Skip to content
/ endorse Public

stochastic prediction of a safety indicator of excavation damage zone for a deep geological repository


Notifications You must be signed in to change notification settings


Repository files navigation

Endorse - EDZ safety indicator simulations

The software implements specialized safety calculations for the excavation disturbed zone (EDZ) of a deep repository of radioactive waste. It consists of two parts:

  1. determination of the rock parameters using the Bayesian inversion
  2. stochastic prediction of the contamination transport and safety indicator evaluation

It essentially use the simulator Flow123d of processes in fractured rocks.


The software requires a working Docker Desktop installation or SingularityCE installation. The first is better for local desktop usage, while the latter is usually the only option on HPC clusters. The use of clusters is recommended, as stochastic simulations are pretty computationally demanding. Currently, only the Linux installations are tested but should run with little effort on Windows due to containerization.

Quick start

  1. Download the latest version of the sources as a ZIP package.
  2. Extract to the directory of your choice.
  3. Set up the computational container with the proper environment using the bin/endorse-setup tool.
  4. Create a working directory on a filesystem shared between computational nodes.
  5. Prepare main configuration files.
  6. Run Bayes inversion (bin/endorse-bayes) or stochastic transport (bin/endorse-mlmc).

See full documentation for the details.


TACR logo Development of the Endorse software was supported by
Technological agency of Czech republic
in the project no. TK02010118 of the funding programme Theta.


Technical university of Liberec

  • Jan Březina coordination, stochastic transport
  • Jan Stebel hydro-mechanical model in Flow123d
  • Pavel Exner Bayes inversion for the EDZ
  • Martin Špetlík MLMC library and homogenization

Institute of Geonics

  • Stanislav Sysala plasticity model
  • Simona Bérešová core Bayes inversion library surrDAMH
  • David Horák, Jakub Kružík PERMON library integration for fracture contacts in Flow123d


  • David Flanderka Flow123d, optimizations, technicalities
  • Radek Srb containerization
  • Michal Béreš consultation, tests

Developers corner

Repository structure:

  • doc - software documentation and various reports from the Endorse project
  • experiments - various numerical experiments and developments as part of the Endorse project
  • src - main sources
  • tests - various software tests, test data

Development environment

In order to create the development environment run:

As the Docker remote interpreter is supported only in PyCharm Proffesional, we have to debug most of the code just with virtual environment and flow123d running in docker.

More complex tests should be run in the Docker image: flow123d/geomop-gnu:2.0.0 In the PyCharm (need Professional edition) use the Docker plugin, and configure the Python interpreter by add interpreter / On Docker ...


stochastic prediction of a safety indicator of excavation damage zone for a deep geological repository







No releases published


No packages published