Whole Core Reactor High Fidelity Calculations using the RAPID Code System

Nuclear reactors require routine modeling for determining safety margins, core configurations, and other optimizations. These models are typically simulated with a Monte Carlo based code, or a deterministic code based on the Linear Boltzmann Equation (LBE) or transport equation. The problem with these solutions arises when high-fidelity results are necessary. Monte Carlo based codes are able to obtain high fidelity results, but they require a significant amount of computational resources (e.g., processors, memory) and time. The deterministic codes can also produce high fidelity results, but due to the discretization of variables the resources required can exceed that of Monte Carlo based codes.

The RAPID code has been developed based on the Multistage Response function Transport (MRT) methodology. In this methodology, a problem is partitioned into stages or sub-problems that can be simulated using response functions or coefficients. By pre-calculating these functions/coefficients as a function of different parameters using a Monte Carlo code, then solutions are obtained using a liner system of equations in seconds or minutes on one computer core. The RAPID code system has been verified computationally and validated using benchmark problems, and experimentally using the Jo˘zef Stefan Institute (JSI) TRIGA reactor in Slovenia.