Upload
phungcong
View
218
Download
1
Embed Size (px)
Citation preview
Robert Brient
Quality Assurance Specialist
Center for Nuclear Waste Regulatory Analyses
Southwest Research Institute®
San Antonio, Texas, USA
Presentation Outline
Review Objective
Review Approach
Code Documentation and QA Reviews
Review Findings
SKB Code QA Program Review and Findings
Conclusions
Back-Up Slides
Preliminary Initial Review Findings, 2012 May 21–23 2
Review Objective
Determine whether SKB code documentation
and quality assurance (QA) practices are
adequate to provide sufficient confidence that the
SR–Site is supported by valid and accurate
modeling and calculations.
Determine whether additional information
regarding SKB code documentation and QA
practices should be provided by SKB for
subsequent SR–Site reviews.
Preliminary Initial Review Findings, 2012 May 21–23 3
Review Approach
Code documentation and QA reviews
— General review of all codes used in
SR–Site
— Detailed review of selected codes
Supplemental review of SKB code QA
program requirements
Preliminary Initial Review Findings, 2012 May 21–23 4
Code Documentation and QA Reviews General Review
— Primary Source: SR–Site Model Summary
Report TR–10–51
— There are 22 codes described, covering topics such as code
suitability; usage; development process and verification; and
passing data between models
— Codes are categorized by source for applying graded
quality assurance
– Commercial (10 codes)
– Modified-commercial (3 codes)
– Project-specific (11 codes)
– Several codes are in more than one category
Preliminary Initial Review Findings, 2012 May 21–23 5
Code Documentation and QA Reviews
(continued)
Detailed Reviews
— Cross-section of codes in various categories selected
— Code_Bright, CONNECTFLOW, 3DEC, and MATLAB–FPI
— Model Summary Report references are primary sources for
detailed reviews
Preliminary Initial Review Findings, 2012 May 21–23 6
Review Findings
Information regarding suitability of the codes for their
SR–Site applications appears to be adequate
User’s manuals are available where appropriate
Information for the development process and verification
varies widely
— Software standards applied to development
are identified for Abaqus and ConnectFlow
(commercial) and MARFA (project specific);
none mentioned for other codes
Preliminary Initial Review Findings, 2012 May 21–23 7
Review Findings (continued)
— Model Summary Report identifies comprehensive
verification and validation documents for 3DEC, Abaqus,
CODE_BRIGHT, ConnectFlow, DarcyTools, FARF31,
MARFA, MATLAB–COMP23, TOUGHREACT
— PHAST and PHREEQC descriptions indicate extensive
testing, but without references (documentation is
available at U.S. Geological Survey websites)
— Other codes had references for verification activities
Data exchange information provided is adequate for
all codes
Preliminary Initial Review Findings, 2012 May 21–23 8
SKB Code QA Program Review
and Findings
Based on the variability in code documentation and QA
among codes noted in the general and detailed reviews, the
reviewers examined SKB code requirements.
— SKB requirements focus on the Model Summary Report
format and content rather than directly providing
requirements for codes, particularly for the verification
and validation necessary to demonstrate that codes
produce correct results for the purpose for which they are
being used.
Preliminary Initial Review Findings, 2012 May 21–23 9
SKB Code QA Program Review
and Findings (continued)
— SKB applies code development and verification requirements only to modified commercial and project-specific codes. However, the Model Summary Report has information for commercial codes as well. Verification information, in particular, is important for commercial codes and should be developed where needed.
— SKB requirements are much less detailed than recognized software QA standards (i.e., NQA–1, Subpart 2.7 or ISO 9001). Additional guidance in SKB instructions may reduce variability and enhance overall confidence in codes producing correct solutions.
Preliminary Initial Review Findings, 2012 May 21–23 10
Conclusions
Confidence in SKB code documentation
and QA practices could be enhanced with
comprehensive verification and validation
documents for each code, including
commercial codes. DarcyTools verification
and validation is an excellent example.
SKB instructions may need to be
revised for additional detail in order to be
effectively implemented
Preliminary Initial Review Findings, 2012 May 21–23 11
Backup Slides
12 Preliminary Initial Review Findings, 2012 May 21–23
Observations: Commercial Codes
Code Proper Usage
Development Process
Followed Appropriate
Procedures
Produces Accurate Results
ABAQUS User’s manual ISO 9001, U.S. Nuclear QA Information on web site
CODE_BRIGHT User’s manual References in MSR
ConnectFlow User’s manual ISO 9001, TickIT References in MSR
Ecolego User’s manual References in MSR
ERICA Tool User’s manual Affirmative statement, but no
references in MSR
MATLAB None identified in Model
Summary Report (MSR), but
User’s manual found in web
search
No references in MSR, but
wide-user base should identify
and correct errors
MIKE SHE User’s manual References on DHI web site
PHAST User’s manual affirmative statement, but no
references in MSR
PHREEQC User’s manual Affirmative statement, but no
references in MSR
TOUGHREACT User’s manual Affirmative statement, but no
references in MSR
Preliminary Initial Review Findings, 2012 May 21–23 13
Observations: Modified Commercial Codes
Code Proper Usage
Development Process Followed
Appropriate Procedures
Produces
Accurate Results
3DEC User’s manual includes coverage
of the FISH language for
modifications
Development process for
modifications not identified in the
Model Summary Report (MSR)
References in MSR
ABAQUS Available user guides are likely to
cover user-defined subroutines
Development process for
modifications not identified in
MSR
User defined subroutines verified
by using simple test examples, no
references provided in MSR.
CODE_BRIGHT‡ CODE_BRIGHT user’s manual
does not appear to cover
modifications
Development process for
modifications not identified in
MSR
References in MSR
Preliminary Initial Review Findings, 2012 May 21–23 14
Observations: Project-Specific Codes
Code Proper Usage
Development Process Followed
Appropriate Procedures Produces Accurate Results
Analytical model for quantification
of buffer erosion and canister
corrosion
Simple user instructions included
in the file
Excel-based code should not
require a formal development
process
References in MSR*
DarcyTools User’s manual MSR indicates no attempt has
been made to show that
DarcyTools conforms to any
international QA standard
References in MSR provide
comprehensive documentation of
verification, validation, and
demonstration
FARF31 User’s manual Development procedures not
identified in MSR
References in MSR includes
validity document, regression
tests of code changes
MARFA User’s manual NQA–1–2000 based software QA
program
Validation tests in user’s manual,
reference in MSR
MATLAB–COMP23 User’s manual Development procedures not
identified in MSR
References in MSR includes
validity document
MATLAB–FPI No user’s manual mentioned in
MSR
Development procedures not
identified in MSR
References in MSR
MATLAB–Pandora Internal user’s manual, code is
documented in a reference
Development procedures not
identified in MSR
References in MSR
Numerical GIA model No user’s manual, but users work
closely with developer
Development procedures not
identified in MSR
References in MSR
Numerical permafrost model No user’s manual, but
descriptions of the code are
referenced
Development procedures not
identified in MSR
References in MSR
Solubility model (Simple
Functions)
No user’s manual, but
presentation provides instruction
on spreadsheet use
Excel-based code should not
require a formal development
process
References in MSR
UMISM No user’s manual, but MSR
indicates close collaboration of
users with developer
Development procedures not
identified in MSR
References in MSR
Preliminary Initial Review Findings, 2012 May 21–23 15