Verification and Validation Strategy for LWRS Tools

Abstract

SUMMARYThe US Department of Energy Light Water Reactor Sustainability (LWRS) Program is focused on enabling the long-term operation of US commercial power plants. Commercial power plant owners will make the decisions on life extension -the information provided by the research and development activities in the LWRS Program will reduce the uncertainty (and therefore the risk) associated with making those decisions.Modeling and simulation activities are an important part of the LWRS Program. Advanced tools that represent more accurately (when compared with currently used tools) the behavior of nuclear power plants, in particular as they age, can be an important tool to inform life extension decisions, and could potentially be used in licensing activities. However, before these tools can be used to inform decisions or support licensing activities, they will need to be adequately verified and validated (V&V'ed). The acceptable level of V&V is ultimately the decision of the nuclear power plant owners (who would use the tools to inform life-extension decisions) and the NRC (if the tools are used to support licensing). Although this report is focused on the LWRS Program, it is broadly applicable to advanced modeling and simulation activities.Significant and continuing advances in computer simulation and rising costs of building and conducting test facilities are increasing the reliance on complex models in licensing facilities, optimizing designs, improving performance, and understanding the underlying science. At the same time, however, the complexity of these models and associated computational methods present unprecedented challenges for code verification, validation, and uncertainty quantification (VVUQ).In general, the VVUQ Strategy should:1. Be driven by the needs of the users (for this program, the LWR industry) 2. Distinguish between and address the needs of licensing, design and performance, and aging/life extension 3. Demonstrate the effectiveness of the methodology by executing one or more well-defined problems early in the process to establish good practices and instill confidence within the user community 4. Utilize a rigorous and integrated approach to code development and VVUQ that is accepted by industry (to inform life-extension decisions) and regulators (if the tools are used to support licensing) 5. Recognize the complementary nature of system and multiphysics/multiscale codes, addressing the unique challenges in VVUQ including gaps in understanding (physical models) and coupling of phenomena across physics, time and length scales 6. Acknowledge and identify gaps in data and re-establish the pedigree of legacy data, exploiting it in a cost-effective manner.iv v CONTENTS