Rolls-Royce has embarked on a Small Modular Reactor (SMR) design programme, which is developing a safe, deployable and affordable Nuclear Power Plant. An element of the work undertaken is the determination of the performance of nuclear reactor plants through the application of thermal hydraulic analysis methods. These methods are used throughout the product lifecycle; beginning with exploring the performance potential of design options and ultimately providing analysis to support the safety justification.
To provide performance analysis at the earliest stages of design it is imperative to have a flexible analysis method that evolves alongside the design process while being able to inform the system architects at each phase. To achieve this goal an analysis method has been developed that allows for varying plant configurations with adjustable plant component geometries. The process of updating each of the plant component subsystem models has been automated and linked with an overarching database that allows system designers to directly update geometrical parameters.
This approach to early concept design has led to the practice of two-stage verification and validation. The first stage is the traditional understanding of V&V, the verification of the models and supporting scripts through the use of a predefined rigorous test suite which encompasses unit tests through to full system integration tests.
The second stage is data verification and validation, as the method has the ability to constantly update based on data input the verification of this input data is paramount. The design data is hierarchical, which in addition to a mean and an uncertainty distribution, requires knowledge of a transfer function.
Each item of data is owned by the department that is the expert in that field, for example the steam generator dimensions are owned by the components team. It is the responsibility of each of the data owners to verify their data and ensure that any transfer functions are implemented correctly.
As data is used in a cross discipline setting it requires a central storage location, this is the Master Parameter List (MPL). This repository defines the data, its format, its owner, and the last time it was updated among other things. This repository is the single point of truth for every data item used in analysis.
The aim of this paper is to demonstrate the way in which a flexible analysis method is created along with the framework that provides confidence in the analysis results through a two phase data driven V&V approach.
