Run OPM Flow from Python

To run OPM from Python, you need to:

  1. Compile Flow with python support:

  • Add the cmake flags -DOPM_ENABLE_PYTHON=ON and -DOPM_INSTALL_PYTHON=ON.

  • Optionally add prefix -DCMAKE_INSTALL_PREFIX=/opt/opm to install outside the standard directories.

  • Optionally specify python binary -DPython3_EXECUTABLE=/home/user/miniconda3/envs/rkt/bin/python3 if you don’t want to use the system python, e.g. use a python from pyenv or from a conda environment.

Sample compilation on linux:

#! /bin/bash

flags="-DPython3_EXECUTABLE=/home/hakon/miniconda3/envs/rkt/bin/python3 -DOPM_ENABLE_PYTHON=ON -DOPM_INSTALL_PYTHON=ON -DCMAKE_INSTALL_PREFIX=/opt/opm"
for repo in opm-common opm-grid opm-models opm-simulators
do
    cd "$repo"
    mkdir -p build
    cd build
    cmake  $flags ..
    make -j8
    sudo make install
    cd ..
    cd ..
done

  1. Now you should be able to use the module from a Python script.

If you installed in a non-standard directory by specifying -DCMAKE_INSTALL_PREFIX you may need to set the PYTHONPATH environment variable before running your Python script, for example:

PYTHONPATH=/opt/opm/lib/python3.11/site-packages python3 spe1case1.py

Here, spe1case1.py could be:

import os
from opm.simulators import BlackOilSimulator
from opm.io.parser import Parser
from opm.io.ecl_state import EclipseState
from opm.io.schedule import Schedule
from opm.io.summary import SummaryConfig

os.chdir("SPE1CASE1")
deck  = Parser().parse('SPE1CASE1.DATA')
state = EclipseState(deck)
schedule = Schedule( deck, state )
summary_config = SummaryConfig(deck, state, schedule)

sim = BlackOilSimulator(deck, state, schedule, summary_config)
sim.step_init()
sim.step()
poro = sim.get_porosity()
poro = poro *.95
sim.set_porosity(poro)
sim.step()
sim.step_cleanup()