Skip to content

Troubleshooting

Setup debugger

To test the workflow using Snakemake rules, a configuration file launch.json with the following content is required:

Debugger configurations
{
    "version": "0.2.0",
    "configurations": [
        {
            "name": "mock snakemake",
            "type": "python",
            "request": "launch",
            "program": "${file}",
            "console": "integratedTerminal",
            "justMyCode": false,
            "cwd": "${cwd}${pathSeparator}scripts"
        }
    ]
}

Once this file is set up, you can easily run any Python file in the scripts folder under debugger mode for testing.

Infeasibility issues

If your model is infeasible or unbounded, it often means that your input settings are leading to a situation where PyPSA can’t solve one or more objective functions. Common causes are:

  • Insufficient generator capacity at a bus to meet the load across all snapshots.
  • Must-run generators (p_min_pu) produce more power than the load at certain snapshots, causing excess power (i.e., power dumping).
  • Negative values assigned to capacity parameters.
  • Maximum capacity (p_nom_max) is smaller than minimum capacity (p_nom_min).
  • Storage units has inflow and cyclic charging (cyclic_state_of_charge = True) but no discharging capability.

Tips for diagnosing the problem

Try the following steps to identify and fix the issue:

  • Run pypsa.network.consistency_check() to check if any warnings appear.
  • Temporarily disable custom features or user extensions to isolate the cause.
  • Ensure load-shedding generators are added to every bus.
  • Check pypsa.network.generators.p_min_pu to identify any must-run generators. Then verify if their generation exceeds the corresponding load in pypsa.network.loads_t.p_set.
  • Compare the p_nom_max and p_nom_min values for each component to ensure they make sense (i.e., max ≥ min).
  • Try disabling cyclic charging for storage units: pypsa.network.storage_units.cyclic_state_of_charge = False.
  • Double-check for any negative values in:
    • pypsa.network.{component}.p_nom
    • pypsa.network.{component}.p_nom_max
    • pypsa.network.{component}.p_nom_min

More detailed information

PyPSA also provides official guidance on solving infeasiblity issues. You can also explore the link to seek for a good solution.