option. Background 2.
BACKGROUND 2.1 Current 50.59 Regulation and Process l
Section 50.59 was promulgated in 1962 to allow licensees to make certain changes that affect systems, structures, components, or procedures described in the safety analysis report (SAR) without prior approval provided certain conditions were met. In 1968, the rule was revised to modify some of the criteria for when approval was requited. The intent ofthe f 50.59 process is topermitlicensees to make changes to thefacility, provided the changes maintain the level ofsafety documented in the originallicensing basis, such as in the l
safety analysis report. The process is thus structured around the licensing approach ofdesign basis events (anticipated operational occurrences and accidents); safety-related mitigation systems, and consequence calculations for the design basis accidents. Margins and equipment functionality, reliability and availability also may be impacted by facility changes. Therefore, the criteria for requiring NRC approval were directly related to: (1) preserving licensing assumptions concerning initiation of design basis events by not allowing a different type of initiating event or probability of occurrence larger than previously considered; (2) preserving effectiveness (reliability) of the mitigation systems by not allowing introduction of different equipment malfunctions and by limiting increases in probability of malfunction, or reductions in the margin of safety (w hich reflects the capability of the system); and (3) preserving acceptability of consequences by limiting increases in consequences of the postulated design basis events.
While the intent of the regulation to permit licensees to make changes to the facility still remains, in practice, the wording of the regulation has severely limited licensees' ability to make changes. The basic elements l
of 10 CFR 50.59 are illustrated in Figure 2-1.
l
-scora-
~ rara,,,ac&p g-w:n r m g
_ + -
[ tr.:::::. -
M(
i My nr:.-
=
- 't 7..
3,,
3 o
W v.
v.
v.
Q'~
{-
l yMe
~
f m
i tvmotx scM "wd
_s
< =:r x.m 31
~
=
pg 9
gg. a,
v.
u u w,, m
.;w Kb@
x,--
T:f hf:
'~ l$.
- a A l
- W
,i.: opa u..
.g.,
w a,m,,.u,,,,
bre f a W N.
E 4
,[
l Figure 2-1 Current 50.59 Process 1
-. 2 --
l l
. - -. - - -. -. - ~. - -
, - -. - - ~ _.
~.. _ -
i i
(50.59 Risk-Informed Option, Background They include, scope, where changes are screened as either falling under the confines of Q50.59; and parameters, where the impact of the proposed changes that fall within the l50.59 scope are assessed as either not requiring NRC approval prior to implementation or requiring NRC approval prior to implementation (classified as an unreviewed safety question).
10 CFR 50.59 defines the scope as being limited to any description identified in the SAR-i
"(1) The holder ofa license authorizing operation ofa production or utilization facility may (i) make changes in the facility as described in the safety analysis report,(ii) make changes in the procedures as described in the safety analysis report, and (iii) conduct tests or experiments not described in the safety analysis report, without prior Commission approval, unless the proposed change, test or experiment involves a change in the technical specifications incorporated in the license or an unreviewed safety question."
l As illustrated in Figure 2-1, all activities related to technical specifications (TS) require Commission (NRC) approval prior to implementation. Activities not related to TS but described in the SAR are subject to a Q50.59 evaluation to assess ifa the licensee needs NRC approval prior to implementing the proposed change.
l Once it is determined that an activity is described in the SAR, a licensee assesses if the proposed change is categorized as an unreviewed safety question. A proposed change, test, or experiment is identified as an unreviewed safety question (USQ):
(i) if the probability of occurrence or the consequences of an accident or malfunction of equipment important to safety previously evaluated in the safety analysis report may be increased; or (ii) if a possibility for an accident or malfunction of a different type than any evaluated previously in the safety analysis report may be created; or (iii) if the margin of safety as defined in the basis for any technical specification is reduced.
While some of the excessive constraints identified in the rule are being addressed in an ongoing rulemaking initiative, the initiative does not address the consideration of risk, as defined in the Commission's policy statement on use of PRAs. Consideration of risk insights into the regulatory process is a new initiative being pursued by the Commission. The options and issues to be considered when incorporating risk insights into the Q50.59 process are addressed in this paper.
2.2 Limitations of10 CFR 50.59 Regulation and Process 2.2.1 Summary of Limitations l
The following list identified limitations with the existing 50.59 process, in the context of making it risk-l informed:
i Risk insights are not incorporated into the scope of changes requiring NRC approval prior to implementation l
The identified parameters do not consider PRA insights The criteria for the identified parameters do not incorporate risk insights (i.e., do not allow minimal j
increases) l 1 -
- - - - _ - ~.. - -
I c
{ $0.59 Risk. Informed Option. Background The bases for these limitations are discussed in the following sections.
2.2.2 Scope Limitations Section 50.59 permits licensees to make changes to activities not described in the SAR without prior NRC approval. Given the spectrum of SARs amongst licensees (some consisting of a few volumes and others consisting of a book shelf full of volumes), disparity can exist among licensees as to what changes are permitted without prior NRC approval. Therefore, some licensees with limited descriptions of their facility in their SAR may be able to perform modifications without prior NRC approval that other licensees, whose SAR contains detailed descriptions of their facility, are not able to perform. Licensees with limited scope SARs would prefer not to expand the scope beyond the final safety analysis repon (FSAR), while licensees with voluminous SARs would find it advantageous to limit the scope to a subset of the SAR or to limit the l
scope to only high-safety-significant activities and issues as classified through an appropriate PRA l
evaluation.
l As presently written and interpreted, the scope of a QS0.59 review does not take into consideration the risk significance of the change, as evaluated with an appropriate PRA. Consequently, a potential exists that licensees and the NRC staff may be expending resources on activities that have negligible to no impact on l
risk or plant performance. Similarly, the possibility exists that changes are being implemented on risk significant activities not described in the SAR and should be reviewed by the NRC prior to implementation.
To correct or alleviate this inconsistency, opening the scope beyond that described in the S AR was considered as an option. Another option considered risk insights to screen issues described in the SAR that are not risk significant and do not significantly impact the operation of the facility.
l 2.2.3
" Parameter" Limitations During the design process, plant response is evaluated using assumptions that are intended to be conservative to account for uncertainties in analysis or data. In the SAR, analyses are done conservatively to account for known uncertainties in the design, construction, and operation ofnuclear power plants. These conservatisms l
are introduced into S AR analyses in numerous ways. For example, some computer codes model systems and processes in a simplified but bounding fashion. Analysis input assumptions are typically worst case values (consistent with the design and operating limits) ofinstrument drift or error, temperature, pressure, fluid volume and enthalpy, flow rate, system response time, heat transfer rate and heat capiicity, reactivity coefficients, power history and decay heat. A SAR analysis also typically assumes the worst-case single-l active failure of equipment.
National star.dards and other regulatory policies, such as defense-in-depth, constitute additional engineering i
considerations that influence plant design and operation. Commensurate with expected frequency and I
consequences ofchallenges to the system, defense-in-depth could require: (1) multiple means to accomplish safety functions and prevent release ofradioactive material (multiple barriers);(2) reasonable balance among prevention of core damage, prevention of containment failure and consequence mitigation; (3) system redundancy; (4) independence; and (5) diversity.
Various margins exist in a facility design. These margins are based on, for example, assumptions ofinitial conditions, conservatisms in computer modeling and codes, allowance for instrument drift and system response time, redundancy and independence of components in safety trains, and plant response during operating transient and accident conditions. Margin is provided by meeting codes and standards or i
alternatives approved for use by NRC, including the safety analysis acceptance criteria in the SAR and in
?
l
.- 4 l
{$0.59 Risk. Informed Option, Background supporting analyses. Not all margin that exists falls within the purview of" reduction in margin of safety' as denned in the basis for any technical specification."
When a plant is licensed, the NRC states in its Safety Evaluation Report (SER) the basis for finding each l
SAR analysis acceptable. A SAR analysis may be accepted because it was considered to be adequately l
conservative and because the NRC's acceptance criteria for that analysis are met. Frequently, the SER states j
specific conditions the NRC relied upon forconcluding that the analysis was conservative. Examples ofsuch conditions may be the use of an NRC-approved computer code, correlation, or set point methodology, specinc limitations on one or more input assumptions, or penalties inserted in a calculation to account for uncertainties. In addition to being stated in a plant-specific SER, these conditions may be found in other safety evaluations such as for an analysis method proposed by a topical report.
l i
Changes to the basis for licensing occur over the life of the plant through promulgation of new rules, plant-specific license amendments and other analyses and reviews that may be conducted, such as in response to NRC bulletins and generic letters. The NRC prepares a safety evaluation for many of these issues based upon either licensee requests for changes or licensee responses to NRC requests for information. The licensee is required to periodically update the final-SAR to renect effects of these changes so that the SAR (as updated) remains a complete and accurate description and analysis of the facility such that it can serve as the reference document for evaluation of changes made under s50.59.
liowever, once a proposed change is identined as falling within the scope of 50.59, the chances that it will not be classified as an unreviewed safety questions is, by design of the rule, very small. The reason for this limitation is due to the stringent requirements that any change in accident consequence or probability of occurrence; any change in equipment consequences or probability of occurrence; any potential for resulting in a new accident or equipment malfunction; as well as any potential reduction in safety margin require NRC approval prior to implementation.
l Since the promulgation of l50.59, in 1962 and revised in 1968, the staff and the industry have gained l
experience and technological improvements that enable better understanding of the analysis, design and operation of nuclear facilities. At the same time, technological improvements and economic considerations have reached a cross road where a revisit of SAR assessments may be beneficial to the public, licensees and j
the NRC staff. Economic conditions have imposed hardship on selected suppliers such that " safety grade" equipment are special orders and very costly. At the same time, economic enhancements could have improved such that industrial grade equipment may be acceptable substitutes. The existing;;Q50.59 process does not take such changes into consideration.
This inflexibility of the present regulation can result in unnecessary expenditure of analysis and review by both the industry and the NRC. The existing regulation can also prevent implementation ofincremental plant improvements that are not safety significant and do not impact plant operations because the window of opportunity to implement changes escape due to the added expense and time required for NRC review and approval.
2.3 Examples of Facility Changes Using the 10 CFR 50.59 Process l
This section describes actual examples of licensee's facility changes requiring and not requiring NRC l
j approval. Discussions address why approval was required along with the potential risk significance of the i
)
' Margin of safety is not defined in the regulations, although it is mentioned in {5034(a) ["the margins of safety during normal operations and transient conditions anticipated during the life of the facility *). {$0.92(c) ["No significant hazards considerations if the proposed amendment would not involve a significant reduction in a margin of safety") as wc!! as {50.59. I i
{50.59 Risk Informed Option, Background change to provide insights as to the constraints and limitations of the current 50.59 process. In this report, a risk significant change is defined as one that causes one or more risk measures (e.g., CDF or LERF) to exceed a specified value (for illustrative purposes,1 E-7 forACDF, i E-8 ALERF). Other measures could also be applied, such as frequency-consequence curves or meeting the subsidiary objectives of the Safety Goal 3
l Policy Statement, etc.
2.3.1 Facility Changes Needing NRC Approval The following examples were taken from licensee submittals where the licensee identified the change as being an unreviewed safety question (USQ); thus requiring NRC approval. However, based on the information provided in the submittals, it has been determined that the changes were not risk significant and i
thus should not require NRC approval.
2.3.1.1 Error in Dose Calculations for the Process Gas System Proposed Facility Change in this example, the licensee submitted a request to change the Section 15 updated safety analysis report (UFSAR) dose values to reflect as-built conditions. The licensee identified that the realistic dose values for the process gas system rupture in Section 15 of the UFS AR had not been updated to correct the realistic dose at the exclusion area boundary from 6.52E-03 rem to 6.53E-03 rem, and the dose at the low population zone increased from 1.50E-04 rem to 1.51E-04 rem.
Why an UnreviewedSafety Question Existed l
Since the correct dose value was greater than tha*. previously reported in the UFSAR, the consequences of the accident had increased, and a USQ existed.
Risk Significance Negligible. Since the change resulted in an increase ofdose to the exclusion area boundary of IE-5 rem and i
an increase to the low population zone of IE-6 rem, the risk significance of this proposed facility change was judged to be negligible.
2.3.1.2 Errorin Air Volume of Secondary Containment Proposed Facility Change in this example, the licensee identified an error in the free air volume of the sece,dary containment. The i
actual free air volume was 18% less than the UFSAR value. The licensee submitted a request to change the UFSAR to reflect as-built conditions.
Why an UnreviewedSafety Question Exists The licensee's analysis credited 50% mixing in the secondary containment. The decrease in the secondary containment volume, resulted in an increase thyroid dose to in the control room operator. The control room operator thyroid dose increased from 22 rem to 25.6 rem and, therefore, the error represented a USQ due to the reduction in the margin to safety and an increase in the consequences of an accident. To offset this dose increase, the licensee requested a technical specification (TS) change (which also requires NRC approval per 550 59 Risk. Informed Option, Background 650.36) to increase the removal efficier:cy for the standby gas treatment charcoal from 90% to 95% with i
more restrictive surveillance acceptance criteria. With this adsorber emeiency, the thyroid dose actually l
decreased from 22 rem to 21.88 rem.
Risk Significance Insignificant. The error in secondary containment free volume increased the calculated thyroid dose to the operators by 2.6 rem to 25.6 rem. The acceptance criteria for GDC-19 is 30 rem. While the increase was within regulatory limits, the licensee compensated for this increase by modifying the technical specifications to require a 95% removal efliciency for the standby gas treatment charcoal filters. This modification reduced the calculated dose to the operators from 25.6 to 21.88 rem, slightly below the incorrect FSAR value.
2.3.1.3 Calculational Error in IIcad Loss of Emergency Core Cooling Suetion Strainers Proposed Facility Change In this example, the licensee identified a deficiency in the net positive suction head (NPSH) for the emergency core cooling system (ECCS) pumps. The licensee determined that the calculations supporting the original SAR design basis head loss across the ECCS suction strainers was not the actual pressure drop that would exist during a design basis accident (DBA). The licensee requested that the UFSAR description be modified to reDect the calculation results which credited a nominal amount of containment pressure to compensate for the deficiency in NPSH for the ECCS pumps following a DBA during the short-term accident injection phase. In addition, the licensee also requested a TS change (which required NRC approval) to lower the allowable water temperature in the suppression chamber and ultimate heat sink.
Why an Unreviewed Safety Question Exists The pressure drop across the ECCS suction strainers are utilized in the calculations which demonstrate that adequate NPSil was available to support the operation of the ECCS pumps during DBA conditions. A USQ existed because a reduction of NPSH could increase the probability of ECCS equipment malfunction.
Risk Significance None. Through the reanalysis, licensee was able to compensate for the deficiency in NPSH; thus, meeting the original NPSH requirements.
2.3.1.4 Modification of Main Feedwater Valves Proposed Facility Change in this example, the licensee intended to modify the single failure trip logic of the main feedwater control and bypass valves from that described in the UFSAR.
Why an UnreviewedSafety Question Exists While this modification reduced the probability of a reactor trip, it slightly increased the unavailability of the feedwater isolation function. The probability of a reactor trip was reduced by 41.75% (1.20E-1 per year for current design compared to 6.99E 2 per year for proposed design). However, the failure probability of feedwater isolation increased from 2.8E-5 per demand to 6.lE-5 per demand. Therefore, the modification resulted in an increase in the probability of equipment malfunction occurrence.
t l
7.-
i
l
{50.59 Risk-Infonned Option. Background Risk Significance I
None. Overall, the proposed modification resulted in slight reduction in the CDF because of the decreased probability of a spurious reactor trip.
2.3.1.5 Discrepancy Between As-Built System and UFSAR Description Proposed Facility Change l
In this example, the licensee identified that the isolation scheme of the main condenser vacuum pump had not been properly descr; bed in the UFSAR and requested approval to change the UFSAR to reflect as-built plant conditions, i
Why an UnreviewedSafety Question Exists A USQ existed because the as-built system isolation was of a difTerent type than the one previously evaluated in the SER. During plant construction, devices that would automatically trip the pump and close an isolation valve, as described in the SER, were never installed. This resulted in an incorrect licensing basis as defined in the FSAR and SER.
Risk Significance None. As a result of compensatory operator actions, the licensee was able to restore dose limits to within the original licensing basis.
2.3.1.6 Examples of Facility Changes That Are Not Pursued as License Amendment Because of the Expense, Uncertainties, or Timeliness of Amendment Approval i
South Texas Project currently has approximately 1000 components that they would like to reclassify from safety.related to nonsafety-related in implementing their Graded QA program. To do this would require Q50.59 evaluations and license amendment submittals. Implementing the changes through the current 10 CFR 50.59 process would be costly and resource intensive for both the licensee and the NRC. This report j
has chosen to use one of these component changes as a test case for assessing the various p'roposed 650.59 options. The test case is described in Section 5.
2.3,2 Facility Changes Not Needing NRC Approval Section 50.59(b)(2) requires that the licensee submit a report containing a brief description of each change, test, and experiment, including a summary ofits supporting safety evaluation, implemented without prior l
NRC approval in accordance with 50.59. It states that the report may be submitted annually or along with FSAR updates. The following are examples of changes licensees have made to their facilities that did not require NRC approval before implementation.
l 2.3.2.1 Residual Heat Removal Service Water Sump Pump Discharge Check Valves Replacement Proposed Facility Change in this example, the licensee replaced a residual heat removal (RIIR) service water sump pump discharge check valves, as described by the UFSAR, with Mission DUO-CHECK check valves.
i 950.59 Risk. Informed Option, Background l
Why an UnreviewedSafety Question DidNot Exist A USQ does not exist since the new valves serve the same function as the existing valves. Since the failure modes are the same, the consequences of an equipment malfunction are the same. Reliability of the check valves will be increased since it will be less likely that debris will collect (due to a less torturous flow path through the valve) and allow leakage past the valve seats. With the new check valve hinge pin positioned in the vertical direction, the resistance of the new valves is less than the existing valves. Process parameters are not being altered and operating modes are not being changed. Therefore, there are no new accidents or failure modes created by this change.
Risk Signspcance None. Since the reliability of the check valve has increased and the consequences of an equipment j
malfunction have not change, a USQ does not exist.
j 23.2.2 ManualIsolation of the Control Rod Drive Pumps during a Station Blackout or Loss Of Offsite Power Procedure Change Proposed Facility Change In this example, the licensee added procedural steps to manually isolate the control rod drive (CRD) pumps during a station black out (SBO) or loss of off site power (LOOP). This procedure is described in the UFSAR.
Why an UnreviewedSafety Question Does Not Exist A USQ does not exist since this procedure change mitigates the consequence of a design basis accident. In fact, these procedural changes were in response to NRC Information Notice 90-78, which described a previously unidentified radiation release path in the CRD system. The procedure change isolates the CRD system manually from the reactor primary system, w hich prevents the possibility of vessel inventory or radio nuclides leaking backwards through the CRD system into the contaminated condensate storage tanks. This prevents an increase in dose rates, and therefore, the consequences of an accident are not increased. The safety-related function ofthe CRD system is unaffected, because the scram capability is cont!.ined com pletely in the hydraulic control unit. The CRD system can also be used per existing operating procedures for alternate injection or normal system operation.
Risk Signspcance None. A USQ does not exist since this procedure change mitigates the consequence of a design basis accident.
2.3.2.3 RIIR Low Pressure Core Injection Inboard Injection Valve Design Change i
i ProposedFacility Change In this example, the licensee modified an RHR injection valve, described in the SAR, by drilling a 3/16" diameter hole through the reactor recirculation piping inlet (high pressure) side of the valve flex-wedge disc.
l
_9
l 1
i
{50.59 Risk-Informed Option, Background i
Why an Unreviewed Safety Question Does Not Exist
(
A USQ does not exist since the drilled hole improves the reliability of the valve to perform its function to open for low pressure coolant injection by permitting the pressure in the bonnet to equalize with the piping system thereby reducing the differential pressure and eliminating pressure locking susceptibility.
The function of the valve to close for the required primary containment isolation is not affected by this l
change. This change improves the RilR system's capability to respond to the long tem cooling requirement of the reactor as specified in the UFSAR for anticipated transients or accidents. The change is entirely internal to the valve so there are no new interfaces created between the valve and other systems. The change does not create any new failure modes for the valve. Drilling a hole in the valve disc does not increase the probability of a loss of pressure boundary integrity. The valve disc is not part of the system pressure boundary, and the ability to isolate the rector coolant has not been reduced because the outboard valve disc will remain intact.
Risk Significance None. Since the reliability of the check valve would be increased and the consequences of an equipment malfunction would not change, a USQ does not exist.
l i
I.
l (50.59 Risk-Informed Option, Approach 3.
APPROACH FOR IDENTIFYING OPTIONS The staff considered a wide range ofoptions for revising Q50.59. Various combinations of changes to both scope and parameters were considered. While no option was considered unreasonable a priori, an informal screening process was used by the stafTto select those options to carry forward for further evaluation. In selecting the options to carry forward, four general questions were addressed:
- 1. Could the option effectively deal with the problems currently associated with j50.59,
- 2. Could the option be constructed so as to maintain adequate safety,
- 3. Are current regulatory constraints and fundamental operating principles compromised, and
- 4. Is Commission approval of the option likely?
Current problems associated with Q50.59 were discussed in Sections 2.2 and 2.3. Many, but not all, of the problems reflect cases where utilities are precluded from making {50.59 changes inappropriately. That is, NRC review is being required when it should not be necessary. These problems result in additional resource expenditures by both the licensees and the'NRC staff with no corresponding safety benefit. The staff identified options that would allow the scope of {50.59 to be focused down to those items of clear safety significance and that would not require excessive conservatism in engineering calculations. Varying degrees l
of scope and parameter changes were considered. Options were rejected if they were perceived to lead to a different, but equally cumbersome, process.
If an option appeared to improve on the current 50.59 process as noted above, then assurance of safety was considered. That is, the staff considered the potential for the Cornmission's safety goals or adequate l
protection criteria to be exceeded. If that potential was significant, then the option was rejected or mod!fied.
i However, the staff did consider some allowable risk increases that would not violate the safety goals or adequate protection criteria. Various metrics are possible for evaluating allowable risk increases, as discussed further in Chapter 6. The staff's objective in addressing question 2 above was to make sure that factors important to plant safety were not left out. In some cases, this may involve expanding the current scope of 50.59. For example, for some options most plants would be able to exclude some SAR items based on a lack of safety significance. However, those plants with limited SARs may need to include some additional SSCs that have been shown to influence their PRA results even though those SSCs were not included in the SAR.
The staff did not consider options that would allow l50.59 to supercede other regulations and operating principles, such as those imposed by 10 CFR Parts 20, the remainder of Part 50, and 100. These regulations and principles were established to control plant operation and performance such that plant and personnel safety are not unnecessarily challenged. Examples where unreviewed changes would not be allowed under any Q50.59 option include acceptable fuel design limits (minimum DNBR and MCPR), containment design l
pressure, additional challenges to safety systems, and radiation management (ALARA). Review would be
{
required in these cases even ifa PRA showed the risk impact to be minimal. This position is consistent with l
positions established in other recent staff documents dealing with risk-informed regulation, including RG 1.174.
Finally, the Commission has set forth additional policies and positions concerning the regulatory process.
In particular, the Commission has said that the regulatory process should be risk-informed and not risk-based.
Therefore, options based solely on PRA results were excluded from funher consideration, j
Within the context of the discussion above, a range of options was generated. Options were eliminated if the l
answer to any of the four questions above was no. Scope and parameter changes were proposed ranging from J
J
9
{50.59 Risk-Informed Option, Approach minimal to very extensive. Chapter 4 describes the evaluation factors used to produce a systematic evaluation of those options that survived the screening described above.
t l
l i
l 1
l l
I 4
l I i l
.~
,.. ~ -.
l i
650.59 Risk-Informed Option, Evaluation Factors
'4.
EVALUATION FACTORS This section defines and describes the factors that are used to evaluate and assess the options for making
$50.59 risk-informed. These factors are the same as those used in assessing the options for risk-informing Part 50. These factors, as presented in the forthcoming paper to modify Part 50, include the following:
potential for improving safety decisions potential for reducing licensee and NRC burdens e
the anticipated complexity of changes e
NRC resources needed for putting changes in place licensee resources needed for putting changes in place e
calender time for full implementation (NRC and licensee)
These factors were grouped into five functional categories:
PRA Policy Statement implications, regulatory implications, a
licensee implications, resource implications, and
=
time implications, in addition, two new factors were added:
risk implications and facility coverage implications The following summarizes the different factors used in the evaluation and their potential ratings.
PRA Policy Statement Imphcations 1
Enhanced Safety Decisions no/yes More Eflicient Use of Agency Resources no/same/yes Reduction of Unnecessary Burden no/same/yes Facility Coverage implications same/less i
l' Risk Implications Confidence Required from Risk Results high/ low Required PRA Completeness a
scope completc/ full / medium / minimal n
level of detail detailed / simple Deterministie vs Risk-Informed deterministic / deterministic-risk-informed / risk-informed i
Regulatory implications 1
Regulatory Guidance Needed yes/no Impact on Other Regulations yes/no I
New Rulemaking Needed yes/no
- Impact on Inspectionffraining yes/no
- 1-
?-
I
)
650.59 Risk-Informed Option. Evaluation Factors Licensee implications Internal Procedures Needed yes/no Impact on PRA Activities yes/no Impact on Training yes/no Impact on Inspection increase /same/no Impact on Facility Changes yes/no Resource implications e
Short term minimal / moderate /significant Long term more/same/ fewer Time implications Short term (development) short/ medium /long Long term (day-to-day implementation) more/same/less 4.1 PRA Policy Statement Implications A PRA Policy Statement was issued which stated "...the Commission 's intention to encourage the use offRA andto expandthe scope ofPRA applications in allnuclear regulatory matters... Implementation ofthepolicy statement will improve the regulatoryprocess in three areas: Foremost, through safety decision making enhanced by the use of PRA insights; through more efficient use of agency resources; and through a reduction in unnecessary burdens on licensees.. " In evaluating the proposed options, the degree to which each option accomplishes the goal of the Commission's PRA Policy Statement is an essential factor. The three objectives of the goal are discussed below with their ratings.
4.1.1 Enhanced Safety Decisions This measure defines whether the option will enable the staff and the industry to make enhanced safety decisions. Enhanced safety decisions are made when using PRA insights that focus the decision-making process on the more risk significant changes. In evaluating this factor, two different ratings are considered:
No -PRA insights are not incorporated into the decision-making process Yes -PRA insights are incorporated into the decision-making process 4.1.2 More Efficient Use of Agency Resources This measure defines whether the option will enable the staff to make more efficient use ofagency resources for each individual change. The resources required to make the option part of the day-to-day operation is addressed in Section 4.4. More efficient use of agency resources occurs when PRA insights are used because of a more risk-informed decision-making process. In evaluating this factor, three different ratings are considered:
No -a decrease in efficiency is expected Same -the same amount of resources are expected Yes -an increase in efficiency is expected
j i
{50.59 Risk.Infonned Option. Evaluation Factors 4.13 Reduction in Unnecessary Burden on Licensees This measure defines whether the option will reduce unnecessary burden on licensees for each individual change. The resources required to make the option part ofday-to-day operation is addressed in Reduction in unnecessary burden on licensees occurs when PRA insights are used because of a more risk-informed decision making process. Unnecessary burden is defined in this report to mean requiring licensee submittals for changes with negligible or no risk significance. In addition..... reduce screening or evaluation process.
In evaluating this factor, three different ratings are considered:
No -an increase in burden is expected l
Same -same amount of effort required e
i Yes -a decrease in burden is expected 1
4.2 Facility Coverage Implications This measure defines to what extent the facility is affected by each of the options. The existing QS0.59 includes any changes to the entire facility as described in the FSAR; it does not limit facility changes to a single radiological hazzard. In evaluating this factor, two different ratings are considered:
l Same -covers the entire facility, consistent with the current Q50.59 process
=
Less -does not cover the entire facility (e.g., only addresses reactor core hazard) 4.3 Risk Implications In evaluating each of the options, how " risk" is used is an essential discriminator. The confidence needed i
from the risk (e.g., PRA) results, the sophistication of the risk analysis needed, and the degree to which risk is used to implement the option are three measures in discriminating between the proposed options.
I 43.1 Confidence Required From Risk Results I
This measure defines the extent to which the option is dependent on the fidelity of the ris'. results. The j
fidelity of the results can be measured without actually examining the details of how tia 6k analysis was performed if the PRA was done in accordance with an endorsed standard. In evaluating this factor, two different ratings are considered:
High - to incorporate this option, high confidence in the results is needed, and therefore, it is assumed that the PRA needs to be performed in compliance to an endorsed standard.
Low -to incorporate this option, low confidence in the results is needed, and therefore, it is assumed that the PRA does not need to be performed in compliance to an endorsed standard.
43.2 PRA Completeness This measure defines the extent to which the completeness of the risk analysis needed to suppon the evaluation that is needed to support the change in question. The completeness of the risk analysis is defined by the scope and level of detail.
The scope identifies how complete the list of hazards, initiating events and operational states must be included in the PRA and determines the minimum level of PRA analysis required to support the determination of the changes in specified risk metrics. In evaluating this factor, three different ratings are considered:
l
{50.59 Risk-Informed Option Evaluation Factors Complete - a complete-scope risk analysis is needed to implement the option; a risk analysis addressing all hazards, risk measures (i.e., health effects), operating states and both internal and external events Full-a full-scope risk analysis is needed to implement the option; a risk analysis addressing only e
core damage / melt, risk measures (i.e., health effects), operating states and both internal and external events Medium -only a risk analysis addressing core damage (e.g., Level 1 PRA) for internal events (excluding intemal fires) at full-power operation including a simplified Level 2 estimating a large i
carly release frequency and with a qualitative analysis for other initiators snd oprational states Minimal-only a Level 1 PRA for internal events (excluding internal fires) at full-power operation and with a qualitative analysis for other initiators and operational states l
The level of detail defines the degree to which systems, components and human performance must be represented in the PRA models. In evaluating this factor, two different ratings are considered:
Detailed -system, component and human failures must be represented by all failure mechanisms as defined in the endorsed standard Simple - system, component and human failures need only be represented by dominant failure mechanisms in the risk analysis to implement the option 4.3.3 Deterministie vs Risk-Informed This measure defines the extent to which risk insights are incorporated into the option (i.e., identifies the
" level of risk-informedness" of the option. In evaluating this factor, three different ratings are considered:
Deterministic-no numerical determination of risk is made, only a qualitative determination is used Deterministic / Risk-Informed-a numerical determination of risk is made; however, no changes are allowed to DBAs Ris A-Informed -a numerical determination of risk is made and changes to DBAs are allowed e
4.4 Regulatory Implications in evaluating each option, how the regulatory process is impacted is an important factor. Your impacts on the regulatory process were identified for evaluation: regulatory guidance, other regulation, rulemaking and inspection.
4.4.1 Regulatory Guidance Needed This measure defines whether regulatory guidance is needed to make the option part ofday-to-day operations.
In evaluating this factor, two different ratings are considered:
i res -to incorporate this option, regulatory guidance is needed
=
No -to incorporate this option, regulatory guidance is not needed 4.4.2 Impact on Other Regulations This measure defines whether it is necessary to modify other regulations to make the option pan of day-to-day operations. In evaluating this factor, two different ratings are considered:
650.59 Risk. Informed Option, Evaluation Factors Yes -to incorporate this option, other regulations are impacted and may need to be modified No -to incorporate this option, no other regulations are impacted 4.4.3 New Rule Making This measure defines whether new rulemaking is required with the option. In evaluating this factor, two different ratings are considered:
Yes -to incorporate this option, new rulemaking is needed.
No -to incorporate this option, no new rulemaking is needed.
4.4.4 Impact on Inspection /I' raining This measure defines the extent to which the option impact overall inspections; that is, it measures whether new training and new inspection modules will be required to make the option part of day-to-day operations and whether additional inspector time will be required to inspect for compliance. In evaluating this factor, three different ratings are considered:
Yes - to incorporate this option, new training and new inspection modules are required No - to incorporate this option, no new training or inspection modules are needed and no additional time is required to inspect for compliance 4.5 Licensee Implications In evaluating each option, how the licensee is impacted is an important factor. Three impacts on the licensee's process were identified for evaluation: internal procedure guidance, inspection and training.
4.5.1 Internal Procedures Needed t
This measure defines whether internal procedures are needed by the licensee to make the option part of the day-to-day operations. In evaluating this factor, two different ratings are considered-Yes -to incorporate this option, new internal procedures are needed No -to incorporate this option, existing internal procedures are sufficient, no changes required 4.5.2 Impact on PRA Activities This measure defines whether the licensee will be required to enhance, maintain, and update their PRAs. In evaluating this factor, two different ratings are considered:
1 Yes -to incorporate this option, the licensee will be required to enhance, maintain, and update their PRA No -to incorporate this option, the licensee will not be required to enhance, maintain, and update their PRA 4.5.3 Impact on Yraining l
This measure dr <ines whether additional training is needed to make the option part of the day-to-day I
operations; that L it measures whether new training is needed in understanding the option, how to respond, l
i l
l I
l 650 59 Risk-Informed Option. Evaluation Factors 1
etc. In evaluating this factor, two different ratings are considered:
Yes -to incorporate this option, new training is needed No -to incorporate this option, new training is not needed 4.5.4 Impact on Inspection This measure defines the extent to which the licensee will need to support and respond to inspections; that is, it measures whether other licensee personnel, i.e., PRA staff, will be required to support the inspection process. In evaluating this factor, three different ratings are considered:
Increase -to incorporate this option, the support provided by the licensee is increased Same -to incorporate this option, the suppon provided is not impacted No -to incorporate this option, the support provided by the licensee is decreased 4.5.5 Impact on Facility Changes This measure defines whether the number of cost-beneficial changes of negligible risk significance desired by licensees needing staff approval is impacted. In evaluating this factor, two different ratings are considered:
Fes -incorporation of this option is anticipated to result in more changes without staff approval No -incorporation of this option is not anticipated to result in more changes without staff approval 4.6 Resource Implications In evaluating each of the options described above, the resources required by the NRC and the licensees to implement the proposed option also must be examined. There are both short-term and long-term resource implications.
For short-term (i.e., make the option part of day-to-day operation), three different ratings are considered:
Minimal-to incorporate this option, it is anticipated that less than 1 person-year of effort will be needed Moderate -to incorporate this option, it is anticipated that more than 1 person-year ofeffort but less a
than 2 staff-years will be needed Significant -to incorporate this option, it is anticipated that more than 2 person-years of effort will be needed For long-term (i.e., the ongoing application of the new option) implications, three different ratings are considered:
l More -to incorporate this option, more resources are required than currently expended on the existing {50.59 process Same - to incorporate this option, the same amount ofresources are required as currently expended j
on the existing %50.59 process Fewer -to incorporate this option, less resources are required than currently expended on the e
existing s50.59 process l
150 59 Risk-Informed Option, Evaluation factors
. 4.7 Time Implications in evaluating each of the options described above, the time needed to implement the proposed option also must be examined. The calendar time needed by both the NRC and licensees to make the option part of day-to-day operations, three different ratings are considered:
Short -less than 6 months Medium -6 months to 1 year Long -greater than 1 year in evaluating each of the options described above, the long-term time needed to implement the proposed option must be examined. Three different rating are considered for the calendar time needed by both the NRC and licensees:
More - more time is needed than currently expended on the existing 50.59 process Same - the same amount of time is needed as currently expended on the existing {50.59 process Less - less time is needed than currently expended on the existing 50.59 process i
\\
r I
l 650.59 Risk. Informed Option, Test Case 5.
Test Case Description This section describes a potential facility change to the South Texas Electric Generating Station (STEGS) that would fall under the scope of the existing Q50.59 process and would require the licensee, Houston Lighting and Power (IIL&P) to determine whether NRC approval was needed. The evaluation must answer the seven parameter questions. The example' will be used further as a test case in evaluating all potential options of a risk-informed Q50.59 process that have been or will be discussed in this paper. The facility change under consideration is to reclassify the safety-related screen wash booster pumps to non-nuclear safety. These pumps are described in the FSAR, and therefore, fall under the scope of Q50.59.
For example purposes, it is assumed that the assessments made by HL&P are correct. These assessment have not been reviewed by the staff and would be subject to audit at a later time.
5.1 Description of Proposed Change At the STEGS, the Essential Cooling (EC) Water (ECW) provides cooling water to the essential cooling water chillers, the diesel generators, and the component cooling water heat exchangers. The ECW system is comprised of three independent single pump trains that draw suction from the EC pond. Each ECW pump is housed in an intake structure with traveling water screens and screen wash booster pumps. Trash bars upstream of the traveling water screens prevent large debris from fouling the traveling water screens. The function of the traveling water screens is to keep the intake structure free of smaller debris. Any debris trapped by the traveling water screens is dislodged from the screen using the screen wash booster pumps and is directed via a sluice to trash pits, thus removing that debris from the EC pond. The screen wash booster l
pumps are present to wash debris from the ECW traveling screens to assure adequate ECW pump suction.
The screen wash booster pumps are started in response to high delta-pressure which might indicate l
l obstructions on the ECW traveling screens. The screen wash booster pumps are also au:omatically initiated on a safety injection signal.
The EC pond is treated with biocide to prevent growth such as algae. The surrounding environment is considered a coastal plain (e.g., a wet marsh-land) and the EC pond is normally debris free. Figure 5-1 provides an aerial view of the site and surrounding area. The history of the plant indicates that the ECW screen wash booster pumps have only twice been auto initiated on a high delta-pressure. In May of 1994, an infestation of algae affected the EC pond. The operating ECW trains could generally manage a week without having an automatic start of the screen wash system. The algae disappeared when treated with biocide and has not reappeared since then. The second auto initiation on high delta-pressure occurred in the fall of 1996 where transiting ducks were believed to deposit pondweed. Most of the pondweed was blocked by the trash bars.
l a
E i
i 1
j l l
i (50 59 Risk-informed Option, lest Case
- J; <. _ ~
. L. e..
.._-.=.=:...- =. ;> y ~. -
.n....-
- -m m_p;p. a.
v.. m
,.+ - e w
~.m
- xv
- ~. u...,
w -
w -
_. a. n p. s %p %,.,
z :::
..a-.,..x.,..
qq;'mm..
c r7
~.- = +. s.g y?:
~..
3T~'Q.w s,? '.f...g:" '-
2..
c-2 p., c -*.. - w a.c
- w ;.;
.g.gy
- Q h.
g;.~y~~~
- 5. v u..v
,a,, N Aw-g% g
- ) N /.a-t u-up~M. 2 7yM:
x4.o%d w -
' a c 4. y**.
r-g*-y ep.j
.. -- U Q-
- *4
- y uw s -
r.
n 2 3 "' '-
,4 w..
w--
~. -. *h
~.
s
,o,
....W
',k j s a
Figure 5-1 Aerial View of South Texas Nuclear Power Station
~Ihe EC pond is sunounded by an eight foot embankment. The water supply to the EC pond is a reservoir, which is surrounded by an even higher embankment. Figure 5-2 provides a cross-sectional view indicating the relatis e eles ations. The water in the reservoir is fed from the Colorado river. There are two major dams (i c., large lakes) upstream of the site (Mansfield and Buchanan dams).
Cross-Sectional View E n h.mla nen ts Iw, = " r p7;
- c..l.....L gjj ger ini,ay.
iem.,
f/)
M liCW ECT
,,,7
,-;',;~g z' c; f/)
t%nd
////
Y nd f
f/)
////
7.:t.2:#: % ;'a.
- rp: t/)
////
7 n:m2 f/)
y 1///
. T p.b ' w ; f/rwwwa//
//// ~*/
>f?
n % Ev-2c^^ t//////// "i
////
'////
- gt. W:f Sf//////// Ma
- r..:
'////
f//////// raw n////////rawv.'////
- g[. ~ W f////////////////////////// ^
e s o$ > > > #$ 6 $ >$
$$a o
a Figure 5-2 Cross-Sectional View of Reservoir and EC Pond.
{50.59 Risk. Informed Option. Test Case t
l The different mechanisms that could fail the function of the ECW pumps in response to internal and external l
events were identined and analyzed. In particular, the need for the screen wash booster pumps to ensure ECW pump operation was examined.
For internal events, the failure of the screen wash booster pumps were examined and it was determined that their failure did not prevent the ECW pumps from operating. This determination was arrived at because the EC pond is normally debris free (see above discussion). It was, therefore, concluded that the screen wash booster pumps would not' be required to support the ECW in the mitigation of an internal event at full (normal) power operation. Consequently, it was concluded that there is negligible consequence from their failure, i.e., they have negligible risk signincance (again, only in regard to internal events, full-power operation).
For extemal events, the loss of ECW function from a severe storm, tornado, or flooding was examined.
These events are denned in the FSAR as " incredible events," and as such, are not evaluated in Chapter 15 of the SAR. The potential for debris in ECW pond and fouling the traveling water screens from high winds or tornadoes was examined. The occurrence of these external events are of very low frequency (i.e.,~1E-6/ry). In addition, for these events, analysis indicaes that these pumps could not provide adequate mitigation, and therefore, their failure had no impact and were not modeled. The loss of the embankment surrounding the reservoir, or failure of dams on the Colorado river were also examined. The occurrence of these events which results in failure of the ECW function are ofvery low frequency (i e., ~3E-7/ry). Consequently, their risk significance is negligible (in regard to external events).
The above is a brief summary of a detailed analysis performed by liL&P. This analysis (and any uncertainties) has not been reviewed by the staff. For purposes of this report, it is assumed that the analysis is accurate in that these screen wash booster pumps have negligible risk signficance.
5.2 Evaluation Against 50.59 Parameters i
i The facility change under consideration is to reclassify the ECW screen wash booster pumps (which have been shown to have negligible risk significance) from safety related to non-nuclear safety. Reclassifying the i
ECW screen wash booster pump from Safety Class 3 to non-nuclear-safety (e.g., industrial grade) may result in the application ofless rigorous quality assurance requirements and reduced testing. The reclassification would not result in a change in the function or operation of the pumps or their performance requirements.
)
The evaluation below follows the format of IIL&P's {50.59 procedure (transmitted to the staff for l
information only). The evaluation concludes that a USQ would exist in answering question (3) below. This i
conclusion is reached because of the deterministic "may" terminology of {$0.59 and because of the definitions of "important to safety" and " safety-related" The deterministic definitions of"important to safety" and " safety-related" drive this determination. The screen wash booster pump has negligible risk significance at STEGS. In addition, the use of"may" in the {50.59 wording does not even allow for a mere potential increase in probability.
May the change increase theprobability ofoccurrence ofan accidentpreviously evaluatedin the SART If a screen wash booster pump inadvertently actuates, there is no effect on the operation of the ECW system.
If a screen wash booster pump fails to actuate on high delta-pressure, the associated ECW train may be impaired or fail; however, failure of the ECW train would not cause an accident previously evaluated in the SAR. Consequently, reclassifying the screen wash booster pumps to non-nuclear safety would not increase the probability of occurrence of an accident evaluated in the SAR. 1
--. _ - - -....~
_ - ~.. - --
i l
l t
6$0 59 Risk-Informed Option. Test Case May the change increase the consequences of an accident previously evaluated in the SAR?
Reclassifying the ECW screen wash booster pumps would not result in a change in the operation of the pumps or their performance requirements, although some unquantifiable reduction in reliability may be implied. There is no credible basis to postulate external events that would generate debris in the ECW pond concurrent with a DBA. Consequently, the screen wash booster pumps would not be required to support the ECW in the mitigation of the event and there is not potential for an increase in on-site or off-site dose.
Therefore, the consequences of an accident previously evaluated in the SAR are not increased.
May the change increase theprobability ofoccurrence ofa malfunction ofequipment ingportant to safety previously evaluated in the SART
\\
Reclassifying the ECW screen wash booster pumps from Safety Class 3 to non-nuclear-safety may result in the application of less rigorous QA requirements and reduced testing used in its ASME certification.
Although the reclassification would not result in a change in the function or operation of the pumps or their performance requirements, some unquantifiable reduction in reliabilitymay be implied. Becauk the screen wash booster pump is a Safety Class C component, it is considered to be important to safety. Therefore, the reclassification may reduce the reliability of the pump and increase the probability that it will malfunction.
May the change increase the consequences of a malfunction of equipment important to safety previously evaluatedin the SART The reclassification would not result in a change in the function or the operation of the pumps or their performance requirements, although some unquantifiable reduction in reliability may be implied. There is no credible basis to postulate external events that would generate debris in the ECW pond concurrent with a DBA. The screen wash booster pumps would not be required to support the ECW in the mitigation of the event so there is no consequence form their failure. Therefore, there is no potential for an increase in on-site or off-site dose and the consequences ofmalfunction ofequipment important to safety previously evaluated in the SAR are not increased.
May the change create the possibility of an accident of a dVferent type than any previously evaluatedin the SART No. The screen wash booster pumps are not accident initiators. No possibility of a different type of accident than previously evaluated is created.
May the change create the possibility of a dVferent type of malfunction than any previously evaluatedin the SAR?
Postulated failure of a screen wash booster pump may cause malfunction ofits associated ECW train.
Malfunction of ECW has been evaluated in the SAR; therefore, the possibility of a different type of malfunction is not created.
Does the change reduce the margin ofsafety as defined in the basisfor any TS?
The TS of interest is the ECW specification and the requirements for operability of the ECW.
Declassification of the screen wash booster pumps will not have a material effect on their function or performance. Normal preventive maintenance will assure reliability and availability. The screen wash booster pumps will still perform as required to support the ECW; therefore, the margin ofsafety for the ECW TS is not reduced.
(50.59 Risk-infbrmed Option, Test Case booster pumps will still perform as required to support the ECW; therefore, the margin of safety for the ECW I
TS is not reduced.
(Note that if the screen wash booster pumps were reclassified as non-seismic, the response to this question would be "Yes" since LOCA and seismic are required to be considered concurrently in the regulations. A non-seismic screen wash booster pump could be cor rued as a reduction in the seismic safety margin ofthe ECW.)
i
)
l
m.
I f 50.59 Risk Informed Option, Options I
6.
Options This section describes various options the staff considered for incorporating risk insights into the Q50.59 process. As illustrated in Figure 2-1, the existing Q50.59 process involves consideration of" scope" and
" parameters." The following describes, evaluates, and provides observations for each option separately.
6.1 Scope and Pa.rameter Changes The staffreviewed options for incorporating risk insights into the decision making process for both the scope and parameter evaluation of Q50.59. These are summarized below in Table 6.1-1. An evaluation of each option is discussed in the following sections.
Table 6.1 1 Scope and Parameter Options s
l 6.3.2 l Limit the Scope to SAR Risk-Significant Changes l
6.3.3 Expand the Scope to Risk Significant Changes 6.4.1 Minimal increase in Probability and Consequences to Current s50.59 Parameters l
l 6.4.2 l Reduction in Margin with Control inputs l
ll6.4.3 ll Delete " Margin of Safety" as a Criterion l
l6.4.4 ll Define Margins with Safety and Regulatory Limits l
l6.4.5 l Define Margins with Fission Product Barrhrs - Definition l l6.4.6 l Define Margins with Specified Parameters l
l 6.4.7 ll Define Margins with Mitigation Capability l
ll6.4.8 l Define Margins with No Reduction l
l6.4.9 ll Define Margins with Minimal Increase l
l l 6.4.10 Define Margins with % Reduction Between Calculated and l
Acceptance Criteria l
l 6.4.11l NEI 96-07 Report l
l l 6.4.12 ll Replace Parameters with Regulatory Guide 1.174 Principles l
.+
l 6.4.13 l Frequency-Consequence Curves l
6.4.14 Replace Parameters with Modified Regulatory Guide 1.174 l
Principles i o
o, (50.59 Risk-Informed Option, Options Table 6.1-1 Scope and Parameter Options 6.4.15 Risk Increase Interval (RAW) i 6.4.16 Maintenance Rule Metrics 6.4.17 Core Damage Frequency 6.4.18 Large Early Release Frequency 6.2 Evaluation Process
)
When identifying options to change the QS0.59 process, a variety of possible approaches were considered as demonstrated in Table 6. L Each approach (option) listed in Table 6.1 contains elements addressing either the scope of the Q50.59 process as well as the use of parameters for making the Q50.59 process more risk-informed. " Options" for scope and " options" for parameters are the subject of the evaluations discussed in Sections 6.3 and 6.4 of this document.
Based on the desired intent of making 50.59 risk-informed, possible options which did not make use of either risk insights or, more specifically, qualitative or quantitative risk-informed parameters, were screened from further evaluation. Thus, for instance, the "no change" scope approach is not covered in this review.
j Furthermore, possible policy-related or other non-risk related wording changes being considered for l50.59 are not the subject of this review. In short, the options ev'aluated for how to change the scope or which parameters might be used, had to include the use ofrisk-related information and implementation ofthe option had to be consistent with the guidmg pnnciples summarized in Section 3 of this document.
l l
l As a result, only the risk-informed SAR and risk-informed facility options are addressed but not evaluated in this document. The evaluation of possible parameters to be used in the Q50.59 process was limited to investigating risk-informed parameters based on information presented in SECY-98 171, NEl 96-07, as well as illustrative quantitative risk parameters consistent with the principles in RG 1.174.
As will be described in the following sections, three risk-informed options are evaluated for how to decide whether a licensee proposed change would fall within the scope of $50.59. Additionally, a number of parameters based on information presented in the SECY and NEI documents mentioned above are evaluated with regard to their potential use in deciding whether the proposed change would require NRC review in a risk-informed 50.59 process. Both qualitative and quantitative risk metrics (parameters) are evaluated as possible options.
Following a brief description of each option, the evaluations consist of characterizing the scope and individual parameter options by the evaluation factors presented in Section 4 ofthis document. An additional evaluation is also performed considering the use of that option on the test case presented in Section 5 and the examples discussed in Section 2. For each option, the result for each evaluation factor and for its use on the test case and examples are presented and a summary observation is provided that highlights the findings of the evaluation ofeach option. Ultimately, the purpose ofthese evaluations is to identify the feasibility issues, the resource / cost and timing effects, as well as the portion (s) of Q50.59 affected, for each option; and demonstrate the use of that option using the test case.
If an option was purely deterministic in nature it did not go through the evaluation process (i.e., assessed against the factors) and not compared with the examples or test case.
i '
i m
. ~. -.
1 j
$50.59 Risk-informed Option. Options J
]
Table 6.2-1 provides a summary of the evaluation factors and the associated possible ratings. This table will be used in the evaluation ofeach option. Severalof the functional evaluation factors (e.g., Risk implications, Regulatory implications, Licensee Implications) have resource and time implications. How resource and i
time impacts these factors is described below.
I l
l Table 6.2-1 Evaluation Factors and Possible Ratings 1
(
l RESOURCE
- TIAfE*
EVALUATION FACTOR RATING
Enhanced Safety Decisions no/yes i
1 More Eincient Use of Agency s
no same/yes Feduction of Unnecessary Burden no/same/yes FACILITY COVERAGE same/less IMPLICATIONS RISK IMPLICATIONS Conndence Required from Risk high/ low Results PRA scope com/ full /mcimm Completeness level of detail detaile& simple Deterministic vs Risk. Informed det/det4t/n REGULATORY IMPLICATIONS Regulatory Guidance Needed yes no min / mod' sign shoremediang s
impact on Other Regulations yes/no mm/ mod / sip shoremediong New Rulemaking Needed yes/no mmimod/ sip short/mediong impact on inspectionfrraining yes/no min /mo& sign more/same/ fewer short/mediong morWsameSess LICENSEE IMPL ICATIONS Internal Procedures Needed yes/no min / mod / sign shoremediang impact on PRA Activi*ies yes/no min /mo& sign more/same/ fewer shon/mediong more/same4ess impact on Training yevno min /mo& sign short/medions impact on Inspection yevno more/same/ fewer more/same/less impact on Facility Changes yes/no more/same/ fewer more/sameless
- See Section 4 for a detailed description of the evaluation factors and ratings.
FRA Policy Statement Implications: This factor solely examines whether the option is addressing the objectives of the Commission's PRA Policy Statement. The ratings considered are, therefore, "no," "same,"
or "yes." The resource and time implications in meeting these objectives are addressed in the regulatory and licensee implications factors.
~ 27 -
(50.59 Risk-Informed Option. Options Facility Coretage Implications: This factor solely examines if the option is addressing the same scope (i.e.,
entire facility) as the current 50.59 regulation. The ratings considered are, therefore, "same" or "less."
There are no resource and time implications associated with this factor.
RisA Implications: This factor examines how much and what quality ofrisk analysis is needed to support the I
option. The examination addresses the con 0dence required from the PRA results with possible ratings of either "high" or " low;" the completeness needed for the PRA with scope ratings of " complete," " full,"
" medium" or " minimal," and level of detail ratings of" detailed" or " simple;" and how much are risk insights integrated into the decision making process with ratings of " deterministic," " deterministic / risk-informed,"
or " risk-informed." The resource and time implications associated with this factor are addressed in the regulatory and licensee implications factors.
Regulatory Implications: This factor examines the impact of the option on the regulatory process. There are both short term and long-term implications. To implement the option, make it part of the day-to-day operation (i.e., short term. implications), the need for regulatory guidance, the impact on other regulations, the need for Rulemaking, the impact on inspection (e.g., development ofinspection modules) and training are examined considering the resources required and the time needed. In looking at the resource and time implications, the ratings considered are " minimal," " moderate" or "signi0 cant" and "short," " medium" or "long," respectively. Once the option is pan of the day-to-day cperation, how the regulatory process could be impacted (i.e., long term implications) is also examined. The extent to which inspection is impacted, considering both resources and time, is addressed. In looking at the resource and time implications, the ratings considered are "more," "same" or " fewer" and "more," "same" or "less," respectively.
Licensee Implications: This factor examines the impact of the option on the licensee. There are both short term and long-term implications. To implement the option, make it part of the day-to-day operation (i.e.,
short term implications), the need for internal procedures, the impact on licensee's PRA activities, the impact on training are examined considering the resources required and the time needed. In looking at the resource and time implications, the ratings considered are " minimal," " moderate" or "significant" and "short,"
"rnedium" or "long," respectively. Once the option is part of the day-to-day operation, how the licensee's activities could be impacted (i.e., long term implications) is also examined. The impact on the support the licensee provides inspections and the impact on the number of facility changes requiring NRC approval, considering both resources and time, are addressed. In looking at the resource and time implications, the ratings considered are "more," "same" or " fewer" and "more," "same" or "less," respectively.
6.3 Scope Change Options Three options to modify the scope of $50.59 are identined, described and evaluated in this section. These three scope options include:
Limit the scope ;to Chapter 15 events Limit the scope to SAR risk signiGcant changes e
Expand the scope to risk significant changes (not limited to the SAR)
6.3.1 Option
Limit the Scope to Chapter 15 Events Description The existing 50.59 regulation deGnes scope as " changes in thefacility as describedin the safety analysis report." This option would revise the existing wording in Q50.59(1)(i)(ii) and (iii),"..as described in the safety analysis report.. " by replacing with "..as described in Chapter 15 of the safety analysis report.. "
(50 59 Risk-informed Option, Options l
Evaluation l
Reducing the scope from " changes in thefacility as described in the SAR" to " changes in thefacility as l
described in Chapter 15 of the SAR," a subset of the SAR, will not lead to enhanced safety decisions, l
irrespective of the use of PRA insights (for this case PRA insights are not used) beyond decisions currently made. This is attributed to potential changes to the facility not described in Chapter 15 that can negatively l
impact plant risk and not receive NRC review prior to implementation. Agency resources will be made more l
efficient in that less time will be required to process 50.59 changes, thereby focusing limited resources on l
more risk-significant activities. For the licensee, unnecessary burden to perform licensing evaluations of a proposed change is reduced since the scope is being limited to a subset of the SAR. The risk implication could be higher if the licensee implements changes by applying improper practices to activities described in the SAR, described outside Chapter 15. Since PRA is not used in the decision-making process, the issue of confidence and completeness required from the PRA is not applicable to this option.
1 i
The regulaton implications for this option includes the development of a regulatog guide that addresses the i
reduced scope of the new 50.59 process. While this appears to be a trivial change, the assumption made is that even if the existing 50.59 process was not modified, there still exists uncertainty within the industry on what constitutes an acceptable method for implementing the 50.59 rule. Therefore, the staff believes that a regulatory guide would still be useful. For this option, no other regulations are impacted. However, a rulemaking initiative is needed to reduce the scope from the SAR to information described in Chapter 15.
This rulemaking could be completed within one year, since it involves a relaxation and is not considered a backfit. In addition, since the inspectors are familiar with the existing regulatory process and the only change to the process involves reducing the scope to changes impacting only Chapter 15 of the SAR, short-term impact on inspectors would be minimal and long-term impact on inspectors would remain the same as for the existing process.
l Impacts on the licensee include, short-term resources needed to modify internal procedures and training l
programs for performing 50.59 assessments, and long-term reduction in resources for implementing changes that no longer require NRC approval. No change in resources required to support NRC inspections is expected under this option since the inspection process would be similar in nature. It is expected that the licensee will implement more changes to its facility under this option.
Table 6.3.1-1 summarizes the results from the evaluation of this option against the evaluation criteria described in Section 4.0.
l l
l Table 63.1-1 Evaluation Results for : Limit the Scope to Chapter j
15 Events l
RESOURCE
- TIME
- EVALUATION FACTOR RA TING
Enhanced Safety Decisions no More Emcient Use of Agency yes l
Reduction of Unnecessary Durden yes l
l,
I 1
I a
(50 59 Risk-informed Option. ( >ptions Table 6.3.1-1 Evaluation Results for : Limit the Scope to Chapter 15 Events RESOURCE
- TI.%IE
- EVALUA TION FACTOR R.-t TIA G
- ll h
RISK l%1P1.lCATIONS Confidence Required from Risk NA Results PRA scope NA Completeness lesel of detail NA Deterministic n Risk-Informed devemum nis REGl' TAI OR Y 1%1Pl.lCA'ilONS Regulatory Guidance Needed
>cs mimmum shon
-g Impact on Other Regulations na N4 NA e
-9.
New Rulemaking Needed sei mmimum medmm impact on inspection /Traming s es mmemum mmimum shon
.ame LICENSEE I%1Pl.ICAllONS Internal Procedu:es Needed yes mmimum shon impact on PRA Actisitses NA NA NA N'4 NA impact on Trammg l
ic.
mmimum shon Impact on inspection na same same sw.. -
Impact on facilits Changes y es
_ _ __ _ i -
less less Evaluation Using Test Case The scope under the current %50.59 process includes everything described in the SAR. This option limits the scope to Chapter 15 descriptions only. The service water booster pump is described in Chapter 15 and therefore would not be screened out by reducing the scope to issues described in Chapter 15. Therefore, the licensee would have to evaluate the change by each parameter; and consequently, the evaluation of the proposed facility changes would still result in an USQ needing NRC approval.
Evaluation ofthe Exasspies Describedin Section 2.3 Under the current 50.59 process, the scope includes all changes described in the SAR. This option limits the scope to Chapter 15 changes. Each of the following five examples are described in Chapter 15 and therefore falls under the scope of 50.59 and needs to be evaluated for a potential USQ. Consequently, evaluation of the proposed facility change would still results in a USQ requiring NRC approval.
. 3 0 --
~
- ~ -. -,,
?
s
{50.59 Risk-Informed Option. Options Example 2.3.1.1 Error in Dose Calculations for the Process Gas system Example 2.3.1.2 Error in Air Volume of Secondary Containment Example 2.3.1.3 Calculational Error in Head Loss of Emergency Core Cooling Suction
~
strainers Example 2.3.1.4 Modification of Main Feedwater Valves Example 2.3.1.5 Discrepancy Between As-Built system and UFSAR Description e
The following three examples were evaluated under the existing 50.59 process as not resulting in an USQ.
This option would not have altered those findings.
Example 2.3.2.1 Residual Heat Removal Service Water Pump Discharge Check Valves Replacement Example 2.3.2.2 Manual Isolation of the Control Rod Drive Pumps During a Station Blackout or Loss of Offsite Power Procedure Change Example 2.3.2.3 RHR Low Pressure Core Injection Inboard Injection Valve Design Change Observation Limiting the scope from the entire SAR to a subset of the SAR (Chapter 15) provides a licensee greater freedom to make changes to its facility without the need to obtain NRC approval. This option does not incorporate risk insights provided by PRAs and does not consider risk-significant and operational activities beyond Chapter 15 events. Therefore, plant risk can increase if the licensee makes improper changes to its facility as described in the SAR and screened out as not requiring NRC approval because it is not described in Chapter 15. This option does not impact the need to perform parameter evaluations for the test case and example problems.
6.3.2 Option
Lirnit the Scope to SAR Risk-Significant Changes i
l Description Narrow the scope of the existing 50.59 rule to a subset of the SAR that is risk-significant as provided by insights gained from a PRA. In this option risk-insights from a PRA are used to screen out proposed changes that are not risk significant and therefore do not require prior NRC approval to implement. This option would revise the existing wording in 50.59, "... changes...as describedin the safety analysis report...," by replacing l
the words with ".. changes...as described in the safety analysis report which operating experience or l
probabilistic risk assessment has shown not to be sigmficant to public health and safety. " This wording 'is l
_ consistent with those used in Criterion 4 of 950.36, Technical Specifications. Risk-significant can be defined as:
l an SSC or activity that either can challenge the plant or given a challenge is needed to mitigate the i
consequence of the challenge; and its contribution to risk is less than 10%.
e l
l Evaluation J
This option discriminates on risk insights when assessing changes requiring prior NRC review and approval.
+
I For non-risk-significant changes, NRC revieiv and approval would not be required prior to the licensee implementing the change. Based on insights provided by South Texas Project, significant reduction of
' unnecessary' burden to the licensee can be achieved ifnon-risk-significant changes can be screened out from the scope of QS0.59. Safety decisions are enhanced in that the licensee will more directly focus its attention.
+r--
{50 59 Risk-informed Option, Options on changes that impact plant risk. The efficiency of NRC resources will increase as limited resources would focus on review of risk-significant changes and away from non-risk-significant changes. The facility coverage will remain unchanged, in that all activities described in the SAR will be reviewed for their safety implication prior to implementing a change.
To implement this option, the NRC will require high confidence in the licensee's determination on risk-significance. Applying a complete scope and detailed PRA, a licensee can screen out non-risk-significant changes at the onset of this program. Alternatively, a licensee could identify the risk-significance ofa change on a case-by-case basis. Both options have their pros and cons that a licensee can assess to its needs. The latter option may not require the level of completeness and detail as needed for initial screening of all or most of the plant changes.
To implement this option, a regulatory guide and rulemaking initiative would be needed. The short term resources would be moderate to significant, depending on the level of detail the guidance is to provide. A significant up-front expenditure of resources would provide significant long term gains.
Similarly, licensees implementing this option would need to modify their internal procedures, obtain NRC acceptance of their PRA (i.e., meeting acceptable standards), train their personnel on the proper implementation of the new process, and provide similar support to NRC inspections (as needed under the present process). On the longer term, licensee's will be able to make significantly more changes to their plant without the need to obtain prior NP.C spproval.
Table 6.3.2-1 summarizes the results from the evaluation of this option against the evaluation criteria described in Section 4.0 Table 6.3.2-1 Evaluation Factors for Limiting the Scope to Risk-Significant Changes to information Described in the SAR l
CE*
TihlE*
EVALUA TION FACTOR RA TING
- h T
LT ST LT PRA POLICY STATEMLNT IMPLICATIONS Enhaaced Safety Decisions yes More Emcient Use of Agency yes Reduction of Unnecessary Burden yes FACILITY COVERAGE same IMPLICATIONS RtSK Ilt!PLICATIONS Confidence Required from Risk high l
Results PRA scope complete Completeness level of detail detailed Deterministic vs Risk-Informed deic REGULAlORY IMPLICATIONS
, 1 1
l f 50.59 Risk-Informed Option. Opuons Table 6.3 2-1 Evaluation Factors for Limiting the Scope to Risk-Signincant Changes to information Described in the SAR l
l RESOURCE
- Tlh1E*
- l ST LT ST LT Regulatory Guidance Needed ses moderaie medmm Impact on Other Regulations yes sign long New Rulemaking Needed yes sign long j
Impact on Inspection / training yes moderate more medmm less LICENSEE IMPLICATIONS Internal Procedures Needed yes modernie medium impact on PRA Activities yes sigmricani same long same Impact on Training yes moderaie medmm Impact on Inspection yes same same impact on Facility Changes yes less less
- See Section 4 for a detailed description of the es aluation factors and ratings Evaluation Using Test Case Under the current 50.59 process. the essential cooling water screen wash booster pump is described in the SAR and would require a parameter evaluation. Under this option, the booster pump would be categorized as non-risk-significant and would not require a detailed parameter evaluation. However, the declassi0 cation of the pump from safety grade to industrial grade would require a licensing submittal as required by other pans of 10 CFR 50. In order for the licensee to reclassify the pump without NRC approval, other parts of the regulations would have to be modified.
Evaluation ofthe Examples Described in Section 2.3 Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System This test case identified a non-risk-insigni6 cant error in calculating dose values for the process gas system rupture, as described in Chapter 15 of the SAR. Applying a risk-informed scoping criteria as addressed above, the licensee would be able to implement the change without performing detailed parameter evaluations.
Example 2.3.1.2 Error in Air Volume ofSecondary Containment t
Under the current 50.59 process, the scope includes all changes to SSCs or procedures described in SAR.
{
The volume in the secondary containment is described in the SAR and, under the current regulation, would have to be evaluated by each parameter. This option limits the scope to risk-dgni6 cant changes and, for this example, would permit the licensee to implement the change without obtaining NRC approval. Modi 6 cation to the technical speciHcations would not be needed to offset the increase in calculated dose, llowever, if the
(
licensee elected to modify its technical speciGcation on the charcoal efficiency, NRC approval would be required.
-- 3 3 -
?
I l
l
{50.59 Risk. Informed Option. Options Example 2.3.L3 CalculationalErrorin HeadLoss ofEmergency Core Cooling Suction Strainers Under the current {50.59 process, the scope includes all changes to SSCs or procedures described in the SAR.
The design and operation of the emergency core cooling system is described in the SAR and, under the current regulation, the change must be evaluated by each parameter. Under this option, changes that are
[
l not risk-insignificant would be screened as not requiring a parameter assessment. For this example, the increase in pressure drop across the ECCS suction strainers could render the emergency coolant systems inoperable due to vapor lock at the pumps. That is a risk-significant change that would still require NRC approval. To ensure adequate NPSH, the licensee requested to change its technical specifications to lower l
the water temperature in the suppression changer and ultimate heat sink. These changes would still require NRC approval. This option would categorize the example change as risk-significant and, as the current regulation, require a parameter evaluation.
Example 2.3.L4 Modification ofMain Feedwater Valves Under the current 50.59 process, the scope includes all changes of SSCs and or procedures described in SAR. The modification of the main feedwater valves are described in the SAR and, under current requirements, fall within the scope of the $50.59 regulation, thereby requiring a detailed parameter l
assessment. Under this option, only risk-significant changes would fall within the 50.59 scope. The
)
modification of the feedwater valves would not be considered risk-significant and would be screened out by this option.
Example 2.3.L5 niscrepancy Between As-Built System and UFSAR Description Under the current 50.59 process, the scope includes all changes of SSCs or procedures described in SAR.
l The main condenser vacuum pump is described in the SAR and therefore the change must be evaluated by each parameter. The change was not risk-significant since it was previously approved by the NRC and found l
acceptable. Under this option, updating the S AR with the as-built configuration would not require prior NRC approval.
The following three examples were evaluated under the existing Q50.59 process as not resulting in an USQ.
Example 2.3.2.1 Residual Heat Removal Service Water Pump Discharge Check Valves Replacement Example 2.3.2.2 Manual Isolation of the Control Rod Drive Pumps During a Station Blackout or Loss of Offsite Power Procedure Change Example 2.3.2.3 RiiR Low Pressure Core injection Inboard Injection Valve Design Change o
This option would not have changed the USQ assessment and would have screened the options from requiring a parameter evaluation because they were not-risk significant.
Observation This option is risk-informed only as it relates to applying risk insights from a PRA. It neglects other facility activities described in the SAR that may be non-risk significant but impact operating limits designed to meet other regulatory requirements, such as Parts 20 and 100, that limit radiation exposures to levels well below i
impacting public health and safety. These limits were established to erisure, for example, that anticipated
}
operational occurrences do not result in breaching the first burier to fission product releases (e.g., fuel design limits). However, this option does provide a good risk screening discriminator of changes that are not risk i
significant and could reduce unnecessary burden to the licensee and the NRC..
f 50.59 Risk-informed Option. Options
6.3.3 Option
Expand the Scope to Risk-Significant Changes Description The existing 50.59 regulation defines scope as changes of SSCs or procedures described in the SAR. This option expands the scope to all SSC and procedural changes that are risk-significant within the facility. This can include SSCs and procedures beyond those described in the SAR. This option would revise the existing wording in 50.59, ".hhanges...as described in safety analysis report... " by replacing with
" changes....which operating experience orprobabilistic risk assessment has shown not to be sigmficant to public health andsafety " This wording is consistent with those used in Criterion 4 of 50.36, Technical Specifications. Risk-significant can be defined as:
an SSC or activity that either can challenge the plant or given a challenge is needed to mitigate the consequence of the challenge; and its contribution to risk is less than 10%.
e Evaluation Expanding the scope from S SCs and procedures described in the SAR to SSCs and procedures in facility does not significantly alter the evaluations for option 6.3.2 (maintaining the scope within the SAR and incorporating risk insights as a discriminator). Safety decisions are enhanced in that SSCs and procedures not described in the S AR could be risk-significant and captured by this option. Limited industry and age.ncy resources would be utilized more efficiently in that non-risk significant changes could be implemented in a more timely manner with significantly less oversight by the NRC. Risk-significant changes would receive more NRC resources for review than presently expanded, This would be more significant if the agency reduces in FTEs.
As with the previous option, up-front investment in a detailed and full scope PRA would lead to a more stable
.egulatory environment as both risk and non-risk significant SSCs and procedures are identified early in the process. Long term efficiency would increase as a stable regulatory process is established.
While this option is expected to require some additional resources than the previous option, it is anticipated that the increased regulatory stability would provide licensees greater flexibility to implement more changes without NRC approval. Impact on PRA activities, internal procedures, rulemaking initiatives, regulatory guide initiatives, etc., are anticipated to be similar to that required to implement the above option, since most PRAs already cover SSCs and procedures beyond the scope of their associated SAR.
Table 6.3.3-1 summarizes the results from the evaluation of this option against the evaluation criteria described in Section 4.0 l
l Table 6.3.3-1 Evaluation Factors for Expanded Scope to Risk-Significant l
Changes Throughout the Facility l
RESOURCE
- TIME
- EVALUATIONFACTOR RA TING
.~
-. - -. - - ~.. _,.
- ~ -..-.
- ~ - - -. - -... _..
(50 59 Rkk-informed Op: ion (1pt:orr Table 633-1 Es aluation I actors for Expanded Scope to Risk-Significant Changes Ihroughout the Facility resol RCE*
TUIE*
El'ALUA TION l'A CTOR R 4 TI%
j. [
f l_nhanced Safety Decisions t es
,,9 g Lw....7 ' 'V4 '....
More Efficient Use of Agene)
>es i
~
^
~n.
_w.
Reduction of Unnecessar) 13urden
> cs
.,.,.y g.,.
..w..,.
- j. d .I U 2[ -
FACILII Y COVEllAGE more 14tPLICAllONS yg
}
Risk 1%IPLICA IIONS Confidence Required from Rhk high
(
f' ljif' l
Results h..,-D
'.M n..,
r
- sysE%a y,@.k.
l
.r...-s-
~
PRA scope umpierc w
Completeness MOME E*~'
... - -. a L.
.. l.
((kv hr.hh f
.'}._ff les el of detait detaded h[?[hkk
- E$;b bk_ h_
Determinista u Risk. Informed ort n i
HFGL L %10R) 1%IPI ICAllONS k~.v.
w,w Regulatory Guidance Needed yet m@ rate medium
_ _ _ J
,h'( _
1mpact on Other Regulations
> es urn kmg l
[
New Rulemakmg Needed yet ugn kmg
( medmm impact on inspection Trammg es mMace m re less
.. a a
1.lCENSEL 1%It'LICAllONS
-r-m I
internal Procedures Seeded ves moderate medmm
-- s-q l l
Impact on PRA Actis stics s es urmrmam same kmg same h#g medmm l
impact on Irairieng yes moderaie k
Impact on Inspection
, e.
same same hNIUkk impact on I acilits Changes les*
less ses
- See Section 4 for a detailed description of the esalu.ition factors and ratmgs
&aluation Using Test Case Under the current {50.59 process, the essential cooling water screen wash booster pump is described in the SAR and would require a parameter evaluation. Under this option, the booster pump would be categorized as non-risk-significant and would not require a detailed parameter es aluation. However, the declassification i
of the pump from safety grade to industrial grade uould require a licensing submittal as required by other j
parts of 10 CFR 50 in order for the licensee to reclassify the pump without NRC approval, other pans of the regulations would have to be modified.
. 36 -
- ~~.--
s i
1 f 50 59 Rirk. Informed Option. Options Evaluation of the Examples Describedin Section 2.3
)
Exampic 2.3.1.1 Error in Dose Calculationsfor the Process Gas System This test case identified a non-risk-insignificant error in calculating dose values for the process gas system
{
rupture, as described in Chapter 15 of the SAR. Applying a risk-informed scoping criteria, the licensee j
would be able to implement the change without the need to perform detailed parameter evaluations.
I Example 2.3.1.2 Error in Air Volume ofSecondary Containment Under the current Q50.59 process, the scope includes all changes to SSCs or procedures described in SAR.
The volume in the secondary containment is described in the SAR and, under the current regulation, would have to be evaluated by each parameter. This option limits the scope to risk-significant changes and, for this l
example, would permit the licensee to implement the change without obtaining NRC approval. Modification I
i to the technical specifications would not be needed to offset the increase in calculated dose. However, if the licensee elected to modify its technical specification on the charcoal efficiency, NRC approval would be required.
Example 2.3.1.3 Calculational Error in ilead Loss ofEmergency Core Cooling Suction Strainers Under the current G 50.59 process, the scope includes all changes to SSCs or procedures described in the S AR.
The design and operation of the emergency core cooling system is described in the SAR and, under the current regulation, the change must be evaluated by each parameter. Under this option, changes that are not risk-insignificant would be screened as not requiring a parameter assessment. For this example, the increase in pressure drop across the ECCS suction strainers could render the emergency coolant systems inoperable due to vapor lock at the pumps. That is a risk-significant change that would still require NRC approval. To ensure adequate NPSH, the licensee requested to change its technical specifications to lower the water temperature in the suppression changer and ultimate heat sink. These changes would still require NRC approval. This option would categorize the example change as risk-significant and, as the current regulation, require a parameter evaluation.
Example 2.3.1.4 Modification ofMain Feedwater Valves Under the current f 50.59 process, the scope includes all changes of SSCs and or procedures described in SAR. The modification of the main feedwater valves are described in the SAR and, under current requirements, fall within the scope of the {$0.59 regulation, thereby requiring a detailed parameter assessment. Under this option, only risk-significant changes would fall within the f 50.59 scope. The i
modification of the feedwater valves would not be considered risk-significant and would be screened out by this option.
Example 2.3.1.5 Discrepancy Between As-Built System and UFSAR Description l
Under the current Q50.59 process, the scope includes all changes of SSCs or procedures described in SAR.
l The main condenser vacuum pump is described in the SAR ar.d therefore the change must be evaluated by I
cach parameter. The change was not risk-significant since it was previously approved by the NRC and found l
acceptable. Under this option, updating the SAR with the as-built configuration would not require prior NRC l
approval.
The following three examples were evaluated under the existing Q50.59 process as not resulting in an USQ. -
. ~ ~ -
.__ - - - ~ ~ - - _-
1 l
Q$0.59 Risk-Informed Option. Options l
Example 2.3.2.1 Residual Heat Removal Service Water Pump Discharge Check Valves Replacement
=
Example 2.3.2.2 Manual Isolation of the Control Rod Drive Pumps During a Station Blackout or Loss of Offsite Power Procedure Change Example 2.3.2.3 RHR Low Pressure Core injection inboard injection Valve Design Change This option would not have changed the USQ assessment and would have screened the options from requiring a parameter evaluation because they were not-risk significant.
1 Observation This option is risk-informed only as it relates to applying risk insights from a PRA. It neglects other facility activities described in the SAR that may be non-risk significant but impact operating limits designed to meet i
other regulatory requirements, such as Parts 20 and 100, that limit radiation exposures to levels well below
{
impacting public health and safety. These limits were established to ensure, for example, that anticipated operational occurrences do not result in breaching the first barrier to fission product releases (e.g., fuel design limits). However, this option does provide a good risk screening discriminator of changes that could reduce unnecessary burden to the licensee and the NRC. This option also provided enhanced regulatory stability in that issues raised regarding risk-importance of SSCs and procedures within the facility are clearly identified. The PRA risk insights discriminate non-risk significant activities from the 50.59 and supports the Commission's policy statements on Safety Goals and Severe Accidents. However, expanding the scope throughout the facility would constitute a backfit, unless it is shown that the benefits from burden reduction overcompensates any additional requirements on the licensees.
6.4 Parameter Change Options in the following subsections, descriptions and evaluations are presented ofoptions consisting of the use of l
different qualitative parameters based on information in SECY-98-171, NEI 96-07, and illustrative quantitative parameters consistent with the principles in RG 1.174. For each parameter option, an assessment is made regarding the impacts ofimplementing the parameter option using the evaluation factors presented in Section 4 of this document. To perform these evaluations, judgments are required regarding the impacts on inspector time and training, the extent to which PRA information would be utilized, effects on current l
regulations, the level of resources required to implement use of the parameters, and other factors. The net result is to estimate the impacts ofimplementing the use of each parameter option in a risk-infonned {50.59 process.
i Following the Section 4 individual factor results, another general evaluation is provided considering the use of that parameter option for the test case provided in Section 5. Summary observations are then provided, highlighting the findings of the evaluation for each option and the test case. The focus of these evaluations is to provide decision-makers with the impacts of implementing each parameter option, including demonstration ofits potential use for a test case.
6.4.1 Option
Minimal Increase in Probability and Consequences to Current {50.59 Parameters Description This option is a variation ofone of the options identified in SECY-98-171 in that it would revise the existing paragraph f 50.59(a)(2) (i), "if the probability of occurrence or the consequences of an accident or malfunction of equipment important to safety previously evaluated in the safety analysis report may be increased, " by replacing "may be increased" with "would result in more than minimal increase, " as welI as l
J
$50.59 Risk-Infonned Option. Options incorporate insights from a PRA. Minimal increase is defined as:
for probability of occurrence of equipment malfunction-less than a factor of two a
for probability of occurrence of accident-less than 10%
for consequence of an accident-less than 10%
=
for consequence of equipment malfunction-less than 10%
Evaluation This option would enhance safety decisions by incorporating risk insights into the decision making process.
More efficient use of agency resources and a decrease in unnecessary burden on licensees is expected as PRA insights are used because of a more risk-informed decision making process. This option would not impact facility coverage implications and would still cover the entire facility consistent with the current 50.59 process. To incorporate this option high confidence is needed in the results of the PRA along with a complete scope risk analysis, to properly quantify the magnitude of the increase in probability and consequences. This option was fopnd to be mainly deterministic utilizing risk-informed insights to allow minimal increases in probability and consequences. To clarify this position a regulatory guide would be required to implement this option. No other regulations were found to be impacted. New rulemaking would be required which would entail more resources and time to implement. A moderate impact on the inspection staff to develop a new inspection modules and to properly train the inspection staff. The licensee would need to modify existing administrative procedures which would require additional resources over the short term. The PRA activities could be impacted significantly, as licensees would be required to enhance, maintain and update their PRA's. This would also require increased support by the PRA staff for training of the licensees staff and suppon ofinspection activities.
Table 6.4.1-1 presents the results from the evaluation of this option against the evaluation criteria described in Section 4.
Table 6.4.1-1 Evaluation Results for Minimal Increase in Probability and Consequences Option RESOURCE
- TIAIE*
EVALUA TION FACTOR RA TING
- l ST LT ST LT PRA POLICY STATEMENT IMPI,1 CATIONS Enhanced Safety Decisions Yes More Efficient Use of Agency Yes Reduction of Unnecessary Burden Yes FACILITY COVERAGE same IMPLICATIONS RISK IMPLICATIONS Confidence Required from Risk Egh Results PRA scope Complete Completeness level of detail Detailed Detenninistic vs Risk-Informed Det'R-1 d
-.39
i (50 59 Risk Informed Option. Options Table 6.4.1-1 Evaluation Results for Minimal Increase in Probability and Consequences Option RESOURCE
- TihtE*
- l ST LT ST LT REGt!LATORY 1%IPLICATIONS Regulatory Guidance Needed Yes Min Medium impact on Other Regulations No N/A N/A New Rulernakmg Needed Yes Mm Long Impact on Inspection / Training Yes Mm More short same LICENSEE IMPLICATIONS internal Procedures Needed Yes Min sh ert Irnpact on PRA Activities Yes signiricant More Medium More impact on Trainmg Yes Min Shmi Impact on Inspection increase Moee same l
Impact on iacilin Chanres Yes same same
- See Section 4 for a detailed description of the evaluation factors and ratings Evaluation Using Test Case Under the current 50.59 process, this change was determined to be a USQ because it increased the i
probability ofoccurrence ofa malfunction ofequipment important to safety previously evaluated in the SAR.
Under this option the parameter " minimal increases in probability ofoccurrence" would be allowed without NRC approval. Based on the information provided by the licensee in section 5, it appears that the probability of equipment malfunction will not increase above a factor of 2. Therefore, this option would allow the licensee to make the change without NRC approval.
Evaluation of the Examples Described in Section 2.3 Against SECY-98-171 Options As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the licensee, to be USQs and have required NRC approval in accordance with the current 50.59 process. The potential SECY-98-171 { 50.59 options were used to evaluate whether these same facility changes would still I
require NRC review and approval.
In using any potential 50.59 process, the change alone must be evaluated without consideration ofcorrective or compensatory measures the licensee will take to address the facility change. In many cases, the SECY-98-171 options did not alleviate unnecessary burden of requiring NRC approval for facility changes l
that had little or no "overall" risk significance once the licensee took corrective and or compensatory measures, as described in Section 2.3. Consideration should be given to change the 50.59 wording to allow licensee's to evaluate the facility change with any potential corrective or compensatory actions included.
This would allow the collective plant change to give an "overall" plant risk insight.
f 50.59 Risk-Informed Option, Options Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System Under the current l50.59 process, this facility change was determined to be a USQ because it increased the consequences of an accidna as a result of a 1 mrem dose increase from the UFSAR stated dose. Under this SECY-98-171 option, the dose change would be determined to be a " minimal" increase in the consequences i
of an accident and would not, therefore, require NRC approval.
Example 2.3.1.2 Errorin Air Volume ofSecondary Containment 1
Under the current 50.59 process, this facility change was determined to be a USQ because it resulted in a reduction in the margin of safety and an increased the consequences of an accident, due to operator thyroid dose increasing from 22 rem to 25.6 rem. Under this SECY-98-171 option, the dose change would be determined to be a " minimal" increase in the consequences of an accident; however, NRC approval would i
l still be required, due to a reduction in the margin of safety, s
Example 2.3.1.3 Calculational Error in Head Loss ofEmergency Core Cooling Suction Strainers l
Under the current Q50.59 process, this facility change was determined to be a USQ because the lack of enough NPSH could increase in the probability of ECCS equipment malfunction occurrence. Under this SECY-98-171 option, the loss of potential NPSH may not be considered " minimal", as demonstrated by a the TS changes to take additional credit for a nominal amount of containment pressure and to lower the allowable water temperature in the suppression pool. The facility change would still be required to receive NRC approval.
Example 2.3.1.4 Modification ofMain Feedwater Valves Under the current {50.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.lE-5 per demand). Under this SECY-98-171 option, the increase in the probability of equipment malfunction would be determined to be a " minimal" increase and would not, l
therefore, require NRC approval.
Example 2.3.1.S Discrepancy Between As-Built System and UFSAR Description Under the current Q50.59 process, this facility change was determined to be a USQ because the as-built I
system isolation malfunction was of a different type than that previously evaluated in the UFSAR. Under this l
SECY-98-171 option,the change would still require NRC approval (even though the as-built system isolation was previously approved by the NRC but never updated in the UFSAR) because the as-built isolation still creates a possibility for a malfunction of equipment with a different result previously evaluated in the UFSAR.
The following three examples were evaluated under the existing s50.59 process as not resulting in an USQ.
This options would not have altered that finding.
Example 2.3.2.1 Residual Heat Removal Service Water Pump Discharge Check Valves Replacement Example 2.3.2.2 Manual isolation of the Control Rod Drive Pumps During a Station Blackout or Loss of Offsite Power Procedure Change Example 2.3.2.3 RHR Low Pressure Core Injection Inboard Injection Valve Design Change e
Observations l
. 7 j_
{50.59 Risk-Informed Option, Options The existing Q50.59 process does not discriminate between risk significant and non-risk significant changes.
This option is deterministic / risk-informed and would, therefore, incorporate PRA insights in the decision-making process. Overall, this option would make more efficient use of agency resources and result in a i
licensee's reduction of unnecessary burden by reducing the amount of Q50.59 submittals for NRC review.
6.4.2 Option
Reduction in Margin with Control Inputs Description This option would revise the existing paragraph Q50.59(a)(2)(iii),"sythe margin ofsafety as definedin the basisfor any tecknicalspecifIcations is reduced, "by replacing th e existing wording"...basisfor any technical specification.." with ".. associated with any technicalspecification.. " This change redefines reduction in margin of safety. This option concludes that the analyses and infomiation in the SAR establishes the basis for the margins of safety for the TS. Thus, the NRC would redefine " reduction in margin ofsafery associated with any technical specification" to mean that the input assumptions, analytical methods, acceptance conditions, criteria and limits of the safety analyses, presented in the FSAR report (as updated),
that established any technical specification requirement, cannot be altered in a nonconservative manner without NRC approval.
Evaluation The existing Q50.59 process does not discriminate between risk-significant and non-risk significant changes.
This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
I
6.4.3 Option
Delete " Margin of Safety" as a Criterion Description This option would delete the existing paragraph Q50.59(a)(2)(iii),"ifthe margin ofsafety as definedin the basisfor any technicalspecifications is reduced. " This option deletes any criterion focusing upon margins.
Instead, the NRC would rely upon the other criteria in Q50.59, as well as the regulatory requirement that all changes to TS be reviewed and approved by the NRC, to assure that there are no significant adverse changes to margins in design and operation. The NRC would argue that there is no need for prior review ofchanges that do not satisfy any of the other evaluation criteria in view of " risk-informed" insights and greater understanding of the margins that exist through meeting the body of regulatory requirements.
Evaluation The existing Q50.59 process does not discriminate between risk-significant and non risk significant changes.
This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
6.4.4 Option
Define Margins with Safety and Regulatory Limits l
Description l
This option would revise the existing paragraph {50.59(a)(2)(iii). "ifthe margin ofsafety as definedin the l
basisfor any technicalspecifications is reduced," by redefining margin of safety with safety and regulatory limits. 'Illis option examines the results of the safety analyses to determine whether changes to operational.
f 50 59 Risk-Informed Option. Options characteristics or other information described in the FSAR (as updated) would reduce the level of protection afforded by the TS (i.e., by the limiting safety system settings and limiting conditions of operation), as reDected in the results of safety analyses.
I As part of the licensing review for a facility, the NRC established acceptance criteria (or regulatory limits) l l
for certain physical parameters, such as those that define the integrity of the fission product barriers (fuel l
cladding, reactor ecolant system boundary and containment). Satisfying these regulatory limits produces a margin of safety to loss of barrier integrity. The safety analyses presented in the FSAR (as updated) demonstrate that the response of the barriers to the postulated accidents, transients, and malfunctions meets the acceptance criteria. For certain of these parameters, TS safety limits have been established to reasonably protect the integrity of physical barriers that guard against the uncontrolled release of radioactivity.
l However, for other parameters, a licensee must determine the licensing basis of the parameter in question by j
reviewing the plant-specific safety analyses. The acceptance criterion is that value approved by the NRC for a particular parameter or process variable (e.g., ASME Code stress limits, a departure from nucleate boiling ratio limit or maximum critical power ratio limit or containment design pressure). These acceptance criteria may be stated in the SAR, may be in NRC regulations, or may be presented in the NRC Standard Review Plan.
The margins of safety that would be controlled by 50.59 process can be characterized in different ways.
Several variations of this option are discussed below and in Sections 6.4.5 through 6.4.10:
This option would deOne " margin of safety" as follows:
The " margin of safety as defined in any technical specification" (margin of safety) is the amount (quantitative or qualitative) of margin between the operation of the facility as described in the technical specifications and the accedence of safety limits listed in the technical specifications or other regulatory limits. In relation to accident analysis, the margin of safety is typically the difference between calculated parameters (e.g., peak fuel clad temperature and maximum RCS pressure) and the associated regulatory or safety limit. The margin of safety is a product of speciDe values and limits contained in the technical specifications (which cannot be changed without NRC approval) and other values, such as assumed accident or transient initial conditions or assumed safety system response times, which are not speci0cally contained in the technical speci0 cations. Any change to the values not specifically contained in the technical speci0 cations must be evaluated for impact on the margin between the calculated result of an accident or transient,and the safety or regulatory limit.
With this option, before changing operational characteristics described in the UFS AR (not directly controlled by TS), a safety evaluation must be performed to determine, among other things, if the change results in a reduction in the level of protection afforded by the TS [ margin of safety as defined in any TS]. Such a reduction would typically occur only if the operational characteristic had been used as a bounding condition in the analysis upon which the selection of TS was based, or in analysis where the acceptability of selected I
TS values was demonstrated. Licensees could make desired changes to operational characteristics without prior NRC approval, provided that the change does not result in accident analysis results that are nearer the regulatory, or safety limits than the corresponding results that the NRC used in evaluating the acceptability of the TS during licensing of the facility.
{$0.59 Risk-infonned Option, Options Evaluation The existing {50.59 process does not discriminate between risk-significant and non-risk significant changes.
This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
6.4.5 Option
Define Margins with Fission Product Barriers - Definition Description This option would revise the existing paragraph 650.59(a)(2)(iii), "ifthe margin ofsafety as definedin the 1
basisfor any technical specifications is reduced " by redefining margin of safety with fission products 1
barriers definition.10 CFR 50.36 has criteria for when TS are to be provided that specifically are tied to l
design basis accident or transient analysis that either assumes the failure of or presents a challenge to the integrity of a fission product barrier. Thus, the margin as defined in the basis for any TS can be reasonably viewed as that margin associated with preserving integrity of these barriers. This option considers a more i
explicit liakage to the response of the three fission product barriers generally relied upon to provide protection from uncontrolled release ofradioactive materials from a reactor facility. Under such a proposal, the text of the rule would explicitly state that it is the response of fission product barriers (fuel, reactor coolant j
system, and containment) to accidents, transients, and malfunctions that is being controlled.
The following could be given as a definition of margin of safety and of fission product barrier response.
Regulatory guidance would explicitly list the parameters (for PWRs and BWRs) that are to be controlled.
i The margin of safety for any fission product barrier response is the difference between the calculated l
value and its associated acceptance criteria.
Fission product barrier response means those parameters that must be satisfied in the event of postulated design basis events to demonstrate integrity of the fuel, reactor coolant system and containment system barriers.
The following parameters would be included: Fuel and cladding performance (peak cladding temperature, or energy deposition, DNBR or MCPR, oxidation), RCS performance (pressure, flows, stress), and containment performance (peak pressure, containment leakage).
Evaluation The existing Q50.59 process does not discriminate between risk-significant and non-risk significant changes.
This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
6.4.6 Option
Define Margins with Specified Parameters Description This option would revise the existing paragraph 50.59(a)(2)(iii), "ifthe margin ofsafety as definedin the basisfor any technical specifications is reduced,"by redefining margin of safety to include specific parameter ofinterest. This option actually lists the parameters ofinterest directly in the criterion for prior review, as for instance, the criterion could read:.
V l
{50.59 Risk-Informed Option, Options l
Result in a change to the FSAR (as updated) calculated value of RCS peak pressure, containment peak pressure, or fuel performance (DNBR/MCPR, others), etc.
1 i
This option has the advantage of being more precise, but the rule language would need to be crafted to account for various reactor types.
Evaluation The existing 50.59 process does not discriminate between risk-significant and non-risk significant changes.
This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
6.4.7 Option
Define Margins with Mitigation Capability Description This option would revise the existing paragraph 50.59(a)(2)(iii), "ifthe margin ofsafety as definedin the basisfor anytechnicalspecifications is reduced,"by redefining margin ofsafety by including performance parameters. This option would preserve the integrity of both prevention and mitigation capabilities available in the plant, and would include both features within the " margin" criterion. If this approach were adopted, the definition or the list of parameters would be supplemented with the performance parameters for the accident mitigation capability of the plant, as for instance, ECCS performance (pressures, flows, actuation values), engineered safety feature performance (flows, pressures, spray effectiveness, system efficiencies).
Evaluation The existing l50.59 process does not discriminate between risk-significant and non-risk significant changes.
This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
6.4.8 Option
Define Margins with No Reduction Description i
This option would revise the existing paragraph f 50.59(a)(2) (iii), "ifthe margin ofsafety as defined in the basisfor any technicalspecifications is reduced. "by not allowing a reduction in margin. This option would require that the safety analysis, considering the effect of the change, must show that the accident analysis results are not nearer to any safety or regulatory limit, thus, a "no reduction in margin" standard. Possible i
rule text:
Changes, or the net effect of multiple changes, which result in a reduction in the margin of safety require prior NRC approval. Changes, or the net effect of multiple changes, which do not cause a reduction in the margin of safety do not require prior NRC approval.
Evaluation The existing Q50.59 process does not discriminate between risk-significant and non-risk significant changes.
His option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples..
(50 59 Risk-informed Option. Options
6.4.9 Option
Define Margins with Minimal Increase Description This option would revise the existing paragraph 50.59(a)(2)(iii), "ifthe margin ofsafety as definedin the basisfor any technical specifications is reduced " by allowing minimal increases in margin. This option would allow licensees some flexibility in making changes, through development of a " minimal increase" standard. In considering margins, the NRC is thus weighing how such a concept could be applied. This option would require NRC npproval for a change, test, or experiment if the output values (calculated in the SAR) are altered by more than a minimal amount. The " margin" criterion would be modified to state that a change in calculated result of "more than a minimal amount" would require prior review and approval.
Either in the rule itself, or in guidance, the NRC would define " minimal amount", modeled upon the options offered for minimal increases in consequences. For example, there could be a fixed amount (percent change) i in margin. as long as regulatory limits are still met. Ifguidance itemizes the parameters, such guidance could also custo mize how " minimal" should bejudged for each particular parameter (allowing greater amounts for certain paameters depending on precision of calculations, sensitivity of results and other considerations).
For instance, the definition of" margin of safety reduction..." might be stated as follows:
l Reduction in margin of safety means that as a result of a change, the [ MARGIN] is altered in a nonconservative manner by more than a minimal amount.
Evaluation The existing 50.59 process does not discriminate between risk-significant and non-risk significant changes.
i This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
' 6.4.10 Option: Define Margins with % Reduction Between Calculated and Acceptance l
Criteria 1
i Description l
This option would revise the existing paragraph Q50.59(a)(2)(iii). "ifthe margin ofsafety as definedin the basisforany technicalspecifications is reduced,"by allowing a % reduction in margin. This option defines
" minimal" in the rule itselfin terms of the results or analyses for barrier response, with respect to meeting the acceptance criteria for those barriers. For example, rule language could read as follows:
l License amendment needed if as a result of a change :
there is more than a 10% reduction in the difference between the calculated value and the acceptance criteria for fission product barrier response to accidents evaluated in the SAR.
j If such an approach is followed, the NRC would propose to include a definition of acceptance criteria, such as follows:
Acceptance criteria are those values, established by NRC regulation or review guidance, to which the licensee is committed through its FSAR (as updated), as the basis for acceptability of response to the postulated accident, transient or malfunction..
r--
u
l l
l 650.59 Risk-Informed Option, Options Evaluation l-l The existing $50.59 process does not discriminate between risk significant and non-risk significant changes.
)
This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
i 6.4.11 Option: NEI 96-07 Report Description 1
This option was discussed by Nuclear Energy Institute in NEI 96-07, " Guidelines for 10 CFR 50.59 Safety Evaluations," Revision 0, dated September 1997. Under this option, an increase in consequences and reduction in margin would not be deemed to be an USQ unless the established acceptance limits were exceeded. In addition, this option would allow small increases in probability, consequences and margin. NEI 96-07 would not consider a change needed where the change in probability, consequences, or margin were so small or the uncertainties in determining whether a change in consequences, probability or margin has occurred are such that it cannot be concluded reasonably that it has actually changed. In such cases, the change need not be considered a reduction.
1 Evaluation The existing l50.59 process does not discriminate between risk-significant and non-risk significant changes.
i This option is deterministic and does not incorporate risk insights. Therefore, this option was not evaluated against the factors nor test case or examples.
6.4.12 Option: Replace Parameters with Regulatory Guide 1.174 Principles Description Under this option, the NRC would replace the existing parametric criteria in Q50.59 with the five principles discussed in RG 1.174. In particular, paragraph (a)(2) of {50.59 now reads:
A proposed change, test, or experiment shall be deemed to involve an unreviewedsafety question (i) of theprobability ofoccurrence or Ihe consequences ofan accident or malfunction ofequipment important I
to safetypreviously evaluated in the safety analysis report may be increased; or (ii) ofa possibilityfor an accident or malfunction ofa diferent t>pe ihan any evaluatedpreviously in the safety analysis report 1
may be created; or (iii) ofthe margin ofsafety as defined in the basisfor any technical specification is reduced.
1 This would be deleted and replaced with:
A proposed change, test, or experiment shall be deemed to require Commission approval (i) of the proposedchange does not meet the currert regulations, (ii) aftheproposedchange is not consistent with the defense-in-depth philosophy, (iii) ofsufficient safety margins are not maintained, (iv) of the core damagefrequency or risk increases are not small and consistent with the intent of the Commission 's Safety Goal Policy Statement, or (v) an appropriate performance monitoring strategy is not included.
Each of the five items in the above statement are discussed in the form of guiding principles in Section 2 of RG 1.174. Acceptance guidelines for items i, ii, iii, and v are discussed in RG 1.174. Implementation ofthis option is likely to require additional specific guidance on these items. Acceptance guidelines for item iv v
l
{ 50 59 RisL-Infonned Option. Options
)
l (CDF and LERF) are also provided in RG 1.174. However, RG 1.174 does not identify levels of risk or risk increase such that no NRC review is necessary. For the implementation of this option for 50.59, values of 1.0E-7/ year for an acceptable change in CDF and 1.0E-8/ year for an acceptable change in LERF are proposed such that NRC review would not be required. These numbers are sufficiently small that the current level of CDF and LERF at a plant would not be an important consideration.
The nature of this option is such that it focuses on accidents involving the reactor core, i.e., CDF and LERF are the primary risk measures. Consideration of radiological hazards outside the reactor core could either be based on principle i (meet current regulations) or through combination with another option.
Evaluation This option is consistent with the objectives of the PRA Policy Statement. This option enhances safety decision making by focusing NRC review on only those changes that impact one or more of the five principles of RG 1.174, thus using NRC resources more efficiently. Unnecessary burden on industry is reduced because additional changes that are relatively unimportant to safety and risk would be allowed without a submittal.
In general, less of the facility will be covered than under the current Q50.59. Portions of the plant that do not affect CDF or LERF, or that are not affected by any of the other principles of RG 1.174 would be excluded.
This will eliminate some SSCs or activities that are included in the SAR.
This option makes use of quantitative CDF and Risk results. Therefore, a high confidence is needed in the results for those changes that are compared to the quantitative criteria. The scope rating is stated as
" medium;" because that level is expected to be needed for a substantial fraction of the changes. That is, 1
external events or low power and shutdown may not need to be considered for many changes. For a particular change, the needed scope can range from minimum to full.
The level of detail must be commensurate with the change, although a uniform level of detail for all aspects of a PRA may not be necessary for a particular change. Since this option represents implementation of Regulatory Guide 1.174 and its five principals, and those principles are a combination ofdeterministic and risk-informed principles, this option is classified as deterministic / risk-informed.
New rulemaking and regulatory guidance will be needed to implement this option, especially in the area of acceptance limits for the four principals that are deterministic in nature. However, regulatory guidance is needed for the current {50.59, and some guidance is already available in RG 1.174. Therefore additional time and resources are not required beyond that expected for the other options. This option will have minimal impact on other regulations, since that is one of the principles of RG 1.174. Nevertheless, the inspectors will require training in methods to verify compliance. Once new inspection processes are in place, it is assumed that the total inspection time will remain the same, although a larger fraction of that time will be focused on PRA related activities.
The licensees will need to develop their own intemal procedures for implementing this option. This will involve additional resources and time for implementation in the beginning. The actual PRA activities will be impacted due to the need for a high confidence PRA with appropriate scope and level ofdetail. Significant time and resources will be needed at many plants to achieve this level of confidence. Ongoing maintenance ofthe PRA will also be necessary. At many plants, additional training will be necessary for the licensee staff.
The nature of interactions with inspectors will change to be more focused on PRA-related activities and this will impact needed licensee resources. The potential exists for a significant return on investment for the licensees if large numbers of cost-effective changes are allowed by this process. If the process works properly, less time and resources will be spent on each individual change, because the PRA results can be used to screen things out quickly.
{50.59 Risk-informed Option, Options i
Table 6.4.12-1 presents the results from the evaluation of this option against the evaluation criteria described i
m Section 4.
Table 6.4.12-1 Evaluation Resultsfor RG 1.174 Option RESOURCE TIME EVALUA TION FACTOR RA TING ST LT ST LT PRA POLICY STATEMENT IMPLICATIONS i
Enhanced Safety Decisions yes More Efficient Use of Agency yes Reduction of Unnecessary Burden yes FACILITY COVERAGE less IMPLICATIONS RISK IMPLICATIONS Confidence Required from Risk high Results PRA scope medium Completeness level of detail detailed Deterministic vs Risk-Informed det-ri REGULATORY IMPLICATIONS Regulatory Guidance Needed yes mod med Impact on Other Regulations no NA NA New RuleMaking Needed yes mod med Impact on inspection /Traming yes sign same long same LICENSEE IMPLICATIONS Internal Procedures Needed yes sign med Impact on PRA Activities yes sign more long more Impact on Training yes sign long impact on inspection yes same same impact on Facility Changes yes less less a
Evaluation Using Test Case Under the current Q50.59 process, this change was determined to be a USQ because it increased the probability ofoccurrence of a malfunction ofequipment important to safety previously evaluated in the SAR.
Under this proposed option, NRC review will still be required, because the components are designated as safety-related, and therefore the change impacts the other regulations. If this option were combined with options to change the scope or definition of what is considered safety related, then it is likely that this change _.
{50.59 Risk-informed Option. Options would not need review. The change does not affect safety margins or defense in depth. The licensee would have to determine that the change resulted in CDF and LERF increases less than the threshold values presented above and would have to implement a monitoring program as described in RG 1.174. The CDF and LERF results would need to address the complete set ofinitiating events.
Evaluation of the Examples Describedin Section 2.3 Against RG 1.174 Option As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the licensee, to be USQs and have required NRC approval in accordance with the current Q50.59 process. The potential RG 1.174 50.59 option was used to evaluate whether these same facility changes would still j
require NRC review and approval. This option did not alleviate the unnecessary burden for all of the l
examples, since technical specifications were impacted in some cases.
Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System Under the current 50.59 process, this facility change was determined to be a USQ because it increased the consequences of an accident because doses increased by I mrem from the UFSAR stated dose. Under this RG 1.174 option, the dose change would be acceptable according to each of the five principles and would not require NRC approval.
Example 2.3.1.2 Errorin Air Volume ofSecondary Containment i
Under the current 50.59 process, this facility change was determined to be a USQ because it was a reduction in the margin of safety and an increased the consequences of an accident, because the operator thyroid dose increased from 22 rem to 25.6 rem. Under this RG 1.174 option, the dose change would represent a change in technical specifications and would still require NRC approval.
Example 2.3.1.3 CalculationalError in lleadLoss ofEmergency Core Cooling Suction Strainers Under the current G50.59 process, this facility change was determined to be a USQ because the lack of enough NPSH could increase in the probability of ECCS equipment malfunction occurrence. Under this RG 1.174 option, changes to technical specifications are involved and the change would still require NRC approval.
Example 2.3.1.4 Modification ofMain Feedwater Valves Under the current Q50.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.l E 5 per demand). Under this RG 1.174 option, if the CDF did not increase more than IE-7, then NRC approval would not be required.
Example 2.3.1.5 Discrepancy Between As-Built System and UFSAR Description Under the current 50.59 process, this facility change was determined to be a USQ because the as-built system isolation malfunction was ofa different type than that previously evaluated in the UFSAR. Under this RG 1.174 option, the change would not require NRC approval as long as the licensee determined that the CDF and LERP impacts were less than IE-7 and IE-8. Respectively. -
._.._.m_
_ _ ~
f 50.59 Risk-Infonned Option, Options Example 2.3.1.6 Replace RHR Service Water Sump Pump Discharge Check Valve This facility change was not a USQ under the current {50.59 process and would also not be a USQ under the RG 1.174 option.
l Exangpie 2.3.1.7 Add ProceduralSteps to Manually Isolate the Control RodDrive Pumps During SB0 his facility change was not a USQ under the current f 50.59 process and would also not be a USQ under the I
RG 1.174 option.
1 Example 2.3.1.8 Adda 3/I6 " Diameter Hole Through the Reactor Recirculation Piping Inlet Side ofthe Valve Flex-Wedge Disc This facility change was not a USQ under the current G50.59 process and would also not be a USQ under the RG 1.174 option.
Observations This option is consistent with the intent of the PRA policy statement and is more risk-informed than the current 650.59. It requires both deterministic evaluations and a determination of risk. It requires significant resources by both the NRC and the licensees to develop and implement the process; however, once in place, the process can result in significant savings.
6.4.13 Option: Frequency-Consequence Curves l
Description This option uses Frequency-Consequence curves as a quantitative measure of the risk effects in a risk-l informed 50.59 process. The use of this option is described in the Attachment to the July 16,1998 letter from ACRS Chairman Seale to the chairman concerning proposed changes to 10 CFR 50.59. In particular, i
paragraph (a)(2) of 50.59 now reads:
i A proposed change, test, or experiment shall be deemedto involve an unreviewedsafety question (i) of theprobability ofoccurrence or Ihe consequences ofan accident or malfunction ofequipment important to safetypreviously evaluatedin the safety analysis report may be increased; or (ii) ofapossibilityfor an accident or malfunction ofa diferent type than any evaluatedpreviously in the safety analysis report may be created; or (iii) ofthe margin ofsafety as defined in the basisfor any technicalspecsfication is reduced His would be deleted and replaced with:
A proposed change, test, or experiment shall be deemed to require Commission approval ofthe risk. in terms offrequency times consequences, ofan accident is increased to be inconsistent with the intent of the Commission 's Safety Goal Policy.
This approach makes use of frequency-consequence curves (risk curves). The actual risk curves would be provided in a Reg Guide. PRAs would have to be expanded to include level 3 (consequences). Current PRAs deal primarily with Class 8 and 9 accidents. Additional risk information would be generated to cover the remaining accident classes in Chapter 15 of the SAR. This option would allow tradeoffs between the frequency and consequences of accidents.
- $1 -
t
~,
~,
y
-q l
f 50.59 Risk-Informed Option. Options Evaluation This option supports the objectives of the PRA Policy Statement with respect to being risk informed. By itself, this option does not address those aspects of the PRA Policy Statement dealing with defense in depth or other deterministic criteria. This option enhances safety decision making by focusing NRC review on only those changes that impact risk, thus using NRC resources more efficiently. Unnecessary burden on industry is reduced because additional changes that are relatively unimportant to safety and risk would be allowed without a submittal.
In general, less of the facility will be covered than under the current 50.59. Portions of the plant that do not affect the frequency or consequences of accidents would be excluded. This will eliminate some SSCs or activities that are included in the SAR.
This option makes use of quantitative Risk results. Therefore, a high confidence is needed in the results for those changes that are compared to the quantitative criteria. The scope rating is stated as " medium;" because that level is cipected to be needed for a substantial fraction of the changes. That is, external events or low power and shutdown may not need to be considered for many changes. For a particular change, the needed scope can range from minimum to full. The level ofdetail must be commensurate with the change, although a uniform level of detail for all aspects of a PRA may not be necessary for a particular change. Since this option does not address deterministic principles, it is classified as risk-informed.
New rulemaking and regulatory guidance will be needed to implement this option, especially in the area of acceptance limits for the frequency-consequence curves. Regulatory guidance is needed for the current 50.59; however, it is expected that the guidance for developing and implementing frequency-consequence curves for all accidents will be time consuming to develop. In considering some classes of accidents, the consequences will be onsite. A consistent set of frequency-consequence curves may be needed for all accident types. This option will have minimal impact on other regulations. Nevertheless, the inspectors wili require training in methods to verify compliance. Once new inspection processes are in place, it is assumed that the total inspection time will remain the same, although a larger fraction of that time will be focused on PRA-related activities.
The licensees will need to develop their own internal procedures for implementing this option. This will involve additional resources and time for implementation in'the beginning. The actual PRA activities will be impacted due to the need for a high confidence PRA with appropriate scope and level ofdetail. Significant time and resources will be needed at many plants to achieve this level of confidence. Additional work will be required to estimate risk for accidents in Classes 1 through 8. Ongoing maintenance of the PRA will also be necessary. At many plants, additional training will be necessary for the licensee staff. The nature of interactions with inspectors will change to be more focused on PRA-related activities and this will impact needed licensee resources. The potential exists for a significant return on investment for the licensees iflarge numbers ofcost-effective changes are allowed by this process. If the process works properly, less time and resources will be spent on each individual change, because the PRA results can be used to screen things out quickly.
Table 6.4.13-1 presents the results from the evaluation of this option against the evaluation criteria described in Section 4.
~ $2 -
l
- su 59 unt-Informed Optmn opnon, Table 6.4.13-1 Evaluation Resultsfor Frequency-Consequence Curves Option RLMH 'RCI:
TlifE LT iIF-t 110% I ACTOR R.4 TL%G
(
l LT l
l
.% T ST LT PRA POLICY STA'IE%1ENT INIPLICATIONS qp r cy;rwne:. u v.x n i hanced Safe!) Decisions y es Ph 4K..f. -w14WC-Q;Ms(%E kpgm s.*
n n
s.ofM;h.' _ _ _ __ -.g..:
5 2
-f 7
mm w.#a.w hihhl2h;khhh [,
More 1.tticient Use ol' Agency
> es k [hh.h fff Reduction of Unnecessary llurden y es FACIL11Y COVERAGE less
%EfME!'g'NE9NN?AMkTK[...
'..),y j [ j '-
IslPl.lCA'llONS
%h h j;{}
'[}'qd RISK IMPl.lCNIIONS N[ca egxy' -
Ndh Nh$/kkM M
NNNk Confidence Required from Risk high y
id m.
+bL$2dildk?$ l D5A? &p.8W' WiW YC O
.sw Completeness hffh Ihffhf PRA scope medium
.7 j,.g g
u_za.-._ m. A___
E L.a.,
hyhk(kjf"[Thhih k
[
Determinsue u Risk. Informed ri REGl:LVIOR) IMPl.lCATIONS E g y. w e u long g$fh-u,w Reculaton Guidance Needed jfM 3 es sicn M::c sHMw.e impact on Otner Regulations x
N%e Ph % %
no NA gggmyt) NA g,g New RuleNiaking Needed y es mod
[
h!
hh med i
impact on Inspection:Traming y es sign same long same 1 K ENSEF IMPI ICATIONS h,hh{.[h Interna: Prosedures Needed yes sign h[
f med impact on PR A Activihes yes sign more long more
- [
long Impact on T raming y es sign A %:ny+fg gg,A4gy6M gpyg,tw 3:
1mpact on inspection y es same same
+..m D,d$+$&T*2 wm gup 44 Shykgpgh3 less impact on faubt) Chances yes less Evaluation Using Test Case Under the current #50.59 process, this change was determined to be a USQ because it increased the probability ofoccurrence ofa malfunction ofequipment important to safety previously evaluated in the SAR.
Under this proposed option, NRC review will not be required, assuming that the licensee demonstrates that after the change, the risk remains below the limits set by the frequency-consequence curses.
. 5 3 --
~
v l
{$0.59 Risk. Informed Option. Options Evaluation of the Examples Describedin Section 2.3 Against RG 1.174 Option As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the licensee, to be USQs and have required NRC approval in accord.ance with the current 50.59 process. The poteritial frequency-consequence curve option was used to evaluate whether these same facility changes would still require NRC review and approval. This option did not alleviate the unnecessary burden for all of the examples, since technical specifications were impacted in some cases.
5 Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System Under the current Q50.59 process, this facility change was determined to be a USQ because it increased the consequences of an accident because doses increased by 1 mrem from the UFSAR stated dose. Under this frequency-consequence curve option, the dose change would be likely be acceptable relative to the curves and would not require NRC approval.
Example'2.3.1.2 Error in Air Volume ofSecondary Containment Under the current Q50.59 process, this facilitychange was determined to be a USQ because it was a reduction in the margin of safety and an increased the consequences of an accident, because the operator thyroid dose increased from 22 rem to 25.6 rem. Under this frequency-consequence curve option, the dose change would be factored into the consequences used in the risk calculation. Assuming that the limits were not exceeded, NRC approval would not be required.
Example 2.3.1.3 CalculationalErrorin HeadLoss ofEmergency Core Cooling Suction Strainers Under the current Q50.59 process, this facility change was determined to be a USQ because the lack of enough NPSH could increase in the probability of ECCS equipment malfunction occurrence. Under this frequency-conseq uence curve option, the increase in accident frequency and consequences would be assessed and NRC approval would not be required if the limits were not exceeded.
Example 2.3.1.4 Modification ofMain Feedwater Valves Under the current Q50.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.lE-5 per demand). Under this frequency-consequence curve option,if the risk did not exceed the limits of the curves, then NRC approval would not be required.
Example 2.3.1.5 Discrepancy Between As-Built System and UFSAR Description Under the current 50.59 process, this facility change was determined to be a USQ because the as-built system isolation malfunction was of a different type than that previously evaluated in the UFS AR. Under this frequency-consequence curve option, the change would not require NRC approval as long as the licensee determined that the risk impacts did not cause the curves to be exceeded.
Example 2.3.1.6 Replace RHR Service Water Sump Pump Discharge Check Valve This facility change was not a USQ under the current 50.59 process and would also not be a USQ under the frequency-consequence curve option.
~ $4 -
V
{50.59 Risk-Informed Option. Options Example 2.3.1.7 A dd Procedural Steps to Manually Isolate the Control Rod Drive Pumps During
=
SBO This facility change was not a USQ under the current 50.59 process and would also not be a USQ under the frequency-consequence curve option.
Example 2.3.1.8 Add a 3/16 " Diameter flote Through the Reactor Recirculation Piping inlet Side ofthe Valve Flex-ll' edge Disc His facility change was not a USQ under the current 50.59 process and would also not be a USQ under the frequency. consequence curve option.
l Observations 1
This option is consistent with the intent of the PRA policy statement and requires a more complete PRA (e.g.,
level-3 conseyuence evaluation) than any of the other options. It provides maximum flexibility to the licensees, but requires the most extensive PRA information of any of the options. The need for quality in the PRA results is extremely important. It requires signincant resources by both the NRC and the licensees to develap and implement the process; however, once in place, the process can provide significant savings.
Implemented without augmentation, this option is risk-based. It can be modified to be risk-informed by l
combining it with the principles identified in RG 1.174.
6.4.14 Option: Replace Parameters with Modified Regulatory Guide 1.174 Principles Description i
This option extends the approach describe in Section 6.4.12 to be more risk-informed. In this extended option, selected parameters from RG 1.174 would be used to replace the existing parametric criteria in 50.59. In particular, paragraph (a)(2) of Q50.59 now reads:
A proposed change, test, or experiment shall be deemed to involve an unreviewed safety question (i) if theprobability ofoccurrence or the consequences ofan accident or malfunction ofequipment important to safetypreviously evaluated int eh safety analysis report may be increased: or (ii) ofa possibilityfor an accident or malfunction ofa different type than any evaluatedpreviously in the safety analysis report may be created; or (iii) ofthe margin ofsafety as definedin the basis for any technicalspecification is reduced.
This would be deleted and replaced with:
A proposedchange, test, or experiment shall not requirepr c-Commission approval (i) oftheproposed change meets the current regulations, (ii) iftheproposedchange does not remove a layer ofdefense in depth, and (iii) if the core damagefrequency or risk increases are small and consistent with the intent ofthe Commission 's Safety Goal Policy Statement.
The first item is the same as described in RG l.174 and would be implemented as described therein.
Defense-in-depth is defined differently here, as compared to RG 1.174, in this modified option, defense in depth refers to functional defense in depth that encompasses prevention of accident initiators, prevention of core damage, prevention ofradioactive releases, and emergency response. Unlike the approach in the current RG 1.174, the degree to which each layer is effective and the manner ofimplementation are not restricted, except as they affect CDF and LERF. Further, the need for safety margins is covered in the regulations.
l
~ 55 -
{50.59 Risk-Informed Option, Options Unlike the approach of RG 1.174, this option allows the licensee leeway to reduce margin in excess of that required in the regulations. As described in Section 6.4.12, values of 1.0E-7/ year for an acceptable change in CDF and 1.0E-8/ year for an acceptable change in LERF are proposed. These numbers are suf6ciently small that the current level of CDF and LERF at a plant would not be important. Performance monitoring as described in RG 1.174 is not included here as the maintenance rule would provide adequate assurance of component and system performance.
l Evaluation The evaluation results are the same for this option as for the option in Section 6.4.12. However, there are differences in the degree to which the criteria are met. This option is consistent with the objectives of the i
PRA Policy Statement. This option enhances safety decision making by focusing NRC review on only those i
changes that impact either the current regulations or risk, thus using NRC resources more efHeiently.
Unnecessary burden on industry is reduced more than any of the other options, because additional changes that are relatively unimportant to safety and risk would be allowed without a submittal.
In general, less of the facility will be covered than under the current l50.59. Portions of the plant that do not affect CDF or LERF, or that are not affected by any of the other regulations would be excluded. This will eliminate many SSCs that are included in the SAR.
This option makes use of quantitative CDF and Risk results. Therefore, a high con 6dence is needed in the results for those changes that are compared to the quantitative criteria. The scope rating is stated as
" medium;" because that level is expected to be needed for a substantial fraction of the changes. For a particular change, the needed scope can range from minimum to full. The level of detail must be detailed with respect to the change, although a uniform level of detail for all aspects of a PRA may not be necessary for a particular change.
Since this option still includes compliance with other regulations, and those regulations are largely deterministic, this option is classified as deterministic / risk-informed, although it is more risk-informed than any of the other options.
New Rulemaking and regulatory guidance will be needed to implement this option. However, regulatory guidance is needed for the current 50.59, and some guidance is already available in RG 1.174. Therefore additional time and resources are not required beyond that expected for the other options. This option will have minimal impact on other regulations. Nevertheless, the inspectors will require training in methods to verify compliance. Once new inspection processes are in place, it is assumed that the total inspection time will remain the same, although a larger fraction of that time will be focused on PRA-related activities.
The licensees will need to develop their own internal procedures for implementing this option. This will involve additional resources and time for implementation in the beginning. The actual PRA activities will be impacted due to the need for a high confidence PRA with appropriate scope and level ofdetail. Signincant time and resources will be needed at many plants to achieve this level of con 0dence. Ongoing maintenance ofthe PRA will also be necessary. At many plants, additional training will be necessary for the licensee staff.
The nature ofinteractions with inspectors will change to be more focused on PRA-relatec' activities and this will impact needed licensee resources. The potential exists for a significant return on investment for the licensees if large numbers of cost-effective changes are allowed by this process. If the process works l
properly, less time and resources will be spent on each individual change, because the PRA results can be used to screen things out quickly.
Table 6.4.14-1 presents the results from the evaluation of this option against the evaluatien criteria described in Section 4.
l un w Rist. informed Oroon. Opnons Table 6.4.14-1 Evaluation Resultsfor.11odified RG 1.174 Option l
RESOURCE IlitL L i'4 LI C4110 Y litC10R R 411.%G l
L1 ST LT ST PRA POLICY STA1 LMt NT 1%tPl.lCAllONS
,,s.nv~v4 1:nhanced Safety Decisions pygggf(*. w*ye+ igg y~pfgg g,., m:gg wm.
y yes jg ggg h ]a,.., 7..T {ej.' O( hhfh h More Efficient Use of Agency y es f.
Reducuan of Unnecessary Hurden
- , g gp lk
{(
3 cs y,m y_:.. -......,._.
,..,.y...,.
FACILITY COVERAGE less gf
. m.
<y.
Tb.....
IMPLICAllONS NJ.
'?
- ^4 W.
Ti '
'm.
....' M
~
~~
p.
. m.;;,e RISK 1%!Pl.lCAllOM Contidence Required from Risk lugh
^Nd
. N d.% A I k WN b[ddIN i
k3;, j.
.j'fMl)";
h Results hNfjfYh
- ;{.
h hfh PRA swpe medium Completeness khpv. ? -l, J,, ]y',[(.%- h,,~(h-../
4,,,
lesel of detad detaded Determuusoc ss Risk informed det-n h% '.'l cj h$ki h,ff[l REGt'LA10RY IMPLIC A 110%
Regulaton Guidance Needed h
y es mod j
med Impact on Other Regulations h) khkk no NA NA J am&9 NP.: w v A Wrva New RuleMaking Needed yes mod
~, pr med
% g ;gtg 3 Impact on inspection'i rammg y es sign same long same LICENSEI: IM PLICA TIO%
Internal Procedures Needed
,;e w.m
.,,9 fe'$6)$m$)
g';"jj
)es sign med Impact on PRA Acuuhes y es sign more long more impact on T raimny
> es sign f Y.7
?.! ( ~long ek i
(af.M...M c
ji$gd ~{~{
impact on Inspccuon
> cs g;
same same kb tess YNd N ddh p Impact on Facihty Changes 3 es less Evaluation Using Test Case l
Under the current 50.59 process, this change was determined to be a USQ because it increased the probability of occurrence of a malfunction ofequipment important to safety previously evaluated in the S AR.
Under this proposed option, NRC res iew will still be required, because the components are designated as safety-related, and therefore the change impacts the other regulations. If this option were combined with options to change the scope or definition of w hat is considered safety related, then it is hkely that this change i
would not need review. The change does not affect defense in depth. The licensee would have to determine that the change resulted in CDF and LERE increases less than the threshold values presented above. The CDF and LERF results would need to address the complete set ofinitiating events.
. 5 7..
i
o e
l l
f 50.59 Risk-informed Option, Options l
Evaluation ofthe Examples Describedin Section 2.3 Against RG 1.174 Option As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the licensee, to be USQs and have required NRC approval in accordance with the current Q50.59 process. The potential modified RG 1.174 50.59 option was used to evaluate whether these same facility changes would still require NRC review and approval. This option did not alleviate the unnecessary burden for all of the i
examples, since technical specifications were impacted in some cases.
1 l
Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System l
Under the current Q50.59 process, this facility change was determined to be a USQ because it increased the consequences of an accident because doses increased by I mrem from the UFSAR stated dose. Under this modified RG 1.174 option, the dose change would be acceptable and would not require NRC approval.
1 Example 2.3.1.2 Error in Air Volume ofSecondary Containment Under the current Q50.59 process, this facility change was determined to be a USQ because it was a reduction in the margin of safety and an increased the consequences of an accident, because the operator thyroid dose increased from 22 rem to 25.6 rem. Under this modified RG 1.174 option, the dose change would represent a change in technical specifications and would still require NRC approval.
Example 2.3.1.3 Calculational Error in Ilead Loss ofEmergency Core Cooling Suction Strainers Under the current 50.59 process, this facility change was determined to be a USQ because the lack of enough NPSH could increase in the probability of ECCS equipment malfunction occurrence. Under this modified RG 1.174 option, changes to technical specifications are involved and the change would still require NRC approval.
Example 2.3.1.4 Modification ofMain Feedwater Valves Under the current Q 50.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.lE-5 per demand). Under this modified RG l.174 option, if the CDF did not increase more than 1E-7, then NRC approval would not be required.
Example 2.3.1.5 Elscrepancy Between As-Built System and UFSAR Description l
Under the current Q50.59 process, this facility change was determined to be a USQ because the as-built system isolation malfunction was ofa different type than that previously evaluated in the UFSAR. Under this modified RG 1.174 option, the change would not require NRC approval as long as the licensee determined that the CDF and LERF impacts were less than 1E-7 and 1E-8, respectively.
Example 2.3.1.6 Replace RilR Service Water Sump Pump Discharge Check Valve This facility change was not a USQ under the current Q50.59 process and would also not be a USQ under the modified RG 1.174 option.
1.
7 650.59 Risk-Infonned Option. Options Exan.ple 2.3.1. 7 A dd ProceduralSteps to Manually isolate the Control Rod Drive Pumps During SBO This facility change was not a USQ under the current 50.59 process and would also not be a USQ under the modified RG 1.174 option.
Example 2.3.1.8 A dd a 3/16 " Diameter flote Through the Reactor Recirculation Piping Inlet Side ofthe Valve Flex-Wedge Disc This facility change was not a USQ under the current Q50.59 process and would also not be a USQ under the modified RG 1.174 option.
Obsen ations This option is. consistent with the intent of the PRA policy statement and is more risk informed than any of the other options. It provides maximum flexibility to the licensees while requiring adherence to current regulations. The need for quality in the PRA results is extremely important for this option. It requires significant resources by both the NRC and the licensees to develop and im plement the process; however, once in place, the process can result in significant savings.
6.4.15 Option: Risk Increase Interval (RAW)
Description This option would replace the existing parametric criteria in Q50.59 (a)(2)(i) and (a)(2 )(ii) with the following:
"ofthe risk increase interval associated with aproposed change, test, or experiment is greater than 1E-6/yearfor CDF or 1E-7/yearfor LERF.*
This option uses the risk increase interval (Ril) as a quantitative measure of the risk effects in a risk-informed Q50.59 process. Rll measures, on an absolute scale, the change in risk (e.g., CDF, LERF, etc.) that would occur.if a component's failure probability increased from its nominal value to 1.0. Since most Q50.59 changes would not be expected to cause a component's failure probability to increase to 1.0, use of this measure provides robustness in identifying proposed Q50.59 changes that would require NRC approval before the change could be implemented.
Use of this option requires the establishment of an acceptable Ril. The acceptable RIl is based r *e acceptance guidelines for CDF and LERF provided in RG 1.174, where a minimal change in CDF is d
,d to be less than IE-6/ year (IE-7/ year for LERF). Using IE-6/ year (IE-7/ year) as a definition of an acceptable Rll ensures that the change in risk associated with a proposed j50.59 change would not violate the risk-acceptance guidelines in RG 1.174. In reality, the actual change in risk associated with a proposed l50.59 change could be significantly less because most changes would not be expected to cause the component's failure probability to increase to 1.0. To eliminate truncation problems, RIl should be calculated l
by setting the cor.1ponent's failure probability to 1.0, resolving the PRA model to calculate the new CDF (LERF), and then subtracting the original CDF (LERF) from this new value.
The nature of this option is such that it only addresses accidents involving the reactor core, i.e., radiological hazards outside the reactor core are not addressed..
_m l
(50 59 Risk-Informed Option. Options Evaluation This option could be used to support the objectives of the PRA Policy Statement. It enhances safety decision making by focusing NRC review on only those changes that are important to CDF or risk (i.e., LERF), thus using NRC resources more efficiently. It also reduces unnecessary burden on industry by allowing the
^
licensee to identify and minimize requirements associated with components that are not CDF or LERF important or which have minimal CDF or LERF importance without having to receive NRC approval before the change can be put in place.
In general, less of the facility will be covered than under the curTent %50.59. Portions of the plant that do not affect CDF or LERF would be excluded. This will eliminate some SSCs that are included in the SAR.
l This option makes use of quantitative CDF and LERF results. Therefore, a high confidence is needed in the results for those changes that are compared to the quantitative criteria. The scope rating is stated as
" medium;" because that level is expected to be needed for a substantial fraction of the changes. That is, external events or low power and shutdown may not need to be considered for many changes. For a particular change, the needed scope can range from minimum to full. The level of detail must be commensurate with the change, although a uniform level of detail for all aspects of a PRA may not be necessary fc,r a Particular change. Since Rils for CDF and LERF form the basis for deciding whether NRC approval is needed or not needed, the option is classified as risk-informed.
New rulemaking and regulatory guidance will be needed to implement this option. While it is expected that this option will have no impact on other regulations, the inspectors will require training and modules to verify compliance.
1 The resources required to put both the regulatory guidance and rulemaking in place to support this option is estimated to be between one and two staff-years; thus,the " minimal" designation for resources. This estimate is based on the fact that resources have already been expended to develop a risk-informed approach, as described in RG 1.174, and that this option only requires the comparison of Rils for CDF and LERF to some specified value. Since this option is expected to have no impact on other regulations, there are no associated resource requirements. The resources required to train the inspectors and to modify the inspections modules is estimated to be less than one staff-year; thus, a " minimal" designation for resources. This estimate is based on the fact that inspectors have, or will soon have, had training on PRA principals, and that most of the resources will be spent on modifying the inspection modules. Once the training and modules are in place, it is expected that this option will require the same amount of resources to perform the inspections, although a larger fraction of the resources will be focused on PRA-related activities.
The time required to develop regulatory guidance, put new rules in place, and develop new inspection modules and train the inspectors is estimated to be between six months and a year, a " medium" designation.
Again, these estimates are based on the reasons provided in the above paragraph. Once the training modules are in place, it is expected that the inspectors will spend the same amount of time performing inspections, although a larger fraction of their time wiP be spent on examining PRA-related information.
l l
The licensees will need to develop their own internal procedures for implementing this option. PRA activities will be impacted due to the need for a high confidence PRA with appropriate scope and level of detail. At ruany plants, additional training will be necessary for the licensee staff. The nature of interactions with inspectors will change to be more focused on PRA-related activities. The potential exists for a significant return on investmerit for the licensees if targe numbers ofcost-effective changes are allowed by this process.
l i
450 59 Rnk-Informed Optmn. Options Table 6.4.15-1 Evaluation Resultsfor Rid increase Interval Option RESOURCE 11stE El'AL UA 110% FACTOR R 411%
l LT l
ST )
LT l
ST 8,, Cw.cma:n m,..
1,;.
'_hxk.~*{nh~f$himyg lesel of detail detailed Q
h [. -).l p," I.. p g
{
Determmistic vs Risk informed risk-h,j Q' 3n_ _
k,.5 : 7..r -
mtormed REGUI.ATORY INtPl.lCA I10%
Regulatory Guidance Needed yes mmimal medium
(' h h
h Impact on Other Regulations no N/A N/A r
_^T_2._
New RuleMakmg Needed
>cs mmimal
.h.
([;h medium impact on Inspectionfirammy y es mimmal same medium same I.lCEMEE INtPl.lCATIOM
[ h;
,b)hhfh Internal Procedures Needed yes mimmal short Impact on PRA Actnittes y es sigmficant more long more l
- ~ - -
y.,5geh-g.
Impact on frammg y es moderate short
@hjy-hf,(f Impact on inspection
>es same same N
less hhNIk impact on facility Changes s es lew Evaluation Using Test Case Under the current {.50.59 process, this change was determined to be a USQ because it increased the probability ofoccurrence of a malfunction of equipment important to safety previously evaluated in the S A R.
Under this proposed option, NRC review will still be required, because the components are designated as safety-related, and therefore the change impacts the other regulations. How ever,it this proposed option were the only basis for deciding uhether NRC res iew would be required, such a review would not be necessary, since loss of the screen uash pumps:
has e no risk increase inten al for either CDF and LERF for power operation initiators, have no impact on accident sequences in other operational states, and the only external initiators that might pose a problem also fail the essential cooling w ater pumps.
Evaluation of the Examples Described in Section 2.3 Against RG 1.174 Option l
As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the licensee, to be USQs and have required NRC approval in accordance with the current h50.59 process. The potential risk increase intenal option was used to evaluate whether these same facility changes uould still require NRC review and approval. This option did not allesiate the unnecessary burden for all of the examples, since technical specifications were impacted in some cases.
l i
i l
62 --
. ~-
_ - - -. ~.. -
)
l 650.59 Risk-Informed Option, Options The resources required to by the licensee to develop their own internal procedures is estimated to be less than one staff year, a minimal designation. This estimate is based on the assumption that licenseer. currently have procedures in place to control 50.59 reviews and submittals and that only minor modifications will be needed to the procedures to check whether the Rils for CDF or LERF exceed the specified values. Initially, PRA activities will require significant resources, more than two staff-years. These resources include those needed to bring their current PRA up to appropriate standard levels. Resources for long term PRA activities are expected to be more than currently required because of the need for ongoing maintenance of the PRA and the increased use ofPRA to support specific 50.59 change activities. The resource requirements for licensee training arejudged to be moderate, more than one stafr-year but less than two staff-years. Thisjudgment is based on the assumption that most licensees have some PRA capability and would not be required to start a PRA group from scratch. Once the new process is in place, no additional inspection or facility changes resources are expected. Nevertheless, it is expected that more 50.59 changes can be processed owing to the risk-informed nature of the option.
The time required for the licensee to develop their internal procedures is estimated to be short, less than six l
months. This is based on the assumption that the existing procedure would be replace by one that only required a PRA calculation and a check of the Riis for CDF and LERF for the change being proposed. In the short term, the time required for PRA activities is expected to be long, more than one year. This estimate is base on the assumption that, for most licensees, it will require more than one year to enhance their PRA such that high confidence can be placed in the numerical results from the PRA. In the long term, PRA activities will require more time since the PRA will become an important tool in the G50.59 decision making l
i process. The time required to implement training is estimated to be short, less than six months. This is based on the assumption that most licensees already have 50.59 processes in pbce and that only training in how to use the information for the PRA to make risk-informed {50.59 decisions would be needed. While it is expected that the same amount of resources will be used for inspection and facility changes, the time spem l
on each individual inspection or change will be less owing to the risk-informed nature of the option.
Table 6.4.15-1 presents the results from the evaluation of this option against the evaluation criteria described in Section 4.
l Table 6.4.1S-1 Evaluation Resultsfor Risk increase Inten al Option l
t RESOURCE TIME EVALUATION FACTOR RA TING l
ST LT l
ST LT 1
PRA POLICY STATEMENT IMPLICATIONS Enhanced Safety Decisions yes More Efficient Use of Agency yes Reduction of Unnecessary Burden yes FACILITY COVERAGE less IMPLICATlONS RISK IMPLICATIONS l
Confidence Required from Risk high Results f
PRA scope medium LOmpleLChCh5 _
(
~
650.59 Risk-informed Option. Options
(
Example 2.3.1.1 Error in Dose Ctaculationsfor the Process Gas System Under the current Q50.59 process, this facility change was determined to be a USQ because it increased the consequences of an accident because doses increased by 1 mrem from the UFSAR stated dose. Under this risk increase interval option, the dose change would be acceptable because the change does not affect (i.e.,
increase) CDF; thus, not requiring NRC approval.
Example 2.3.1.2 Error in Air Volume ofSecondary Containment Underthecurrent 50.59 process, this facility change was determined to be a USQ because it was a reduction l
in the margin of safety and an increased the consequences of an accident, because the operator thyroid dose increased from 22 rem to 25.6 rem. Under this risk increase interval option, the dose change would represent a change in technical specifications and would still require NRC approval.
Example 2.3.1.3 CalculationalErrorin HeadLoss ofEmergency Core Cooling Suction Strainers Under the current Q50.59 process, this facility change was determined to be a USQ because the lack of enough NPSH could increase in the probability of ECCS equipment malfunction occurrence. Under this risk increase interval option, changes to technical specifications are involved and the change would still require NRC approval.
Exangle 2.3.1.4 Modification ofMain Feedwater Valves Under the current iS0.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.1E-5 per demand). Under this risk increase interval option, if the RRIs for CDF and LERF were less than IE-6 and I E-7, then NRC approval would not be required.
Exampic 2.3.1.5 Discrepancy Between As-Built System and UFSAR Description Under the current Q50.59 process, this facility change was determined to be a USQ because the as-built system isolation malfunction was of a different type than that previously evaluated in the UFSAR. Under this risk increase interval option, the change would not require NRC approval as long as the licensee determined that the RIIs for CDF and LERF impacts were less than 1E-6 and IE-7.
Example 2.3.1.6 Replace RHR Service Water Sump Pump Discharge Check Valve This facility change was not a USQ under the current 50.59 process and would also not be a USQ under the risk increase interval option.
Example 2.3.1.7 Add ProceduralSteps to Manually isolate the Control Rod Drive Pumps During SB0 This facility change was not a USQ under the current {50.59 process and would also not be a USQ under the risk increase interval option.
l
[
Example 2.3.1.8 A dd a 3/16 " Diameter Hole Through the Reactor Recirculation Piping inlet Side
\\
ofthe Valve Flex-Wedge Disc l
'This facility change was not a USQ under the current Q50.59 process and would also not be a USQ under the risk increase interval option..
{50.59 Risk-Informed Option, Options Observations This option could be used to support the objectives of the PRA policy statement and is more risk-informed than the current {50.59. It requires a determination of risk (both CDF and LERF). In total, it requires signiGcant resources by both the NRC and the licensees to develop and implement the process; however, once in place, the process can result in significant savings.
This option will only provide coverage for class 9 accidents using current PRA information and provides a more conservative representation of the change in risk.
For the test case, if the option is the sole basis for determination, it would allow a change in safety designation of the pumps.
6.4.16 Option: Maintenance Rule Metrics (RRW/ RAW) x Description This option would replace the existing parametric criteria in 50.59 (a)(2)(i) and (a)(2)(ii) with the following:
"ifthe risk reduction worth associated with aproposedchange, test, or experiment is greater than L OOSfor CDF orfor LERF or the risk achievement worth associated with aproposed change, test or experiment is greater that 2.0for CDF orfor LERF."
i This option uses the risk reduction worth (RRW) and the risk achievement worth (RAW) as quantitative measures of the risk effects in a risk informed 50.59 process. This approach would be very similar to that used by many licensees to identify high and low risk significant systems, structures, and components (SSCs) for purposes ofimplementing the Maintenance Rule, { 50.65, and the related RG 1.160. That regulatory i
guide finds the guidance provided in NUMARC 93-01 as an acceptable approach for implementing the Maintenance Rule. The NUMARC guidance suggests the use of a RRW ratio of greater than 1.005 and a RAW ratio of greater than 2.0 as thresholds for indicating high vs. Iow risk significant SSCs. Using this approach, the licensee would evaluate the proposed change for which equipment is affected by the change and then use a PRA to calculate the RRW and RAW for the collective set of equipment that is affected.
Comparison of the resulting RRW/ RAW values to acceptable limits would identify whether there is a l
" minimal" impact on risk, and hence NRC review of the proposed change is not required. Possible new accidents associated with the proposed change would have to be modeled and included when calculating the l
RRW/ RAW values.
To eliminate truncation problems, the RAW ratio should be calculated by setting the component's failure probability to 1.0, resolving the PRA model to calculate the new CDF (LERF), and then dividing the original CDF (LERF) value into the new value.
The nature of this option is such that it only addresses accidents involving the reactor core, i.e., radiological I
hazards outside the reactor core are not addressed.
i i
Evaluation This option could be used to support the objectives of the PRA Policy Statement. It enhances safety decision making by focusing NRC review on only those changes that are important to CDF or risk (i.e., LERF), thus using NRC resources more ef0ciently, it also reduces unnecessary burden on industry by allowing the i
licensee to identify and minimize requirements associated with components that are not CDF or LERF 64--
i
_ = _ _ _ _ _ _
_m
{50 59 Risk-Informed Option. Options important or which have minimal CDF or LERF importance without having to receive NRC approval before the change can be put in place.
In general, less of the facility will be covered than under the current 50.59. Portions of the plant that do not affect CDF or LERF would be excluded. This will eliminate some SSCs that are included in the SAR.
This option makes use ofquantitative CDF and LERF results. Therefore, a high confidence is needed in the results for those changes'that are compared to the quantitative criteria. The scope rating is stated as
" medium;" because that level is expected to be needed for a substantial fraction of the changes. That is, external events or low power and shutdown may not need to be considered for many changes. For a particular change, the needed scope can range from minimum to full. The level of detail must be commensurate with the change, although a uniform level of detail for all aspects of a PRA may not be necessary for a particular change. Since RRWs and RAWS for CDF and LERF form the basis for deciding whether NRC approval is needed or not needed, the option is classified as risk-informed.
New rulemal)ing and regulatory guidance will be needed to implement this option. While it this option will have no impact on other regulations, the inspectors will require training and modules to verify compliance.
The resources required to put both the regulatory guidance and rulemaking in place to support this option is estimated to be between one and two staff-years; thus, the " minimal" designation for resources. This estimate is based on the fact that resources have already been expended to develop a risk-informed approach, as described in RG 1.174, and that this option only requires the comparison of RRWs and RAWS for CDF and LERF to some specified value. Since this option is expected to have no impact on other regulations, there are no associated resource requirements. The resources required to train the inspectors and to modify the inspections modules is estimated to be less than one staff-year; thus, a " minimal" designation for resources.
This estimate is based on the fact that inspectors have, or will soon have, had training on PRA principals, and that most of the resources will be spent on modifying the inspection modules. Once the training and modules are in place, it is expected that this option will require the same amount of resources to perform the inspections, although a larger fraction of the resources will be focused on PRA-related activities.
The time required to develop regulatory guidance, put new rules in place, and develop new inspection modules and train the inspectors is estimated to be between six months and a year, a " medium" designation.
Again, these estimates are based on the reasons provided in the above paragraph. Once the training modules are in place, it is expected that the inspectors will spend the same amount of time performing inspections, although a larger fraction of their time will be spent on examining PRA-related information.
The licensees will need to develop their own internal procedures for implementing this option. PRA activities will be impacted due to the need for a high confidence PRA with appropriate scope and level of detail. At many plants, additional training will be necessary for the licensee staff. The nature ofinteractions with l
inspectors will change to be more focused on PRA-related activities. The potential exists for a significant return on investment for the licer, sees iflarge numbers ofcost-effective changes are allowed by this process.
The resources required to by the licensee to develop their own internal procedures is estimated to be less than one staff-year, a minimal designation. This estimate is based on the assumption that licensees cunently have procedures in place to control Q50.59 reviews and submittals and that only minor modifications will be needed to the procedures to check whether the RRWs and RAWS for CDF or LERF exceed the specified l
values. Initially, PRA activities will require significant resources, more than two staff-years. These resources include those needed to bring their current PRA up to appropriate standard levels. Resources for long term PRA activities are expected to be more than currently required because of the need for ongoing maintenance -
l i
~ _
l d
i f
i i
3 1
(50.59 Risk-infonned Option, Options I
of the PRA and the increased use of PRA to support specific 50.59 change activities. The resource requirements for licensee training arejudged to be moderate, more than one staff-year but less than two staff-years. Thisjudgment is based on the assumption that most licensees have some PRA capability and would 3
j not be required to start a PRA group from scratch. Once the new process is in place, no additional inspection j
or facility changes resources are expected. Nevertheless, it is expected that more 50.59 changes can be processed owing to the risk-informed nature of the option.
The time required for the licensee to develop their internal procedures is estimated to be short, less than six j
months. This is based on the assumption that the existing procedure would be replace by one that only I
required a PRA calculation and a check of the RRWs and RAWS for CDF and LERF for the change being proposed. In the short term, the time required for PRA activities is expected to be long, more than one year.
j This estimate is base on the assumption that, for most licensees, it will require more than one year to enhance their PRA such that high confidence can be placed in the numerical results from the PRA. In the long term, PRA activities will require more time since the PRA will become an important tool in the 50.59 decision l
making process. The time required to implement training is estimated to be short, less than six months. This is based on the assumption that most licensees already have 50.59 processes in place and that only training in how to use the information for the PRA to make risk-informed Q50.59 decisions would be needed. While
)
it is expected that the same amount of resources will be used for inspection and facility changes, the time t
i l
spent on each individual inspection or change will be less owing to the risk-informed nature of the option.
i Table 6.4.16-1 presents the results from the evaluation of this option against the evaluation criteria described I
in Section 4.
Table 6.4.16-1 Evaluation Resultsfor Maintenance Rule Metrics
{
(RR FI'/RA II) Option i
i l
RESOURCE TIME El'ALUA TION FACTOR RATING ST LT l
ST l
LT l
1 PRA POLICY STATEMENT IMPLICATIONS Enhanced Safety Decisions yes More Efficient Use of Agency yes
[
Reduction of Unnecessary Burden yes FACILITY COVERAGE less l
IMPLICATIONS RISK IMPLICATIONS I
l Confidence Required from Risk high Results g
a
{
PRA scope medium a
Completeness
]
level of detail detailed i
j Deterministic vs Risk-Informed risk-j informed 4
REGULATORY IMPLICATIONS j
Regulatory Guidance Needed yes minimal medium i
4' a 1
l (50.59 Risk-Informed Option Options Table 6.4.16-1 Evaluation Resultsfor Maintenance Rule Metrics (RRIl'/RAIf' Option
)
l RESOURCE TlhiE EVALUA TION FACTOR RA TING l
h ST l
LT ST LT Impact on Other Regulations i no N/A N/A New RuleMaking Needed yes minimal
_ medium impact on Inspection / Training yes minimal same medium same LICENSEE IMPLICATIONS Internal Procedures Needed yes minimal short Impact on PRA Activities yes significant more long more impact on Training yes moderate short Impacs o inspection
> cs same same impact on Facil,ity Changes yes less less Evaluation Using Test Case 1
Under the current 50.59 process, this change was determined to be a USQ because it increased the probability ofoccurrence ofa malfunction ofequipment important to safety previously evaluated in the SAR.
Under this proposed option, NRC review will still be required, because the components are designated as safety-related, and therefore the change impacts the other regulations. However, it this proposed option were the only basis for deciding whether NRC review would be required, such a review would not be necessary, since loss of the screen wash pumps:
have no risk reduction worth or no risk achievement worth for either CDF and LERF for power operation initiators, l
have no impact on accident sequences in other operational states, and the only external initiators that might pose a problem also fail the essential cooling water pumps.
Evaluation of the Examples Describedin Section 2.3 Against RG 1.174 Option As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the licensee, to be USQs and have required NRC approval in accordance with the current 50.59 process. The potential risk reduction worth and risk achievement worth option was used to evaluate whether these same facility changes would still require NRC review and approval. This option did not alleviate the unnecessary burden for all of the examples, since technical specifications were impacted in some cases.
Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System Under the current 50.59 process, this facility change was determined to be a USQ because it increased the consequences of an accident because doses increased by 1 mrem from the UFSAR stated dose. Under this risk reduction worth and risk achievement worth option, the dose change would be acceptable because the l
ch age does not affect (i.e., increase) CDF; thus, not requiring NRC approval.
l l
{$0.59 Risk. Informed Option. Options Example 2.3.1.2 Error in Air Volume ofSecondary Containment Under the current 50.59 process, this facility change was determined to be a USQ because it was a reduction in the margin of safety and an increased the consequences of an accident, because the operator thyroid dose increased from 22 rem to 25.6 rem. Under this risk reduction worth and risk achievement worth option, the dose change would represent a change in technical specifications and would still require NRC approval.
I Example 2 3.1.3 - Calculational Error in Head Loss ofEmergency Core Cooling Suction Strainers Under the current Q50.59 process, this facility change was determined to be a USQ because the lack of enough NPSH could increase in the probability ofECCS equipment malfunction occurrence. Under this risk reduction worth and risk achievement worth option, changes to technical specifications are involved and the change would still require NRC approval.
Example 2.3.1.4 Modification ofMain Feedwater Valves Under the current Q50.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.lE-5 per demand). Under this risk reduction worth and risk achievement worth option, if the RRWs and RAWS for CDF and LERF were less than 1.005 and 2, then NRC approval would not be required.
Example 2.3.1.5 Discropancy Between As-Built System and UFSAR Description Under the current {50.59 process, this facility change was determined to be a USQ because the as-built system isolation malfunction was ofa different type than that previously evaluated in the UFSAR. Under this risk reduction worth and risk achievement worth option, the change would not require NRC approval as long 1
as the licensee determined that the Riis for CDF and LERF impacts were less than 1E-6 and 1E-7.
Example 2.3.1.6 Replace RHR Service WaterSump Pump Discharge Check Valve This facility change was not a USQ under the current {50.59 process and would also not be a USQ under the risk reduction worth and risk achievement worth option.
Example 2.3.1.7 Add Procedural Steps to Manually Isolate the Control RodDrive Pumps During SBO This facility change was not a USQ under the current Q50.59 process and would also not be a USQ under the risk reduction worth and risk achievement worth option.
l Example 2.3.1.8 A dd a 3/16 " Diameter Hole Through the Reactor Recirculation Piping Inlet Side ofthe Valve Flex-Wedge Disc This facility change was not a USQ under the current l50.59 process and would also not be a USQ under the risk reduction worth and risk achievement worth option.
Observations This option could be used to support the objectives of the PRA policy statement and is more risk-infonned I
than the current {50.59. It requires a determination of risk (both CDF and LERF). In total, it requires l
l'
~ 68 -
L l
l
j o
f 50.59 Risk-Informed Option. Options significan'. resources by both the NRC and the licensees to develop and implement the process; however, once in place, the process can result in significant savings.
This option will only provide coverage fer class 9 accidents using current PRA information and provides a more conservative representation of the change in risk.
For the test case, if the option is the sole basis for determination, it would allow a change in safety designation of the pumps.'
j 6.4.17 Option: Core Damage Frequency Description This option would replace the existing parametric criteria in 50.59 (a)(2)(i) and (a)(2)(ii) with the following:
\\
"ifthe change in core damagefrequency associated with aproposedchange, test, or experiment is greater than ih-7/ year."
This option uses the change in core damage frequency (CDF) as a quantitative measure of the risk effects in a risk-informed Q50.59 process. Using this approach, the licensee would evaluate the proposed change to identify the equipment affected by the change, determine the affect on the equipment's failure probability.
use the plant PRA to calculate a new CDF reflecting the change in the equipment's failure probability, and then subtract from this new CDF the original CDF (i.e., the CDF before the change was made). Comparison of the resulting change in CDF to acceptable limits would identify whether there is a " minimal" impact on risk, and hence NRC review of the proposed change is not required. Possible new accidents associated with the proposed change would have to be modeled and included when calculating the new CDF value.
Use of this option requires the establishment of a acceptable " minimal" change in CDF. The acceptable
" minimal" change in CDF is based on acceptance guidelines for CDF provided in RG 1.174, where a
" minimal" change in CDF (i.e., one that will be considered regardless of the total CDF) is defined to be less than IE-6/ year. Since one of the goals of s50.59 is to identify changes that do not require NRC review, the definition of " minimal" change in CDF for {50.59 (i.e., one not requiring NRC notification) is IE-7/ year, a factor of 10 less than the ' minimal" change that will be considered by NRC in RG 1.174. Using i E-7/ year as a definition of an acceptable change in CDF ensures that the change in risk associated with a proposed 650.59 change would not violate the risk-acceptance guidelines in RG 1.174.
The nature of this option is such that it only addresses accidents involving the reactor core, i.e., radiological hazards outside the reactor core are not addressed.
Evaluation This option could be used to support the objectives of the PRA Policy Statement. It enhances safety decision making by focusing NRC review on only those changes that are important to CDF, thus using NRC resources more efficiently. It also reduces unnecessary burden on industry by allowing the licensee to identify and minimize requirements associated with components that are not CDF important or which have minimal CDF importance without having to receive NRC approval before the change can be put in place.
In general, less of the facility will be covered than under the current 50.59. Portions of the plant that do not affect CDF would be excluded. This will eliminate some SSCs that are included in the SAR.
._q-l l
650.59 Risk-Informed Option, Options This option makes use of quantitative CDF results. Therefore, a high confidence is needed in the results for those changes that are compared to the quantitative criteria. The scope rating is stated as " medium;" because l
that level is expected to be needed for a substantial fraction of the changes. That is, external events or low power and shutdown may not need to be considered for many changes. For a particular change, the needed scope can range from minimum to full. The level ofdetail must be commensurate with the change, although a uniform level ofdetail for all aspects ofa PRA may not be necessary for a particular change. Since chang in CDF forms the basis for deciding whether NRC approval is needed or not needed, the option is classifie i
as risk-informed.
New rulemaking and regulatory guidance will be needed to implement this option. While it is expected that this option will have no impact on other regulations, the inspectors will require training and modules to v compliance.
l The resources required to put both the regulatory guidance and rulemaking in place to support this option is estimated to be between one and two staff years; thus, the " minimal" designation for resources. This estimate is based on the fact that resources have already been expended to develop a risk-informed approach, as l
described in RG 1.174, and that this option only requires the comparison ofchange in CDF to some specified value. Since this option is expected to have no impact on other regulations, there are no associated resource requirements. The resources required to train the inspectors and to modify the inspections modules is estimated to be less than one staff-year; thus, a " minimal designation for resources. This estimate is based on the fact that inspectors have, or will soon have, had training on PRA principals, and that most of the resources will be spent on modifying the inspection modules. Once the training and modules are in place, it is expected that this option will require the same amount ofresources to perform the inspections, although a larger fraction of the resources will be focused on PRA-related activities.
The time required to develop regulatory guidance, put new rules in place, and develop new inspection modules and train the inspectors is estimated to be between six months and a year, a " medium" designation.
I Again, these estimates are based on the reasons provided in the above paragraph. Once the training modules are in place, it is expected that the inspectors will spend the same amount of time performing inspections, i
although a larger fraction of their time will be spent on examining PRA-related information.
The licensees will need to develop their own internal procedures forimplementing this option. PRA activities will be impacted due to the need for a high confidence PRA with appropriate scope and level of detail. At many plants, additional training will be necessary for the licensee staff. The nature ofinteractions with inspectors will change to be more focused on PRA-related activities. The potential exists for a significant return on investment for the licensees iflarge numbers ofcost-effective changes are allowed by this process.
The resources required to by the licensee to develop their own internal procedures is estimated to be less than one staff-year, a minimal designation. This estimate is based on the assumption that licensees currently have procedures in place to control j50.59 reviews and submittals and that only minor modifications will be needed to the procedures to check whether the change in CDF exceeds the specified value. Initially, PRA activities will require significant resources, more than two staff-years. These resources include those needed l
to bring their current PRA up to appropriate standard levels. Resources for long term PRA activities are expected to be more than currently required because of the need for ongoing maintenance of the PRA and j
' the increased use ofPRA to support specific G50.59 change activities. The resource requirements for licensee l
training arejudged to be moderate, more than one staff-year but less than two staff-years. Thisjudgment is based on the assumption that most licensees have some PRA capability and would not be required to start i
a PRA group from scratch. Once the new process is in place, no additional inspection or facility changes resources are expected. Nevertheless, it is expected that more Q50.59 changes can be processed owing to the l
risk-informed nature of the option..
. _.. - - -. ~.. _ ~..-
~ _ -.. - - - -.. - - - -.
!j 1;
i
{50.59 Risk-Informed Option, Options The time required for the licensee to develop their internal procedures is estimated to be short, less than six months. This is based on the assumption that the existing procedure would be replace by one that only j
required a PRA calculation and a check of the change in CDF for the change being proposed. In the short j
term, the time required for PRA activities is expected to be long, more than one year. This estin, ate is base j
on the assumption that, for most licensees, it will require more than one year to enhance their PRA such that i
high confidence can be placed in the numerical results from the PRA. In the long term, PRA activities will j
require more time since the PRA will become an important toolin the 50.59 decision making process. The i
time required to implement training is estimated to be short, less than six months. This is based on the assumption that most licensees already have 50.59 processes in place and that only training in how to use the j
information for the PRA to make risk-informed 50.59 decisions would be needed. While it is expected that i
the same amount of resources will be used for inspection and facility changes, the time spent on each individual inspection or change will be less owing to the risk-informed nature of the option.
j Table 6.4.17-1 presents the results from the evaluation of this option against the evaluation criteria described j
in Section 4.
Table 6.4.17-1 Evaluation Resultsfor Core Damage Frequency Option l
l RESOURCE TIAIE
}
EVALUA TION FACTOR RA TING ST LT l
ST LT l
i PRA POLICY STATEMENT IMPLICATIONS 4
j Enhanced Safety Decisions yes i
l More Efficient Use of Agency yes Reduction of Unnecessary Burden yes FACILITY COVERAGE less IMPLICATIONS g
RISK IMPLICATIONS I
l Confidence Required from Risk high j
Results PRA scope medium Completeness les el of detail detailed j
Deterministic vs Risk-Informed risk-informed l
REGULATORY IMPLICATIONS Regulatory Guidance Needed yes minimal medium i
Impact on Other Regulations no N/A N/A i
New RuleMaking Needed yes minimal medium I
impact on Inspection $ raining yes minimal same medium same i
LICENSEE IMPL,1 CATIONS i
I internal Procedures Needed yes minimal shon 1
{ -
i 1
f 50.59 Risk-Informed Option. Options Table 6.4.17-1 Evaluation Resultsfor Core Damage Frequency Option l
RESOURCE TIME EVALUA TION FACTOR RA TING l
l 1.T l
ST l
LT l
ST Impact on PRA Activities yes significant more long more Impact on Traming yes moderate short impact on Inspection yes same same Impact on Facility Changes yes less less Evaluation Using Test Case Under the current 50.59 process, this change was determined to be a USQ because it increased the probability ofoccurrence of a malfunction ofequipment important to safety previously evaluated in the S AR.
Under this proposed option, NRC review will still be required, because the components are designated as safety related, and therefore the change impacts the other regulations. However,it this proposed option were the only basis for deciding whether NRC review would be required, such a review would not be necessary, since loss of the screen wash pumps:
involves no change in CDF for power operation initiators, has no impact on accident sequences in other operational states, and the only external initiators that might pose a problem also fail the essential cooling water pumps.
Evaluation of the Examples Describedin Section 2.3 Against RG 1.174 Option As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the licensee, to be USQs and have required NRC approval in accordance with the current {50.59 process. The potential risk increase interval option was used to evaluate whether these same facility changes would still l
require NRC review and approval. This option did not alleviate the unnecessary burden for all of the examples, since technical specifications were impacted in some cases.
Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System Under the current iS0.59 process, this facility change was determined to be a USQ because it increased the consequences of an accident because doses increased by 1 mrem from the UFSAR stated dose. Under this core damage frequency option, the dose change would be acceptable because the change does not affect (i.e.,
increase) CDF; thus, not requiring NRC approval.
Example 2.3.1.2 Error in Air Volume ofSecondary Containment Under the current 50.59 process, this facility change was determined to be a USQ because it was a reduction in the margin of safety and an increased the consequences of an accident, because the operator thyroid dose increased from 22 rem to 25.6 rem. Under this core damage frequency option, the dose change would represent a change in technical specifications and would still require NRC approval.
l l
l !
-.. ~ _. -.,.. -.. -
f 50 59 Risk-Informed Option. Options Example 2.3.1.3 Calculational Error in Head Loss ofEmergency Core Cooling Suction Strainers Under the current Q50.59 process, this facility change was determined to be a USQ because the lack of enough NPS11 could increase in the probability ofECCS equipment malfunction occurrence. Under this core damage frequency option, changes to technical specifications are involved and the change would still require NRC approval.
Example 2.3.1.4 Modification ofMain Feedwater Valves Under the current 50.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.lE-5 per demand). Under this core damage frequency option, if the change in CDF were less than 1E-7, then NRC approval would not be required.
Example 2.3.1.S Discrepancy Between As-Built System and UFSAR Description Under the current Q50.59 process, this facility change was determined to be a USQ because the as-built system isolation malfunction was of a different type than that previously evaluated in the UFS AR. Under this core damage frequency option, the change would not require NRC approval as long as the licensee determined that the change in CDF was less than IE-7/ year.
Example 2.3.1.6 Replace RHR Service Water Sump Pump Dis ge Check Valve This facility change was not a USQ under the current f 50.59 process ar.D auld also not be a USQ under the core damage frequency option.
)
Example 2.3.1.7 Add ProceduralSteps to Manually Isolate the ControlRodDrive Pumps During
'l SBO l
This facility change was not a USQ under the current G50.59 process and would also not be a USQ under the core damage frequency option.
Example 2.3.1.8 A dd a 3/16 " Diameter Hole Through the Reactor Recirculation Piping Inlet Side i
ofthe Valve Flex-Wedge Disc This facility change was not a USQ under the current 50.59 process and would also not be a USQ under the l
core damage frequency option.
Observations This option could be used to support the objectives of the PRA policy statement and is more risk-informed than the current 50.59. It requires a determination of risk (CDF). In total, it requires significant resources by both the NRC and the licensees to develop and implement the process; however, once in place, the process can result in significant savings.
This option will only provide coverage for class 9 accidents using current PRA information.
For the test case, if the option is the sole basis for determination, it would allow a change in safety designation of the pumps..
r
m 650 59 Risk-Informed Option. Options 6A.18 Option: Large Early Release Frequency Description This option would replace the existing parametric criteria in 50.59 (a)(2)(i) and (a)(2)(ii) with the following:
"If the change in large early release frequency associated with a proposed change, test, or experiment is greater than IE-8/ year."
This option uses the change in large early release frequency (LERF) as a quantitative measure of the risk effects in a risk-informed Q50.59 process. Using this approach, the licensee would evaluate the proposed change to identify the equipment affected by the change, determine the affect on the equipment's failure probability, use the plant PRA to calculate a new LERF reflecting the change in the equipment's failure probability, and then subtract from this new LERF the original LERF(i.e., the LERF before the change was made). Comparison of the resulting change in LERF to acceptable limits would identify whether there is a
" minimal" impact on risk, and hence NRC review of the proposed change is not required. Possible new accidents associated with the proposed change would have to be modeled and included when calculating the new LERF value.
Use of this option requires the establishment of a acceptable " minimal" change in LERF. The acceptable
" minimal" change in LERF is based on acceptance guidelines for LERF provided in RG 1.174, where a
" minima!" change in LERF (i.e., one that will be considered regardless of the total LERF) is defined to be less than 1E-7/ year. Since one of the goals of G50.59 is to identify change > that do not require NRC review, the dennition of" minimal" change in LERF for 50.59 (i.e., one not requiring NRC notification) is 1E-8/ year, a factor of 10 less than the " minimal" change that will be considered by NRC in RG 1.174. Using IE-8/ year as a definition of an acceptable change in LERF ensures that the change in risk associated with a proposed 50.59 change would not violate the risk-acceptance guidelines in RG 1.174.
The nature of this option is such that it only addresses accidents involving the reactor core, i.e., radiological hazards outside the reactor core are not addressed.
Evaluation This option could be used to support the objectives of the PRA Policy Statement. It enhances safety decision making by focusing NRC review on only those changes that are important to LERF, thus using NRC resources more efficiently. It also reduces unnecessary burden on industry by allowing the licensee to identify and minimize requirements associated with components that are not LERF important or which have minimal LERF importance without having to receive NRC approval before the change can be put in place.
I In general, less of the facility will be covered than under the current 50.59. Portions of the plant that do not affect LERF would be excluded. This will climinate some SSCs that are included in the SAR.
j This option makes use of quantitative LERF results. Therefore, a high confidence is needed in the results i
for those changes that are compared to the quantitative criteria. The scope rating is stated as " medium;"
because that level is expected to be needed for a substantial fraction of the changes. That is, external events or low power and shutdown may not need to be considered for many changes. For a particular change, the needed scope can range from minimum to full. The level ofdetail must be commensurate with the change, although a uniform level of detail for all aspects of a PRA may not be necessary for a particular change.
Since change in LERF forms the basis for deciding whether NRC approval is needed or not needed, the option is classified as risk-informed.
74 _
~~
l f 50.59 Risk-Informed Option. Options New Rulemaking and regulatory guidance will be needed to implement this option. While it is expected that this option will have no impact on other regulations, the inspectors will require training and modules to ve compliance.
The resources required to put both the regulatory guidance and Rulemaking in place to support this optio is estimated to be between one and two staff-years; thus, the " minimal" designation for resources. This estimate is based on the fact that resources have already been expended to develop a risk-informed app as described in RG 1.174l and that this option only requires the comparison of change in LERF to some speciGed value. Since this option is expected to have no impact on other regulations, there are no associated resource requirements. The resources required to train the inspectors and to modify the inspections modules is estimated to be less than one staff-year; thus, a " minimal" designation for resources. This estimate is based on the fact that inspectors have, or will soon have, had training on PRA principals, and that most of the resources will be spent on modifying the inspection modules. Once the training and modules are in place, it is expected that this option will require the same amount ofresources to perform the inspections, altho a larger fraction of the resources will be focused on PRA-related activities.
The time required to develop regulatory guidance, put new rules in place, and develop new inspection modules and train the inspectors is estimated to be between six months and a year, a " medium" designation.
Again, these estimates are based on the reasons provided in the above paragraph. Once the training modules are in place, it is expected that the inspectors will spend the same amount of time performing inspections, although a larger fraction of their time will be spent on examining PRA-related information.
The licensees will need to develop their own internal procedures for implementing this option. PRA activities will be impacted due to the need for a high confidence PRA with appropriate scope and level ofdetail. At many plants, additional training will be necessary for the licensee stafT. The nature ofinteractions with inspectors will change to be more focused on PRA-related activities. The potential exists for a significant return on investment for the licensees iflarge numbers ofcost-effective changes are allowed by this process.
The resources required to by the licensee to det elop their own internal procedures is estimated to be less than one staff-year, a minimal designation. His estimate is based on the assumption that licensees currently have procedures in place to control {50.59 reviews and submittals and that only minor modincations will be needed to the procedures to check whether the change in LERF exceeds the speciDed value. Initially, PRA activities will require significant resources, more than two staff-years. These resources include those needed to bring their current PRA up to appropriate standard levels. Resources for long term PRA activities are expected to be more than currently required because of the need for ongoing maintenance of the PRA and the increased use ofPRA to support specific 50.59 change activities. The resource requirements for licensee training arejudged to be moderate, more than one staff-year but less than two staff-years. Thisjudgment is based on the assumption that most licensees have some PRA capability and would not be required to start a PRA group from scratch. Once the new process is in place, no additional inspection or facility changes resources are expected. Nevertheless,it is expected that more {50.59 changes can be processed owing to the risk-informed nature of the option.
The time required for the licensee to develop their internal procedures is estimated to be short, less than six months. This is based on the assumption that the existing procedure would be replace by one that only required a PRA calculation and a check of the change in LERF for the change being proposed. In the short term, the time required for PRA activities is expected to be long, more than one year. This estimate is base on the assumption that, for most licensees, it will require more than one year to enhance their PRA such that high con 6dence can be placed in the numerical results from the PRA. In the long term, PRA activities will require more time since the PRA will become an important tool in the {50.59 decision making process. The time required to implement training is estimated to be short, less than six months. This is based on the
._.m
- _ _ _. _ _ _ _. ~ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _ _. _ _ _ _ _ _ _ _.. _ _.. _ _
1 j
650.59 Risk-Informed Option, Options assumption that most licensees already have 50.59 processes in place and that only training in how to use the information for the PRA to make risk-informed 50.59 decisions would be needed. While it is expected that the same amount of resources will be used for inspection and facility changes, the time spent on each individual inspection or change will be less owing to the risk-informed nature of the option.
ll j
Table 6.4.18-I presents the results from the evaluation of this option against the evaluation criteria described in Section 4.
1 Table 6.4.18-1 Evaluation Resultsfor Large early releasefrequency Option El'ALUA TION FACTOR RATING ST LT ST LT PRA POLICY STATEMENT IMPLICATIONSMyes Enhanced Safety Decisions l
More Efficient Use of Agency yes Reduction of Unnecessary Burden yes FACILITY COVERAGE less I
IMPLICATIONS RISK IMPLICATIONS Mhigh Confidence Required from Risk Results PRA scope medium j
Completeness level of detail detailed Deterministic vs Risk informed risk-informed REGULATORY IMPLICATIONS 2
Regulatory Guidance Needed yes minimal medium Impact on Other Regulations no N/A N/A 4
New Rulemaking Needed yes minimal medium a
impact on Inspectionfirainhg yes minimal same medium same J
LICENSEE IMPLICATIONS Internal Procedures Needed yes minimal short impact on PRA Activities yes significant more long more I
impact on Training yes moderate short impact on Inspection yes same same Impact on Facility Changes yes less less i q
{50.59 Risk-Informed Option, Options Evaluation Using Test Case Under the current Q50.59 process, this change was determined to be a USQ because it increased the probability ofoccurrence ofa malfunction ofequipment important to safety previously evaluated in the SAR.
Under this proposed option, NRC review will still be required, because the components are designated as safety-related, and therefore the change impacts the other regulations. However, it this proposed option were the only basis for deciding whether NRC review would be required, such a review would not be necessary, since loss of the screen wash pumps involves no change in LERF for power operation initiators,
=
has no impact on accident sequences in other operational states, and the only external initiators that might pose a problem also fail the essential cooling water pumps.
a Evaluation ofthe Examples Describedin Section 2.3 Against RG 1.174 Option As stated in Section 2.3, the examples listed represent facility changes that have been determined, by the i
licensee, to be USQs and have required NRC approval in accordance with the current Q50.59 process. The potential risk increase interval option was used to evaluate whether these same facility changes would still require NRC review and approval. This option did not alleviate the unnecessary burden for all of the i
examples, since technical specifications were impacted in some cases.
Example 2.3.1.1 Error in Dose Calculationsfor the Process Gas System Under the current Q50.59 process, this facility change was determined to be a USQ because it increased the consequences of an accident because doses increased by I mrem from the UFSAR stated dose. Under this large early release frequency option, the dose chatige would be acceptable because the change does not affect (i.e., increase) LERF; thus, not requiring NRC approval.
Exangpie 2.3.1.2 Error in Air Volume ofSecondary Containment 1
Under the current G50.59 process, this facility change was determined to be a USQ because it was a reduction in the margin of safety and an increased the consequences of an accident, because the operator thyroid dose increased from 22 rem to 25.6 rem. Under this large early release frequency option, the dose change would represent a change in technical specifications and would still require NRC approval.
Example 2.3.1.3 CalculationalError in HeadLoss ofEmergency Core Cooling Suction Strainers Under the current 50.59 process, this facility change was determined to be a USQ because the lack of i
enough NPSH could increase in the probability ofECCS equipment malfunction occurrence. Under this large i
early release frequency option, changes to technical specifications are involved and the change would still require NRC approval.
Exanqple 2.3.1.4 Modification ofMain Feedwater Valves Under the current G50.59 process, this facility change was determined to be a USQ because the modification resulted in an increase in the probability ofequipment malfunction occurrence (failure of feedwater isolation increased from 2.8E-5 to 6.1 E-5 per demand). Under this large early release frequency option, ifthe change in LERF were less than IE-8, then NRC approval would not be required. P I
^ '%
U r*
{50.59 Risk. Informed Option, Options Example 2.3.1.5 Discrepancy Between As-Built System and UFSAR Description Under the current G50.59 process, this facility change was determined to be a USQ because the as-built system isolation malfunction was of a different type than that previously evaluated in the UFSAR. Under this large early release frequency option, the change would not require NRC approval as long as the licensee determined that the change in LERF was less than IE-8/ year.
Example 2.3.1.6 % Replace RHR Service Water Sump Pump Discharge Check Valve This facility change was not a USQ under the current 650.59 process and would also not be a USQ under the large early release frequency option.
Example 2.3.1.7 Add Procedural Steps to Manually Isolate the Control Rod Drive Pumps During SBO This facility change was not a USQ under the current 50.59 process and would also not be a USQ under the large early release frequency option.
Exangle 2.3.1.8 Adda 3/16" Diameter Hole Throngh the ReactorRecirculation PipingInletSide ofthe Valve Flex-Wedge Disc This facility change was not a USQ under the current 50.59 process and would also not be a USQ under the large early release frequency option.
Observations This option could be used to support the objectives of the PRA policy statement and is more risk-informed
.than the current j50.59. It requires a determination ofrisk (LERF). In total, it requires significant resources by both the NRC and the licensees to develop and implement the process; however, once in place, the process can result in significant savings.
This option will only provide coverage for class 9 accidbts using current PRA information.
For the test case, if the option is the sole basis for determination, it would allow a change in safety designation of the pumps.
f I,
.