ML17214A700

Attachment 2: Method of Evaluation Guidance Extract - Track Changes (7/05/2017, E-Mail from K. Cummings/Nei to R. Wharton/Nmss/Dsfm 72.48 and Method of Evaluation)

ML17214A700
Person / Time
Site:	Nuclear Energy Institute
Issue date:	07/05/2017
From:	Cummings K Nuclear Energy Institute
To:	Wharton L NRC/NMSS/DSFM/IOB
Wharton L
Shared Package
ML17214A697	List: ML17214A698 ML17214A699 ML17214A700
References
NEI 12-04
Download: ML17214A700 (22)

Text

DEFINITIONS 3.8 DEPARTURE FROM A METHOD OF EVAL ATION DESCRIBED IN THE FSAR (AS UPDATED) USED lN ESTABLISHING THE DESIGN BASES OR IN THE SAFETY ANALYSES Definition (§72.48(a)(2)):

Departure from a method of evaluation described in the FSAR (as updated) used in establishing the design bases or in the safety analyses means (i) changing any of the elements of the method described in the FSAR (as updated) unless the results of the analysis are conservative or essentially the same; or (ii) changing from a method described in the FSAR to another method unless that method has been approved by NRC for the intended application.

Discussion:

The 10 CFR 72.48 definition of "departure .. . " provides licensees and CoC ho lders with the flexibility to make changes in methods of evaluation that are "conservative" or that are not important with respect to demonstrating that SSCs can perform their intended design functions. See also the definition and discussion of " method of evaluation" in Definition 3.17. Guidance for evaluating changes in methods of evaluation under criterion I 0 CFR 72.48(c)(2)(viii) is provided in Section 6.8.

Conservative vs. Non-Conservative Evaluation Results Gaining margin by revising an e lement of a method of evaluation is considered to be a non-conservative change and thus a departure from a method of evaluation for purposes of I 0 CFR 72.48. Such departures require prior NRC approval of the revised method . In other words, analytica l results obta ined by changing any element of a method are "conservative" relative to the previous results, if they are closer to design bases limits or safety analyses limits (e.g., applicab le acceptance guide lines).

For examp le, a change in an e lement of a method of evaluation that changes the result of a cask peak pressure analysis from 45 psig to 48 psig (with design basis limit of 50 psig) would be considered a conservative change for purposes of 10 CFR 72.48(c)(2)(viii).

Thi is because results closer to limiting values are considered conservative in the sense that the new ana lysis result provides less margin to applicable limits for making future physical or procedure changes without a license or CoC amendment.

!fuse ofa modified e lement of a method of evaluation resulted in a change in calculated cask peak pressure from 45 psig to 40 psig, this wou ld be non-conservative . This is because the change wou ld result in more margin being avai lable (to the design basis limit of 50 psig) for a licensee or CoC holder to make future changes to the cask design or procedures that once again reduce the margin.

"Essentially the Same" Licensees and CoC holders may change one or more elements of a method of evaluation such that results move in the conservative or non-conservative direction without prior NRC approval , provided the results are "essentially the same" as the previous resul _

from the MOE without the modified element(s) . Results are "essentially the same" if they are within the margin of error for the type of analysis being performed. Variation in results due to routine analysis sensitivities or calculational differences (e.g. , rounding errors and use of different computational platforms) would typically be within the analysis margin of error and thus considered "essentially the same."

"Approved by the NRC for the Intended Application" Rather than make a minor change to an existing method of evaluation , a licensee or CoC holder may also adopt ~completely new methodology without prior NRC approval provided the new method is approved by the NRC for the intended application. A new method is " approved by the NRC for the intended application" if it is approved for the type of analysis being conducted and the licensee or CoC holder satisfies applicable terms and conditions for its use. More detailedSpecific guidance for making this determination is provided in Section 6.8.2 .

3.15 INPUT PARAMETERS Definition:

Input parameters are those values derived directly from the physical characteristics of SSCs or processes in the ISFSI facility or cask design , including flow rates, temperatures, pressures, dimensions or measurements (e.g. , vo lum e, weight, size, etcJ, and system response times.

Discussion:

The principal intent of this definition is to distinguish methods of evaluation from evaluation input parameters. Changes to methods of evaluation described in the UFSAR (see Definition 3.17) are evaluated under criterion 10 CFR 72.48(c)(2)(viii), whereas changes to input parameters described in the UFSAR are considered changes to the ISFSI facility or cask design that would be evaluated under the other seven criteria of I 0 CFR 72.48(c)(2), but not criterion (c)(2)(viii).

There are two cases in which an input parameter is treated as an element within a method of evaluation. For guidance regarding these two cases, refer to Definition 3. 17.

Examples illustrating the treatment of input parameters are provided in Appendix A.

Appendix A also provides examples to describe the specific elements of evaluation methodology that would require evaluation under I 0 CFR 72.48(c)(2)(viii) and to clearly distinguish these from specific types of input parameters that are controlled by the other seven criteria of 10 CFR 72.48( c)(2).

3.17 METHOD OF EVALUATION Definition:

Method of evaluation means the calculational framework used for evaluating behavior or response of the JSFSI facility, cask design , or an SSC.

Discussion:

Examples of elements of methods of evaluation are presented below. Proposed activities involving modifications to such methods of evaluation require evaluation under I 0 CFR 72.48(c)(2)(viii) only for methods of evaluation used either in UFSAR safety analyses or in establishing the design bases, and only if the methods are described , outlined or summarized in the UFSAR. Proposed activities involving modifications to methods of evaluation that are subject to I 0 CFR 72.48 include changes to elements of existing methods described in the UFSAR and to changes that involve replacement of existing methods of evaluation with alternative methodologies.

Elements of Methodology Example Data correlations

• Tipover and end drop analysis based on industry-referenced reports Means of data reduction

• ASME methods for evaluating cask parameters, e.g.,

elastic stress intensity Physical constants or coefficients

• Heat transfer coefficients, boundary conditions, burnup peaking factors, and cross-section libraries

• Friction coefficient in a tipover analysis Mathematical models

• Methods of heat transfer assumed (i.e. , conduction ,

convection, and radiation), mesh size, material performance - elastic or plastic Specific limitations of a

• Benchmarking and correlation ranges computer program Specified factors to account for

• Criticality calculation biases, fuel burnup, percent uncertainty in measurements or fuel burnup uncertainty data Statistical treatment of resu Its

• Vendor-specific analysis approach (e.g. , material property testing of composite neutron poison , B-10 content validation from blackness testing, confidence and tolerance limits)

Dose conversion factors and

• lCRP factors assumed source terms

Methods of evaluation described in the UFSAR subject to criterion l 0 CFR 72.48(c)(2)(viii) are:

• Methods of evaluation used in analyses that demonstrate that design basis limits of fission product barriers are met (i.e. , for the parameters subject to criterion I 0 CFR 72.48( c)(2)(vi i)).

• Methods of evaluation used in UFSAR safety analyses, including cask and accident analyses typically presented in the accident analyses section(s) of the UFSAR, to demonstrate that consequences of accidents do not exceed I 0 CFR 72.106 dose limits.

• Methods of evaluation used in supporting UFSAR analyses that demonstrate intended design functions will be accomplished under design basis conditions that the ISFSI facility and cask design are required to withstand , including natural phenomena, environmental conditions, and dynamic effects.

• Methods of evaluation used in UFSAR analyses that demonstrate that radioactive doses from normal operations and anticipated occurrences will be within the limits of 10 CFR 72 . 104.

• Methods of evaluation subject to criterion 10 CFR 72.48(c)(2)(viii) do not include methods used to select fuel for loading into the cask (i.e. , decay heat and burnup determinations, and determination of damaged fuel versus undamaged or intact fuel, etc.) unless those methods are described in the UFSAR.

If a methodology permits the licensee or cask certificate holder to establish the value of an input parameter on the basis of ISFSI facility- or cask design-specific considerations, then that value is an input to the methodology, not part of the methodology. On the other hand, an input parameter is considered to be an element of the methodology if:

• The method of evaluation includes a methodology describingdescription of how to select the value of an input parameter to yield adequately conservative results.

However, if a licensee or cask certificate holder opts to use a value more conservative than that required by the selection method, reduction in that conservatism should be evaluated as an input parameter change, not a change in methodology.

• The development or approval of a methodology was predicated on the degree of conservatism in a particular input parameter or set of input parameters. In other words, if certain elements of a methodology or model were accepted on the basis of the conservatism of a selected input value, then that input value is considered an element of the methodology.

5. 72.48 SCREENING Once it has been determined that I 0 CFR 72.48 is applicable to a proposed activity via the AD process, a screening is performed to detennine ifthe activity should be reviewed against the evaluation criteria of I 0 CFR 72.48(c)(2).

Engineering, design and other technical information concerning the activity and affected SSCs should be used to perform the I 0 CFR 72.48 screening. The I 0 CFR 72.48 screening i perfonned to determine whether the activity or part(s) thereof need to be evaluated in accordance with I 0 CFR 72.48(c)(2) and Section 6 of this guidance. Refer to Section 1.3 for discussion of the use of the word "change" in the context of I 0 CFR 72.48. A full I 0 CFR 72.48 evaluation is required for a proposed activity or part(s) thereof that involve any one of the following:

• A change to a UFSAR-described design function of an SSC or cask design ,

• A change to a UFSAR-described method of performing or controlling a design function ,

• A change to a UFSAR-described method of evaluation or use of an alternative method of evaluation for demonstrating that intended design functions will be accomplished , or

• A test or experiment not described in the UFSAR where an SSC is utilized or controlled in a manner that is outside the reference bounds of the design for that SSC or is inconsistent with analyses or descriptions in the UFSAR.

5.1.5.J Determining if an Activity Involves an MOE The following step-by-step guidance may be used to determine if a proposed activity involves an MOE :

The discussion that follows is organized into four distinct steps:

Step 1 - Distinguish between input parameters and MOEs Step 2 - Determine if an MOE is " ...described , outlined or summarized in the UFSAR."

Step 3 - Determine if the MOE is used for one of the three cited purposes Step 4 - Identification of intended design functions under design basis conditions Each of these steps is discussed in detail below.

Step 1 - Distinguis h Between Input Parameters and Methods of Evaluation This step involves application of two separate definitions. They are :

Definition 3.15 - " Input Parameters" Definition 3.17 - " Method of Evaluation"

The core de finition s for each, a long with explanatory paragraph s are provided be low, with emphasis added :

Input Parameters:

Input parameters are those values derived directly from the physical characteristics of SSCs or processes in the ISFSI facility or cask design, including flow rates, temperatures, pressures, dimensions or measurements (e.g., volume, weight, size, etc.), and system response times.

The principal intent of this definition is to distinguish methods of evaluation from evaluation input parameters. Changes to methods of evaluation described in the UFSA R are evaluated under criterion 10 CFR

72. 48(c)(2)(viii), whereas changes to input parameters described in the UFSA R are considered changes to the ISFSI facility or cask design that would be evaluated under the other seven criteria of 10 CFR 72. 48(c)(2),

but not criterion (c)(2)(viii).

Method of Evaluation:

Method of evaluation means the calculational framework used for e valuating behavior or response of the !SFSI fac ility, cask or an SSC.

... an input parameter is considered to be an element of the methodology if:

• The method of evaluation includes a methodology describing how to select the value of an input parameter to y ield adequately conservative results. However, if a licensee opts to use a value more conservative than that required by the selection method, reduction in that conservatism should be evaluated as an input parameter change, not a change in methodology.

• The development or approval of a methodology was predicated on the degree of conservatism in a particular input parameter or set of input parameters. In other words, if certain elements of a methodology or model were accepted on the basis of the conservatism of a selected input value, then that input value is considered an element of the methodology.

There are exampl es and an ex tended di scuss ion provi ded for eac h of the above elsewhere in this guidance. However, a few points/observations may be useful:

• Input parameters are va lues. Those va lues are derived from physical c haracteristics of SSCs or a process.

• MOEs are the "calcu latio nal framework ." The exa mpl es in Defi nition 3. 17 illustrate that MO Es tend to invol ve some type of math ematica l equations or are re lated to physica l constants of nature.

18&-tn many cases, a simple inspection of whether the topic of consideration is a va lue, a constant of nature, or some fom1 of a mathematical expression would be insightful.

The two definitions, 3. 15 and 3. 17, must be read in their entirety . The stated purpose is to distinguish input parameters from MOEs. This is because the treatment under I 0 CFR 72.48 is entire ly different for input parameters and MO Es. The screening criteria are different and, as noted above, the I 0 CFR 72.48 criteria to be answered are mutua lly exc lusive.

Criteria I through 7 are answe red for input parameters but not for MOEs.

Criterion 8 is so lely for MOEs that require such a review, including the two cases described above in which a n input parameter is considered to be an element of the MOE. The remaining three steps described next wi ll determine if a given modification to an MOE requires a I 0 CF R 72.48 review.

Note that any calcu latio nal framework could potentially satisfy the meaning of "Method of Eva luation" in Definition 3 . 1 7~ This creates a possible source of confu sion because it is common to use the term " MOE change" to mean that any modification to an MOE requires a I 0 CFR 72.48 eva luati on. However, Steps 2 and 3 ind icate that two more attributes are required to achieve that status as described in the " Discussion " section of Definition 3. 17. They are:

• The MOE is described in the UFSA R (Step 2)

• The MOE is subject to 10 CF R 72.48 criterion (c)(2)(viii) review (Step 3)

Step 2 - Determine if an MOE is " ... described, outlined or summarized in the UFSAR" The paragraph entitled "Discussion" from Definition 3. 17 is provided below, with emphas is added:

Examples of methods of evaluation are presented below. Proposed activities involving modifications to such methods of evaluation require evaluation under 10 CFR72. 48(c)(2)(viii) only for methods of evaluation used either in UFSAR safety analyses or in establishing the design bases, and only if the methods are described, outlined or summarized in the UFSAR . Proposed activities involving modifications to methods of evaluation that are subject to 10 CFR 72. 48 include changes to elements of existing methods described in the UFSAR and to changes that involve replacement of existing methods of evaluation with alternative methodologies.

Proposed modifications to such methods of evaluation require evaluation under I 0 CFR 72.48(c)(2)(viii): on ly for:

• evaluations used either in UFSAR safety anal yses or in establi shing the design bases, and

• only if the methods are described, outlined or summarized in the UFSAR.

Step #2 simply identifies if the method is " .. . described , outlined or summarized in the UFSAR."

The intention here is if the MOE was discussed in any fa shion, then the MOE is con sidered to be "described in the UFSAR."

Step 3 - Determine if the MOE is used for one of the three cited purposes The second required feature is that the MOE must be " .. . used either in UFSAR safety analyses or in establishing the design bases .... "

These purposes correspond to the language used in the defined term of I 0 CFR 72.48(a)(2), which is repeated here:

Departure from a method of evaluation described in the FSAR (as updated) used in establishing the design bases or in the safety analyses Definition 3. 17 expands upon the meaning of these purposes. The followin g phrases explicitly describe these three purposes :

Methods of evaluation described in the UFSAR subject to criterion 10 CFR 72. 48(c)(2)(viii) are:

• Methods of evaluation used in analyses that demonstrate that design basis limits offission product barriers are met (i.e., f or the parameters subject to criterion 10 CFR 72. 48(c)(2)(vii))

• Methods of evaluation used in UFSAR safety analyses, including confinement and accident analyses, to demonstrate that consequences of normal operations, off-normal events, and accidents do not exceed 10 CFR 72. 104 or 10 CFR 72.106 dose limits, as applicable

• Methods of evaluation used in supporting UFSAR analyses that demonstrate intended design functions will be accomplished under design basis conditions that the lSFSl or cask is required to withstand, including natural phenomena, environmental conditions, and dynamic effects.

The three purposes can be summarized as MOEs:

I. Results demon strate that design basis limits of fi ssion product barriers are met

2. Used to calculate consequences (on-site and off-site accident dose)

3. That demonstrate intended design functions will be accomplished under design basis conditions Items I and 2 above sho uld be se lf-evident to any I 0 CFR 72.48 evaluator involved in such activities. Item 3 includes two embedded terms, each with their own extended source of guidance. Identification of this usage is the subject of Step 4.

Step 4 - Identification of Intended Design Functions under Design Basis Conditions.

Design function is a critical concept that is used throughout this guidance (Definition 3. 10).

There are two points to be made here:

I. The definition for design function is rather lengthy and is heavily oriented around design bases functions and those functions that support or impact design bases functions.

2 . The term design bases functions comes from NEI 97-04, which is endorsed in Regulatory Guide l . 186.

This discussion will not expand further on the meaning of these two critica l terms, design function and design bases functions , other than to note that a complete understanding of both is required to fully understand the identification of MO Es subject to review under I 0 CFR 72.48. The meanings of design function and design bases functions are included in Definition 3.10.

5.1.5.2 Software Revisions Associated with an MOE This section discusses a revision to existing software that implements an MOE. A change to the software that implements an MOE does not necessarily cause a departure from an MOE, requiring prior NRC approval. Therefore, it is critical to understand the scope and type of changes that were made.

NOTE: This discussion does NOT address the replacement of, or implementation of new, MO Es or software.

Step I - Does the Software Need to be Considered?

The first step in determining the impact of the software revision is to determine if the software fits the definition of an MOE and is, therefore, within the scope of consideration . A discussion of the identification of applicable MOEs is included in Section 5.1 .5.1 .

Step 2 - Performing the Screening Review If the software does meet the criteria for an MOE, then the next step is to understand the scope and type of modification(s) involved. There are many elements to a software package. Determining exactly which e lements are being revised is critical. Examples of elements of methodology are given in Definition 3.17. A simple statement regarding the " revision ," "version" or " modification" identifier as the basis for a I 0 CFR 72.48 screening response is inadequate.

One CoC holdervendor/licensee may use several "versions" of a computer code revision to address errors and minor improvements, thereby saving a new revision for major modeling updates, while another vendor/ licensee may change

" revisions" of a computer code to address a number of minor errors without changing any analytical modeling. Use of a later version or release of a computer code that is already part of an MOE described in the UFSAR would be evaluated as a change to an element of a method of evaluation.

5.1.5.3 Additional Concerns A proposed activity involving an MOE is a change (i.e., screens in) if the modification is not in strict accordance with the constraints and limitations outlined in an NRC Safety Evaluation Report (SER), vendor/ licensee topical report, or the UFSAR (hereafter referred to as "source documents"). A proposed activity to replace an MOE with an alternate MOE (i.e., different software package) always screens in.

The technical description of the MOE in the source documents defines any constraints and limitations on use of the MOE. For example, if a source document for a lattice physics analytical model describes its application to a particular fuel design (e.g., Westinghouse I 5x I 5), the specific reference to the Westinghouse I 5x 15 fuel design shall be viewed as a constraint or limitation on the use of the analytical model unless the source document states that the analytical model may be used for other fuel designs.

For £!...proposed modification to an element of an MOE, it is essential to identify and understand the details of the modification.

_For each modification , the pertinent constraints and limitations associated with the MOE, if any, need to be identified.

Modifications to more than one element may need to consider the cumulative impact of all the modifications on the constraints and limitation . In these cases, the modifications to the MOE may be a " replacement MOE" rather than a

" modification to an element of an MOE."

The I 0 CFR 72.48 screening should identify if a proposed modification to one or more elements of an MOE that is not consistent with the constraints and limitation s affects an element of the MOE or effectively causes the MOE to

become an alternative MOE. This distinction is necessary to correctly apply the Evaluation guidance in Section 6.8.

The following categories of proposed modifications to an element of an MOE are not adverse and would not require a full 72.48 evaluation:

Modifications to an element of an MOE that are administrative, such as changing input/output descriptive labels, changing output table titles, adding/deleting intermediate output results, re-sequencing output tables, adding non-executable comments in the computer coding, etc., would not be considered changes and would not require a I 0 CFR 72.48 evaluation . Note that although these types of modifications to an MOE might be considered editorial or administrative and not require a 72.48 screening (per Subsection 4.3 of this guidance), it is recommended that a I 0 CFR 72.48 screening be performed to provide the explanation that the changes are indeed editorial or administrative and have no impact on the MOE.

Modifications to MOEs that are within the constraints and limitations require case-specific consideration to determine if the modification is a change to an element of an MOE or a different MOE . Typical constraints and limitations may include the following :

Breathing rate of 3.47 E-4 m3/sec from an NRC regulatory guide for inhalation dose calculations Use of dose conversion factors from an ICRP standard Fractional release values for confinement analysis from NUREG-1536 Heat transfer correlations

- Analysis performed "consistent with" a cited topical report.

- Neutron absorber blackness is "appropriately modeled."

- A subroutine iterates to a specified convergence limit

- A boundary condition is set to a specified value.

- Cross sections were obtained by collapsing the library from "x" groups to "y" groups.

If the pertinent constraints and limitations for an MOE are not known or cannot be identified, then the modification to the element of the MOE is considered to be adverse and screens in .

If a source document identifies that a particular feature is included in an MOE, but does not describe how the feature is modeled , a modification to the specific modeling of the feature is not adverse because the modification is consistent with

the terms (and level of detail) of the approved MOE. A full 10 CFR 72.48 evaluation would not be requ ired. However, if a source document states that a particular feature is modeled, a modification to eliminate that feature would not be consistent with the description in the source document and would be adverse~

requiring a fu ll I 0 CFR 72.48 eva luation .

lf an MOE contains a built in mechanism for making modifications to the method of evaluation, then modifications to the MOE made in accordance with the built in change mechanism are not adverse. For examp le, a fuel vendor may have a process for licensing new fuel design that explicitly includes a way of determining ne'N coefficients for the critical power correlation, based on ne*N data. Therefore, new coefficients calculated using the new data are not changes to physical coefficients in the conte>(t of I 0 CFR 72.48 (i.e., adverse) because they are developed using a previously approved mechanism for ca lculating the coefficients.

See Section 6.8 for guidance on completing the l 0 CFR 72.48 evaluationJQ determine if a departure from a method of evaluation is involved .

6.8 DOES THE ACTIVITY RESULT IN A DEPARTURE FROM A METHOD OF EVALUATION USED IN ESTABLISHING THE DESIGN BASES OR IN THE SAFETY ANALYSES?

The cask or ISFSJ or cask UFSAR contains design and licensing basis information for an ISFSI facility or spent fuel storage cask design, including descriptions of how regulatory requirements for design are met (such as the requirements governing normal operations and off-normal eventsanticipated occurrences), and the adequacy of structures, systems, and components provided for the prevention of accidents and the mitigation of the consequences of accidents. Analytical methods are a fundamental part of demonstrating how the design meets regulatory requirements and why the ISFSJ facility ' s or cask ' s response to accidents and events is acceptable. As such , in cases where the analytical methodology was considered to be an important part of the conclusion that the lSFSI facility or cask met the required design bases, these analytical methods were described in the UFSAR and received varying levels ofNRC review and approval during licensing.

Because l 0 CFR 72.48 provides a process for determining if prior NRC approval is required before making changes to the JSFSI facility or spent fuel storage cask design as described in the UFSAR, changes to the methods of evaluation (MO Es) described in the UFSAR may-also fall under the provisions of the I 0 CFR 72.48 process, specifically criterion (c)(2)(viii). In general, licensees or cask certificate holders can make changes to MO Es elements of an MOE without first obtaining a license amendment or cask CoC amendment in one of two ways:

I. Changing one or more elements of the MOE, providedtf the results are essentially the same as, or more conservative than , previous results.

2. Using a new or different MOE, Similarly, licensees or cask certificate holders can also use new or different MOEs without first obtaining a license or cask CoC

amendment if th~ese MO Es ha~ve been approved by the NR for the intended application.

The responsibility and authority for performing the I 0 CFR 72.48 review for a cask design approved under a CoC can be either the general licensee or the CoC holder. I 0 CFR 72.48 responsibility and authority for a specific license are solely the speci fie licensee's. In both cases, however, it is typically the cask design authority that utilizesH-SeS MOEs to perfonn the safety analyses or establish the cask design basis and would have the need to modify an element of an MOE or use a new or different MOE.

As such, the licensee and cask design authority must coordinate efforts closely to ensure roles and responsibilities are clearly delineated and information is shared appropriately to ensure a complete and comprehensive I 0 CFR 72.48 review is perfonned for any change to an MOE.

If the I 0 CFR 72.48 screening for the proposed activity concludes that does not involve a change to an MOE is not involved , then the I 0 CFR 72.48 evaluation screening should reflect that this criterion I 0 CFR 72.48(c)(2)(viii) is not applicable. _If the I 0 CFR 72.48 screening determines that the proposed activity involves only a change to an MOE, then the I 0 CFR 72.48 evaluation should reflect that criteria I 0 CFR 72.48( c)(2)(i- vii) are not applicable.

If a proposed activity (e.g., a physical modification or procedure change) also involves a modified MOE element or a new or different MOE to analyze that physical modification or procedure change, the MOE change should be evaluated first, against I 0 CFR 72.48(c)(2)(viii), before the physical modification or procedure change is evaluated against I 0 CFR 72.48(c)(2)(i) through (vii). This is recommended because if the MOE change requires NRC approval, the licensee or CoC holder may wish to reconsider the phy ical modification or procedure change. If the MOE change is determined not to require NRC approval, these two separate evaluations may be documented in the same, or separate overall I 0 CFR 72.48 reviews, at the discretion of the licensee or CoC holder.

The first step in applying this criterion is to identify the MOEs that are affected by the change. This is accomplished during application of the screening criteria as described in Section 5.1.5.

Next, the licensee or cask CoC holder must determine whether the change constitutes a departure from a method of evaluation that would require prior NRC approval. As discussed further below, for purposes of evaluations under this criterion, the following changes are considered a departure from a method of evaluation described, outlined, or summarized in the UFSAR and used in the safety analysis or establish design bases :

• Changes to any element of an MOE that yield results that are not conservative or not essentially the same as the results from the analyses of record.

• Use of a new or different MOE that is not approved by NRC for the intended application.

By way of contrast, the following changes are not considered departures from a method of evaluation:

• Departures from methods of evaluation that are not described , outlined or summarized in the UFSAR (such changes will have been screened out as discussed in Section 5.1.5);

• Use of a new NRC-approved methodology (e.g. , new or upgraded computer code) to reduce uncertainty, provide more precise results, or other reason, provided such use is (a) based on sound engineering practice, (b) technically appropriate for the intended application, and (c) within the limitations of the applicable SER. The basis for this determination should be documented in the licensee or cask CoC holder evaluation.

See Subsection 6.8.2 for additional guidance on interpreting whether an MOE is approved for the intended application.

• Use of a methodology revision that is documented as providing results that are essentially the same as or more conservative than either the previous revision of the same methodology or with another methodology previously accepted by NRC through issuance of an SER.

• Use of a methodology that is described in the UFSAR, but has not been spec ifically approved by the NRC either through a Topical Report review or through endorsement in the storage system or ISFSI SER to evaluate the impact ofsupport a change that modificationes_!Q the ISFSI facility, cask design, or proceduresinput parameters. For example, the UFSAR describes the methodology, inputs, assumptions, and results of used for the heat tran sfer evaluations of the storage system. The methodology was never submitted to the NRC for approval in a Topical Report, and the storage system SER does not indicate whether the NRC has endorsed or approved the methodology.

In this case, use of the methodology described in the UFSAR to evaluate the impact of support a modification to the LSFSI facility, cask design or procedurechange would NG+-be acceptable and appropriateresult in a departure from a method of evaluation described in the UFSAR .:2 Subsection 6.8.1 provides guidance for making changes to one or more elements of an existing MOE. Subsection 6.8.2 provides guidance for adopting an entirely new MOE to replace an existing one.

It should be noted that the NRC staff, in reviewing dry cask storage designs, does historically has not generically approved methodologies described or referenced in UFSARs_,__ for use by other licensees or vendors. "Generic" approval as used here, means approval for use by any licensee or CoC holder. Instead , the NRC states in its SERs, following the guidance in the Standard Review Plan , that the design has been found to be acceptable in each review discipline area for the particular licensing action reviewed and approved (i.e., initial ISFSL license, cask CoC, or amendment thereto) . See Subsection 6.8.2. l for additional guidance on NRC approval of methods.

If, however, CoC holdersvendors or licensees choose to submit detailed methodologies to the NRC for gel)eric review and approval as part of app li cations for design approval or as separate topical reports, the staff may document NRC endorsement or approval in appropriate SERs on a broader basis. Such end orsements or approva l will faci litate vendors CoC holders and li censees to use the I 0 CF R 72.48 process that deals with approved methodologies .It is important to note, however, that *Nhile e"plieit NRG approval of the MO Es may not appear in the ISFSI or eask SER, the MO Es are NRG approved for that partieular eask system or ISFSI deseribed in the CoC or lieense, to the e>(tent the MOEs are used as deseribed in UFSAR. This is beeause the UFSAR is the basis for the eask CoC or ISFSI lieense and is approved as part of the lieensing proeess when the CoC or lieense is granted. Extrapolation of NRG approval of MOEs beyond the description of the methods in the UFSAR should be avoided, unless the NRG has described sueh extrapolation in the SER supporting that CoC or license. Departures from those UFSAR described MOEs, as defined herein, require NRG approval if those methods are used in the safety analysis or to demonstrate the cask system/ISFSI ean perform its design function.

6.8.1 Guidance for Changing One or More Elements of a Method of Evaluation The definition of " departure ... "provides licensees and CoC holders with the flexibility to make changes under I 0 CFR 72.48 to e lements of MO Es whose results are "conservative" or that are not important with respect to the demonstrations of performance that the analyses provide. Changes to elements of MO Es that yield conservative results or results that are essentiall y the same over the entire range of use for the method wo uld not be departures from approved MOEs.

The guidance is summarized in the fo llowing table.

Margin Results Prior NRC Approval Lost Conservative No Gained Non-Conservative Yes

- Same Essentia lly The Same No One common use of this part of the MOE evaluation is the adoption of a later version of a computer code previously used as an MOE as described in the ISFSI or cask FSAR. In such cases, the MOE user must first verify and validate the updated computer code as acceptable for use and code users are qualified as required by the governing Quality Assurance program. Then, the MOE user must use the updated computer code to analyze one or more benchmark cases that had been analyzed using the previous version of the computer code. The results of old and new sets of benchmark cases are then compared to detennine if the updated computer code produces results that are conservative, non-conservative, or essentially the same for the same benchmark case(s).

To determine if the new results are conservative, non-conservative or essentiall y the same, the guidance in Sections 6.8.1.1 and 6.8. 1.2 is app lied .

6.8.J.J Conservative versus Non-Ceonservative Results Gaining margin by changing one or more elements of an MOE is considered to be a non-conservative change and thus a departure from an MOE for purposes of I 0 CFR 72.48.

Such departures require prior NRC approva l of the revised MOE. Analytical results obtained by changing any element of an MOE are "conservative" relative to the previous results, if they are closer to design bases limits or safety analyses limits (e.g., applicable acceptance guidelines). For examp le, a change from 45 psig to 48 psig in the result of a cask peak pressure analysis (with design basis limit of 50 psig) using a revised MOE would be considered a conservative change when applying this criterion. In other words, the revised MOE is more conservative if it predicts more severe conditions given the same set of inputs. This is because results closer to limiting va lues are considered conservative in the sense that the new ana lysis result provides less margin to applicable limits for making potential physical or procedure changes without a license/Coe amendment.

In contrast, if the use of a modified MOE resulted in a change in calculated cask peak pressure from 45 psig to 40 psig, this wou ld be a non-conservative change. That is because the change wou ld result in more margin being avai lable (to the design basis limit of 50 psig) for the licensee to make more significant changes to the physical ISFSI facility, cask design , or procedures.

6.8.J.2 "Essentially the Same" Results Licensees or cask CoC holders may change one or more elements of an MOE such that results move slightly in the non-conservative direction without prior NRC approval ,

provided the revised result is "essentiall y the same" as the previous result. Results are "essentia lly the same" if they are within the margin of error for the type of analysis being performed. Variation in results due to routine analysis sensitivities or calculational differences (e.g., rounding errors and use of different computational platforms) would typica ll y be within the ana lysis margin of error and thus considered "essentially the same." For example, when an MOE is applied using a different computational platform (mainframe vs. workstation), results of cases run on the two platforms differed by less than l %, which is the margin of error for this type of calculation . Thus the results are essentially the same, and do not constitute a departure from an MOE that requires prior NRC approval.

The determination of whether a new analysis result wou ld be considered "essentially the same" as the previous result can be made through benchmarking the revised MOE to the existing one, or may be apparent from the nature of the differences between the MO Es.

When benchmarking a revised MOE to determine how it compares to the previous one, the analyses that are done must be for the same set of conditions to ensure that the results are comparable, and the revised MOE should only be used where the benchmarking has demonstrated it to be conservative or essentially the same. Comparison of ana lysis MO Es should consider both the peak va lues and time behavior of results, and engineering judgment shou ld be app lied in determining whether two MOEs yield results that are essential ly the same.

6.8.2 Guidance for Changing from One Method of Evaluation to Another 6.8.2.1 NRC Approval of Methods The definition of"departure ... "provides licensees and CoC holders with the flexibility to make changes under I 0 CFR 72.48 from one MOE to another provided that the new MOE is approved by the NRC for the intended app lication. A new MOE is approved by the NRC for intended application if it is approved for the type of analysis (Definition 3.22) being conducted, and the app licab le terms, conditions and limitations for its use as defined in the Safety Evaluation Report and FSAR are satisfied.

Methodologies published by the NRC in NUREGs or NUREG-CRs are not necessarily "approved by the NRC for the intended application" because they may lack suitable ISFSI- or cask-specific context. In order to be considered "approved by the NRC for the intended application," such methods must be approved in an SER or otherwise accepted by the NRC as part of the ISFSI facility's or cask's licensing basis. 1 NRC approva l of an MOE wou ld typica ll y fo llow one of two paths. Some licensees and CoC holders may prepare and obtain NRC approval of topical reports that describe MO Es for the performance of a given type or class of analysis (Definitions 3.22 and 3.7). Through a SER, the NRC wou ld approve the use of the MOEs for a given class of ISFSls or spent fuel storage casks. In some cases, the NRC would accord .:.: generic~

approval of analysis MOEs. Terms, conditions and limitations relating to the app lication of the MO Es wou ld usually be documented in the topical reports, the SER, and correspondence between the NRC and the MOE owner that is referenced in the SER or associated correspondence.

The second path is the acceptancepproval of a specific ana lysis as part of a licensing action rather than generic approval of arr more generic MOE. In these cases, the NRC's acceptanceapproval would typically be part of an ISFSI or cask design's licensing basis and limited to a given ISFSI or spent fuel storage cask design and a given application.

Again, a thorough understanding of the terms, conditions and limitations relating to the application of the MOE is essential. This information should be documented in the origina l license or CoC application or license or CoC amendment request, the SER, and any correspondence between the NRC a nd the MOE owner that is referenced in the SER or associated correspondence.

Regardless of the level ofNRC review/approval for a given analysis methodology, Mmethods of evaluation, to the extent they are described in the ISFSI UFSAR or the generic cask UFSAR (i.e., the current Iicensing basis) are considered approved by the NRC for the intended application (Reference 3, Questions E.22 and E.23), but not generically approved for use by all ISFSI licensees or CoC holders.use in analyzing the design described in the UFSAR, *whether or not the SER states e>cplicitly that the method of evaluation is approved.

1 This, and the following discussion is taken from NEI's Q&A with RC staff on I0 CFR 50.59, Update 4, dated April 200 I (Reference 3). Speci lically, Questions E.16, E.19, E.22, E.23, and G.5 .

This is beeause the NRG approves the UFSAR *Nithout restrietion when they approve the ISFSI or eask design. Furthermore, the 72.48 program is founded upon the information in the UFSAR and the MOEs deseribed in the UFSAR are used to evaluate other ehanges, sueh as to the eask design, input parameters, or operating proeedures. In this respeet, a CoC holder's UFSAR for a generie eask design would funetion similar to a topieal report for lieensees using the MOE or evaluating ehanges to the MOE deseribed in the UFSAR for the eask design approved in the UFSAR. Use of an MOE for a speeifie eask design or lSFSI faeility deseribed in the assoeiated UFSAR for a different eask design or ISFSI faeility depends greatly on the *.vording of the SER.

In sueh eases, use of an MOE in *.vhieh the SER does not e>tplieitly approve or diseuss the aeeeptability of the MOE would not be permitted.6.8.2.2 Statements of Consideration The Statements of Consideration (SOC) for the l 999 rulemaking involving l 0 CFR 50.59 and I 0 CFR 72.48 were considered in developing this guidance. While the SOC are not considered patt of the 10 CFR 50.59 or 72.48 regulations, both NEI 96-07, Revision I and NE! 12-04 are largely consistent with them. Based on extensive public discussions and comment resolution between the industry and NRC staff, two aspects of the SOC were clarified in NE! 96-07, Revision I and endorsed by the NRC in RG I. I 87. NEI l 2-04's predecessor document, NEI 96-07, Appendix B included these same clarifications, which were endorsed in RG 3.72 and are further enhanced here (Reference 3, Question G.5):

First, I 0 CFR 54.21 (d) requires that the UFSAR be supplemented for license renewal with summary descriptions of time-limited aging analyses (TLAAs) and aging management programs (AMPs). The SOC state that changes to this license renewal information requires "evaluation" under I 0 CFR 50.59(c)(viii). The intent of the SOC discussion was to include TLAA and (as applicable) AMPs within the scope of "design bases and safety analyses" for purposes of I 0 CFR 50.59 criterion 8 so that if associated evaluation methods were described in the UFSAR they would fall within the definition of "methods of evaluation" and thus, the scope of l 0 CFR 50.59. The industry guidance reflects that all changes subject to I 0 CFR 50.59 may first be screened to determine if evaluation against the eight criteria of I 0 CFR 50.59 is required. Thus, contrary to the SOC, changes to TLAAs and AMPs for license renewal that screen out based on the definitions and guidance in NEI 96-07, Revision I, do not require evaluation under I 0 CFR 50.59. The guidance in NEI l 2-04 adopts a similar 1

clarification to the SOC for I 0 CFR 72.48 and lSFSl license/cask CoC renewals

• Second, the SOC state that licensees may adopt a new method of evaluation only if it has been specifically approved by the NRC for the intended plant/application or enjoys

" generic" NRC approval. In addition to these cases, NEI 96-07, Revision I, provides that licensees qualified per Generic Letter 83-l l, Supplement l, to perform safety analyses may adopt methodologies approved by the NRC for other plants provided the methodology is technically appropriate for the application. A similar clarification to the SOC is proposed for 10 CFR 72.

lt is i1rn1;1mb©Rt MfHrn th© MS©r ©fa R©'i'i MOJ;; ©V©R 1111 g@R©ri@ally 8f'lf'lFEIV©i by Hrn "HRC t© ©J'ISMrn tl'l©Y 1'18\'1 a drnrnugh MRi©rst1rn1ing Elf th© MOJ;; ifl 1M©Sti©H, th© t©rms

ef its @Histing a1313lieatin1 and eenditirns/lin1itatiens en its Mse. A range ef eensi1eratiens is i1entit:ied helew that may 8e a13vlieahle te determining whether new MOEs are teelrnieally a13fffl13riate fur the intended av13lieatien. The lieensee/CeC helder shelild address these and similar eensideratiens, as a1313liea8le, and deeMment in the IQ CFR 72.~8 evalMatien the hasis fur determining that an MOE is a1313re13riate and a13vrnved ffir the intended av13lieatien. Te e8tain an ade!Mate t1nderstanding ef the MOE and hasis fur determining it is a13prnved fur Mse in drn intended ap13lieatien, lieensees er CeC helders sheMld eensMlt varieMs seMrees, as a1313rn13riate. These inelt1de

£ERs, te13ieal rn13erts, lieensee eerres131rndenee with the MRC and lieensee er CeC l-rnlder 13ersennel familiar with the @Histing a1313lieatien efthe MOK If ade!Mate infurmatien eannet he fuMnd en whieh te hase the intended a1313lieatien ef the M06, the MOE she!ild net l:rn eensidered "a1313rnved by the l'JRC fur the intended a13plieatien."

When eensidering the applieatien efa MD6, it is 1rneessary te ade13t tl1e MOE in its entirety aHd a13ply it eensistent with a1313lieahle terms, eenditiens and liniitatiens.

~.fo{iflg attrihMtes ef new and eJ£isting MOEs is eensidernd a ehange te an ele111ent efan MOE and 111Mst he evalMated as Slieh 13er the g!iidanee in £eetien 0.8.1.

6.8.2.3+ Considerations for Determining if New or Different MO Es may be Considered "Approved by the NRC for the Intended Application"

The guidance for determining when a new or different MOE, which has not been generically approved by the NRC, is "approved for the intended application," per criterion (viii) of I 0 CFR 72.48, would include the following restrictions:

• The new or different MOE must be used only by the same design authority as the originally approved MOE.

• There must be a Generic Letter (GL) 83-11, Supplement I-type qualification process for the user of the new method (Reference 2). (Note that G L 83-1 I is intended for power plant licensees who wish to perform safety analyses that had customarily been performed by nuclear steam supply system vendors. Cask design authorities should have this type of qualification process embedded in their NRC-approved QA programs required as part of the cask design approval.)

• The licensee or CoC holder is able to identify all conditions and limitations under which the MOE received NRC approval. Conditions and limitations include, among other things, the type of analysis, the manner in which the analysis was applied, the physical configuration of the ISF SI facility or cask design, and any licensing basis restrictions. The licensee or CoC holder must assure that all limitations and restrictions are applicable to the ISFSI facility or cask design in question and the MOE is then applied within all conditions and limitations. For example:

o An MOE that is approved for one JSFSI facility or cask design for a particular transient cannot be applied to another lSFSl facility or cask design for a different transient because it was not approved for the other transient.

o A licensee or CoC holder is not permitted to adopt different aspects of different approved MOEs (i.e., mix and match) because the MOE is not being applied in the manner it was approved.

o An MOE that is approved for a particular ISFSI facility or cask design cannot be applied to another ISFSI facility or cask design that has relevant design differences because the method was not approved for the different design configuration.

• The licensee or CoC holder is not permitted to adopt less restrictive licensing basis analysis assumptions even if they have been approved for a different ISFSI facility or cask design (e.g., it is not acceptable to credit not-important-to-safety components in accident analyses, or assume different iodine partitioning even if it is approved for another ISFSI facility or cask design).

6.8.2.4 Additional Considerations

The following questions highlight important additional considerations for determining that a particular application of a different MOE is technically appropriate for the intended application, within the bounds of what has been found acceptable by NRC ,

and does not require prior NRC approval.

• Is the application of the MOE consistent with the ISFSI facility ' s or cask design's licensing basis (e.g. , NUREG-1536, NUREG-1567 , or other ISFSI or cask design-specific commitments)? Will the MOE supersede an MOE addressed by other regulations or the ISFSI or cask Technical Specifications?

ls the MOE consistent with relevant industry standards?

If application of the new MOE requires exemptions from regulations or ISFSI- or cask-specific commitments, exceptions to relevant industry standards and guidelines, or is otherwise inconsistent with an ISFSI facility 's or cask ' s licensing basis, then prior NRC approval may be required. The applicable change process must be followed to make the ISFSI facility 's or cask' s licensing basis consistent with the requirements of the new MOE.

• If a computer code is involved , has the code been installed in accordance with applicable software Quality Assurance requirements? Has the ISFSI- or cask design-specific model been adequately qualified through benchmark comparisons against test data, empirical data , or approved engineering analyses? ls the application consistent with the capabilities and limitations of the computer code? Has industry experience with the computer code been appropriately considered?

The computer code installation and ISFSI or cask design-specific model qualification is not directly transferable from one organization to another. The installation and qualification should be in accordance with the licensee ' s or cask CoC holder's Quality Assurance program.

• Is the ISFSI facility or cask design for which the MOE has been approved designed and operated in the same manner as the ISFSI facility or cask design to which the MOE is to be applied? Is the relevant equipment the same? Does the equipment have the same pedigree? Are the relevant failure modes and effects analyses the same? If the lSFSl facility or cask design is designed and operated in a similar, but not identical, manner, the following types of considerations should be addressed to assess the applicability of the MOE:

- How could those differences affect the MOE?

- Are additional sensitivity studies required?

- Should additional single failure scenarios be considered?

- Are analyses of limiting scenarios, effects of equipment failures, etc. , applicable for the specific ISFSI or cask design ?

- Can analyses be made while maintaining comp liance with both the intent and literal definition of the MOE?

• Differences in the ISFSI or cask design configurations and licensing bases could invalidate the app lication of a particular MOE. For example, the licensing basis of older vintage cask designs may not have been required to consider the same isotopes for offsite dose calcu lations as those in the licensing basis for more recent vintage cask designs. The existence of these differences does not preclude app licatio n of a new MOE to an ISFSI facility or cask design ; however, differences must be identified, understood and the basis documented for concluding that the differences are not re levant to determining that the new app li cation is technically appropriate.

References to be added to NEI 12-04:

I. USNRC Memorandum from E. McKenna to C. Carpenter, " Summary of November 2, 1999 meeting with the Nuclear Energy Institute (NE!) on Revision to NEI 96-07 on Implementation of 10 CFR 50.59 - Methods of Evaluation," dated November 10, 1999 (ML993260078).

2. USNRC Generic Letter 83-1 l, "Licensee Qualification for Performing Safety Analyses in Support of Licensing Actions," dated February 8, 1983 and Supplement I, dated June 24, 1999.

3. Questions and Answers on 10 CFR 50.59 and NEI 96-07, Revision l, Update 4 - April 2001 (provided).