Model Validation Status Review

Abstract

EXECUTIVE SUMMARYThis Model Validation Status Review (MVSR) technical report was prepared as part of the (amended) response to Corrective Action Request BSC-0 1 -D-00 1 (Clark 2001, Krisha 200 1). This report documents a systematic, one-time review of all the Analysis/Model Reports (AMRs) that support the Total System Performance Assessment -Site Recommendation (TSPA-SR), by an independent team under the direction of the Bechtel SAIC Company Chief Science Officer.A total of 125 AMRs plus other documents and data were reviewed. The review team first identified the models documented in the AMRs. Then they determined how the models were used in development of the TSPA-SR. Finally, the models were assigned to three categories ("bins") according to the extent to which model validation was achieved in compliance with Administrative Procedure AP-3.1 OQ, Analyses and Models. Review-team recommendations are provided for improvement of documentation, and in some cases, for fbrther model development, testing, or data collection to improve the models in support of licensing activities.A total of 128 models were identified, of which 17 were assigned to Bin 1 (compliant validation), 77 were assigned to Bin 2 (information readily available for validation but documentation not in compliance), and 34 were assigned to Bin 3 (not in compliance and not readily validated). For the 34 Bin-3 models, impact reviews were conducted and the resulting impact review documentation is summarized in Section 6, and included in Appendix IV to this technical report. For all 34 Bin-3 models, the impact reviews found no significant impact from the model validation review findings, on the conclusions of the TSPA-SR (i.e., the calculated annual dose is negligible). Impact Review Summary for Bin-3 ModelsOf the 34 Bin-3 models, the impact reviews for 24 can be summarized as follows:Sixteen were found not to support TSPA-SR, i.e., the output from these models was not used as input to the system model. All 16 of these models are developmental, and some have been superseded by other products. Of these 16 models, only six were used for screening out of features, events, and processes (FEPs) from the TSPA-SR.Five additional Bin-3 models (not among the 16 discussed above) were originally intended by the authors to be analyses or calculations (i.e., not requiring validation). At the time this work was documented, the authors determined that model validation was not required based on consideration of details such as the analysis methods used and the contribution to TSPA-SR. The review team identified these as models requiring validation, using the criteria from Section 3.4.2 of this report.Three additional Bin-3 models in the Biosphere area were embedded within the GENII-S dose-assessment code and were not previously recognized as discrete models. (The GENII-S code is a widely used and accepted approach, but application to a potential Yucca Mountain geologic repository is not yet hlly documented.)TDR-WIS-MD-000005 REV 00Page vii November 200 1The remaining 10 ...