FINAL REPORT Project Title: Advanced Variance Reduction for Global k-Eigenvalue Simulations in MCNP Award Number: DE-FG07-04ID14608 Covering Period: September 13, 2004 through September 12, 2007 Date of Report: May 16, 2008 Recipient: University of Michigan Department of Nuclear Engineering and Radiological Sciences Cooley Building, North Campus Ann Arbor, MI 48109-2104 Subcontractors: None Other Partners: None Contacts: Edward W. Larsen (PI) (734) 936-0124 edlarsen@umich.edu William R. Martin (co-PI) (734) 764-5534 wrm@umich.edu Project Team: DOE-HQ contact: Nancy Elizondo, elizonna@id.doe.gov Contract specialist: Kate Koorhan, kkoorhan@umich.edu Accountant: Shannon Thomas, noonie@umich.edu Project Objectives: • Develop advanced variance reduction techniques, based on the Variational Variance Reduc- tion (VVR) methodology, for improving the efficiency of global reactor calculations, and incorporate these results into the production Monte Carlo code (MCNP5). • Implement this new version of MCNP5 on a Linux cluster at the University of Michigan, in order to process these histories faster than presently possible on a typical workstation. • Test this methodology for realistic problems. 1