Skip to content

Using the EasyVVUQ Uncertainty Quantification package with the PROCESS fusion power plant systems code.

Notifications You must be signed in to change notification settings

ym1906/easyVVUQ-process

 
 

Repository files navigation

easyVVUQ-PROCESS: Feasibility Uncertainty Quantification (UQ)

A feasibility study using the PROCESS nuclear fusion power plant systems code and the easyVVUQ framework. It aims to assess the feasibility of two different power plant concepts under epistemic uncertainty.

A DEMO-like design is optimised using PROCESS for maximum net electric output, and a second design optimised for minimum major radius. UQ studies are then performed on these two design points to assess their comparative feasibilities under uncertainty.

Installation

It is recommended to use a virtual environment. Firstly, Process needs to be installed. Clone Process

git clone https://github.com/ukaea/PROCESS/process-uq.git

Then install Process:

cd process-uq
cmake -S . -B build
cmake --build build

Secondly, clone this repository and install the infeas package:

cd ..
git clone https://github.com/jonmaddock/process-uq.git
cd process-uq
pip install .

Whether installed locally or on HPC, this should allow any of the notebooks in this repository to run.

Using easyVVUQ to parallelise runs of PROCESS

To summarise the process, you need to create:

  1. A job script which sumbits the job and creates the worker nodes.
  2. A notebook which designs the campaign.
  3. PROCESS inputs and outputs (create an IN.DAT, IN.DAT.template (copy of IN.DAT but with the uncertainties commented in $apsect format))

About

Using the EasyVVUQ Uncertainty Quantification package with the PROCESS fusion power plant systems code.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 80.4%
  • Python 18.1%
  • Shell 1.5%