Verification of RESRAD-build computer code, version 3.1.

RESRAD-BUILD is a computer model for analyzing the radiological doses resulting from the remediation and occupancy of buildings contaminated with radioactive material. It is part of a family of codes that includes RESRAD, RESRAD-CHEM, RESRAD-RECYCLE, RESRAD-BASELINE, and RESRAD-ECORISK. The RESRAD-BUILD models were developed and codified by Argonne National Laboratory (ANL); version 1.5 of the code and the user's manual were publicly released in 1994. The original version of the code was written for the Microsoft DOS operating system. However, subsequent versions of the code were written for the Microsoft Windows operating system. The purpose of the present verification task (which includes validation as defined in the standard) is to provide an independent review of the latest version of RESRAD-BUILD under the guidance provided by ANSI/ANS-10.4 for verification and validation of existing computer programs. This approach consists of a posteriori V&V review which takes advantage of available program development products as well as user experience. The purpose, as specified in ANSI/ANS-10.4, is to determine whether the program produces valid responses when used to analyze problems within a specific domain of applications, and to document the level of verification. The culmination of these efforts is the production of this formal Verification Report. The first step in performing the verification of an existing program was the preparation of a Verification Review Plan. The review plan consisted of identifying: Reason(s) why a posteriori verification is to be performed; Scope and objectives for the level of verification selected; Development products to be used for the review; Availability and use of user experience; and Actions to be taken to supplement missing or unavailable development products. The purpose, scope and objectives for the level of verification selected are described in this section of the Verification Report. The development products that were used for the review and any actions that were taken when these items were missing are documented in Section 5 of this report. The availability and use of user experience were limited to extensive experience in performing RESRAD-BUILD calculations by the verification project manager and by participation in the RESRAD-BUILD workshop offered by the code developers on May 11, 2001. The level of a posteriori verification that was implemented is defined in Sections 2 through 4 of this report. In general, a rigorous verification review plan addresses program requirements, design, coding, documentation, test coverage, and evaluation of test results. The scope of the RESRAD-BUILD verification is to focus primarily on program requirements, documentation, testing and evaluation. Detailed program design and source code review would be warranted only in those cases when the evaluation of test results and user experience revealed possible problems in these areas. The verification tasks were conducted in three parts and were applied to version 3.1 of the RESRAD-BUILD code and the final version of the user.s manual, issued in November 2001 (Yu et al. 2001). These parts include the verification of the deterministic models used in RESRAD-BUILD (Section 2), the verification of the uncertainty analysis model included in RESRAD-BUILD (Section 3), and recommendations for improvement of the RESRAD-BUILD user interface, including evaluations of the user's manual, code design, and calculation methodology (Section 4). Any verification issues that were identified were promptly communicated to the RESRAD-BUILD development team, in particular those that arose from the database and parameter verification tasks. This allowed the developers to start implementing necessary database or coding changes well before this final report was issued